Author Topic: Tonemapping - Plz Halp  (Read 79818 times)

2020-03-11, 06:34:02

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
Bit of a rant - or a cry for help - but I'm either approaching lighting or tone mapping totally incorrectly; it is absolute guesswork every time, and rarely is it 'photorealistic'.

Looking at Juraj's work, Bertrand's work, etc. - there is a consistent look and quality with each project and I'm just not convinced (for now) that they are also, just guessing; I assume they have some sort of method - a procedure. For example, when I worked in a studio, we had a *very* strict method, and it produced a consistent result, but with Corona I feel like I'm guessing once I get to tone mapping.

In Ludvik's, Time to ditch sRGB/Linear as default (?) thread, it is mentioned:

Right now, we perceive linear sRGB as the default, the start line, and we then work with some parameters to bring that sRGB close to photo-realism. We manually have to twist some knobs in order to take a picture, which by default is not realistic to our eyes, and using some controls, turn it into image that our eyes perceive as photorealistic. So why not just skip this process and have renderer(s) by default output same ranges as cameras do. [...] There's no significant reason why renderer should not work the same way. Not by having a dropdown where you can pick numerous response curves, and one of them is called photorealistic, but instead by having it defaulting to a camera, with an option to switch to a very special mode, which will make your output less realistic, but compose-able in post.

This is a very attractive proposition and it speaks to the issue I am referring to. Regardless of whether such an implementation is feasible or not, I'd like to learn how to produce images that are of the quality we expect from the Corona Gallery and I just haven't seen strong tutorials - free or paid - demonstrating such an approach.
« Last Edit: 2020-04-26, 02:30:55 by cjwidd »

2020-03-11, 10:53:21
Reply #1

Jpjapers

  • Active Users
  • **
  • Posts: 1507
    • View Profile
Its not just tonemapping that makes those guys renders look amazing every time. Its attention to detail at every stage and questioning how something actually looks in reality vs what we imagine it looks like. Lacking some minute details is enough to make your brain subconciously question an images' realism so the answer to the question "How do i make it look like [insert artist here]" is usually "Hard work and practice" when you really break it down. I get what you mean though with regard to the consistency of some people images. I think its maybe to do with everyone having a preference of HDRI for a given look, combined with repeating and refining your preferred process over time.

I dont think you need to worry though. Your renders are always awesome!
« Last Edit: 2022-03-06, 22:55:51 by Jpjapers »

2020-03-11, 10:55:43
Reply #2

maru

  • Corona Team
  • Active Users
  • ****
  • Posts: 11392
  • Marcin
    • View Profile
1. "Turn linear sRGB into digital camera product"

Quote
having it defaulting to a camera, with an option to switch to a very special mode, which will make your output less realistic, but compose-able in post.

You can get this by rendering your image at the default settings, then enabling LUT and picking one of the 3 Kim Amland's photographic LUTs. They are captured from real life digital cameras.
Then you can optionally reduce the strength of the LUT to ~0,8 (to your own taste) and play with highlight compression and filmic values (again to your own taste, to reduce highlight burn).


2. "I want to produce images like the masters"

Quote
This is a very attractive proposition and it speaks to the issue I am referring to. Regardless of whether such an implementation is feasible or not, I'd like to learn how to produce images that are of the quality we expect from the Corona Gallery and I just haven't seen strong tutorials - free or paid - demonstrating such an approach.

I am afraid this is not as easy as just following one or two tutorials. Artists have their own secrets. Some of them definitely use Photoshop or other post tools to adjust images in addition to what is done in the VFB, not only by changing the overall tone mapping/coloring, but also locally, e.g. by making some areas more or less noisy, brighter/darker, more clear/blurry, etc.

Marcin Miodek | chaos-corona.com
Chaos Corona Support Team Lead | contact us

2020-03-11, 20:30:32
Reply #3

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@Jpjapers, thank you so much for your thoughtful and detailed response, and @Maru thank you for your advice - I really appreciate it :)

I think both of your responses more or less encapsulate the truth of the matter, which, as you say @Jpjapers - it boils down to hard work and practice. I want to be clear, I'm not grasping for a magical solution, I'm not pining for a "make this look good button", but I do imagine there are techniques - like using False Color LUTs, etc. - to help narrow the path toward an image that speaks to the quality of the 'masters'.

I would also say that I have **scoured** the forums for many of the relevant tools that have been shared here (e.g. Dubcats camera LUT dumps, fStorm LUTs, Adanmq false colour LUTs and log space filmic LUTs, specular to IOR cube (for materials - Quixel Bridge does this by default now), Dubcats ACES emulation recommendations, etc.) and gathered many high quality (?) HDRIs, including Jørgen Herland and Peter Guthrie.

I assume then, that if I am working with high quality tools, then the onus is on me to get the most from the tools which are already doing their part to help.

I'll take a look at the resources you linked @Jpjapers, they seem promising - thank you!

2020-03-11, 22:36:23
Reply #4

Jpjapers

  • Active Users
  • **
  • Posts: 1507
    • View Profile
@Jpjapers, thank you so much for your thoughtful and detailed response, and @Maru thank you for your advice - I really appreciate it :)

I think both of your responses more or less encapsulate the truth of the matter, which, as you say @Jpjapers - it boils down to hard work and practice. I want to be clear, I'm not grasping for a magical solution, I'm not pining for a "make this look good button", but I do imagine there are techniques - like using False Color LUTs, etc. - to help narrow the path toward an image that speaks to the quality of the 'masters'.

I would also say that I have **scoured** the forums for many of the relevant tools that have been shared here (e.g. Dubcats camera LUT dumps, fStorm LUTs, Adanmq false colour LUTs and log space filmic LUTs, specular to IOR cube (for materials - Quixel Bridge does this by default now), Dubcats ACES emulation recommendations, etc.) and gathered many high quality (?) HDRIs, including Jørgen Herland and Peter Guthrie.

I assume then, that if I am working with high quality tools, then the onus is on me to get the most from the tools which are already doing their part to help.

I'll take a look at the resources you linked @Jpjapers, they seem promising - thank you!

Id still say the biggest thing to always keep in mind is the reality aspect of whatever youre doing. If youre making a lamp light, do some googling and find out the lumens output of a lamp bulb. If you ever have to boost the sun intensity. Something else is wrong.  If you ever have a light set to crazy intensity to make it look realistic, something else is wrong. It sort of helps to just keep asking why when you reach those points. It often helps to set everything back to defaults to rebalance your image if you get to a point where youre making a completely unrealistic guess.

But as i say, your renders look great already! Everyone has a style. I certainly have a method when it comes to lighting but the longer you play around with things the more youre likely to develop your own style instead of trying to emulate others.  I didnt think you were asking for a magic button but it really does just boil down to understanding the software, its relation to real world values and knowing the physically accurate values and methods that you should be using to achieve realism. Yes CG is a playground and we arent bound by physics. But if youre going for realism, you should adhere to realism and the rules which apply in the real world.

As Da Vinci said "Simplicity is the ultimate sophistication".

2020-03-12, 00:49:21
Reply #5

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
Thanks man, I really appreciate that, and I definitely feel you with that DaVinci quote ❤️

It often helps to set everything back to defaults to rebalance your image if you get to a point where youre making a completely unrealistic guess.

Haha, yeah I definitely find myself doing this from time to time - get too far down the rabbit hole and just have to reset.

That is a good point about light intensity, and something I should probably do more to be mindful of. I've mostly been in the habit of working with W/(sr.m^2)(?)

Also, thank you for that false color LUT you shared, I really like it! I've used a couple other ones, but I prefer the color code with that one.

2020-03-12, 10:14:40
Reply #6

Jpjapers

  • Active Users
  • **
  • Posts: 1507
    • View Profile
I've mostly been in the habit of working with W/(sr.m^2)(?)

Ive been questioning which units are best to use for a while. I prefer using Lumens because lighting manufacturers pretty much always list lumens as part of their spec sheets.
In my opinion, lumens work alot better as the light output stays the same regardless of the radius, it just becomes more focussed the smaller it gets whereas W/(sr.m^2) is dependent on the size of the light.

2020-03-12, 19:28:39
Reply #7

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
I've mostly been in the habit of working with W/(sr.m^2)(?)

Ive been questioning which units are best to use for a while. I prefer using Lumens because lighting manufacturers pretty much always list lumens as part of their spec sheets.
In my opinion, lumens work alot better as the light output stays the same regardless of the radius, it just becomes more focussed the smaller it gets whereas W/(sr.m^2) is dependent on the size of the light.

What about directionality? I'm asking myself for a while about that... Light power does not seem to take that parameter into account

2020-03-12, 20:31:21
Reply #8

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@Jpjapers this is a great example of what I was originally posting about - I had no idea lumens wasn't affected by the light dimensions (?)

2020-03-12, 23:38:37
Reply #9

Jpjapers

  • Active Users
  • **
  • Posts: 1507
    • View Profile
@Jpjapers this is a great example of what I was originally posting about - I had no idea lumens wasn't affected by the light dimensions (?)

Well i had a few years of experience in commercial and retail lighting so i had to have a very basic understanding of real-world info on theoretical light measurements.
Ill try and give a brief and probably misguided rundown and im sure theres things here that are completely misunderstood or mixed up so ill gladly take corrections if ive got it wrong. I know Ondra and Juraj have talked about this in various threads.

First of all Each of the measurements have a specific function and arent just a personal preference thing.

Lumens is a measure of how much total light a luminaire will emit at the source (luminous power) so it isnt dependent on how big the light is because its an absolute.  So literally "How bright is this bulb?". Its required by law on most lighting spec sheets because there are minimum required lux levels across various spaces by law so this info should be available for every light you could need to include in your scene. I like working with this one because you can be sure that regardless of how big the light is, if you know the lumens value, you can be sure its accurate.

Lux The measure of light output as percieved by your eye ( measures exitance at the source and illuminance at the surface). Its a photometric measurement of how lumens are spread across a surface.

So if you run IR and set the intensity units to something like 2000 w/sr.m^2 and change the radius, youll see the scene get brighter. Whereas lumens, if you make the light bigger, the same amount of light output is spread across a larger area like when you focus and defocus a lens on a torch or a headlamp. See below. The torch doesnt get brighter its just more focussed resulting in higher lux but the same lumens.

You could have 1 really small bright light in your house and get a high lux level on your floor. Whereas if you wanted to light a warehouse with those same lights, you would need alot more of them to achieve the same lux level despite the lumens output of the light being the same. 1 lux is 1 lumen per square meter. I tend to use lux for things like mimicking daylight because its easy to measure lux with a light meter.




Watts per steridean per square meter is a radiometric measure of the lights percieved brightness and it depends on the size of your light so for every size increase, you'll get a power increase. It measures radiance which is sort of how much of the emitted power the eye is recieving. Id like to understand this a bit better.

Candelas (cd) measure luminous intensity. That is, how bright is the light at the source. 1 candela for a wax candle flame is accurate. It uses something called the steridean value which is basiccally the cone of the light. I'm not certain on this one. Again id like to know more.

Here's a little diagram of how the measurements relate to each other




What about directionality? I'm asking myself for a while about that... Light power does not seem to take that parameter into account

I wish there was a corona light meter like the old one in mental ray. Then i could actually do some real tests because theres some stuff about Steridean I dont understand (which i think is the technical term for directionality?) and i thought based on the descriptions of the units, the directionality would affect them differently . I asked a fair few years ago about the implementation and maru had some quotes from somewhere. Hopefully one day we might see some sort of implementation when the big list of future bugs/features is running dry. Unfortunately now mental ray is gone you cant really do much.

Quote
You can paste the value which you obtain in W/(sr.m^2) from image pixel (or from your lightmeter in real life) into Corona light intensity and then change light units to Lux. But from what I understand this still requires some guessing and is more of a "workaround" than real light metering. Unfortunately, it looks like a proper implementation of light metering tools will be required for this to work.

Another idea is to use native 3ds Max photometric lights as Corona supports them. You can then switch to Mental Ray and use pseudo-color exposure controls or other tools which are built into 3ds Max to get the desired light intensity, and switch back to Corona again.

Quote
it seems that the easiest way to recreate real light in Corona (or any other renderer probably) would be to find what is the light's intensity in lumen (it should be on the box/package if this is some regular light bulb, or there should be some manual with such info if this is some more advanced/photographic light). If you know value in lumen, then you should be able to simply paste that value into Corona light. If you know the power of light in W, and you know the light's efficiency, then you should also be able to convert this into lumen. Here is an example calculator: http://www.rapidtables.com/calc/light/watt-to-lumen-calculator.htm

There are also some obstacles like light color, or light shape, so it looks like this is always some simplification.

Quote
If you are not after scientific precision, I would suggest you taking a number of reference photos so that you would know the camera exposure values (shutter speed, iso, f number, etc), then recreating these exposure values in Corona and tweaking light source intensity so that the image is visually similar to the reference photos. Unfortunately I don't think there are currently any easy ways of doing this.
« Last Edit: 2020-03-15, 23:27:02 by Jpjapers »

2020-04-22, 10:22:33
Reply #10

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@Jpjapers at the very least your summary seems like a good shorthand for how to think about these different parameters. You mentioned earlier:

If you ever have a light set to crazy intensity to make it look realistic, something else is wrong.

I was going back and forth with LightMix and realized I was boosting a W/(sr.m^2) light source to >11,000k and then remembered your comment.

Your summary post seemed to indicate that W/(sr.m^2) measurement may include a perceptual component - so there can't really be a 'ground truth' measure for a light of that kind(?)

2020-04-22, 10:59:57
Reply #11

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
1. "Turn linear sRGB into digital camera product"

Quote
having it defaulting to a camera, with an option to switch to a very special mode, which will make your output less realistic, but compose-able in post.

You can get this by rendering your image at the default settings, then enabling LUT and picking one of the 3 Kim Amland's photographic LUTs. They are captured from real life digital cameras.
Then you can optionally reduce the strength of the LUT to ~0,8 (to your own taste) and play with highlight compression and filmic values (again to your own taste, to reduce highlight burn).


2. "I want to produce images like the masters"

Quote
This is a very attractive proposition and it speaks to the issue I am referring to. Regardless of whether such an implementation is feasible or not, I'd like to learn how to produce images that are of the quality we expect from the Corona Gallery and I just haven't seen strong tutorials - free or paid - demonstrating such an approach.

I am afraid this is not as easy as just following one or two tutorials. Artists have their own secrets. Some of them definitely use Photoshop or other post tools to adjust images in addition to what is done in the VFB, not only by changing the overall tone mapping/coloring, but also locally, e.g. by making some areas more or less noisy, brighter/darker, more clear/blurry, etc.




Hey Maru, what do you exactly mean by "rendering with default settings" ? Literally all cam settings untouched, like HC 1, contrast 1, etc ?
Absolutely no highlight compression mostly burns some whites, as we all know. I know that many guys fix this in post as 32 TIF bit or EXR.
In my workflow, for example, I rarely have time for this process... and honestly I don't even like the idea to fiddle around too much in post. Feels fake to me.

However, lately I notice that "correct tone mapping" gets more and more in Corona users´ minds... since, also in my opinion, it is THE key factor for realism.






2020-04-22, 11:11:08
Reply #12

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Its not just tonemapping that makes those guys renders look amazing every time. Its attention to detail at every stage and questioning how something actually looks in reality vs what we imagine it looks like. Lacking some minute details is enough to make your brain subconciously question an images' realism so the answer to the question "How do i make it look like [insert artist here]" is usually "Hard work and practice". Personally, I think photorealism boils down to basically one rule.

I agree only partially with that. Of course stuff in your scene should be realistically modeled , textured, etc.

BUT: if your render engine just does not get that sweet spot of light calculation and the correct proportions in color grading, etc... even the most qualitative scene will look fake.

Lately I watched Johannes Lindqvist again. And as he mentions... the basic part of the realism in his work is due to FStorm and its visibly great tone mapping.
Quote Lindqvist: "even a stupid can generate realistic images in FStorm without tweaking on buttons."

And of course, as you mentioned: on top, his modeling and textures are 99% insane, which makes his images look insane.
As for post, I saw that he often does just a minimal tweaking... well, that´s clear with renders that are naturally tone mapped & lit from the start.



I will give another example: look at real photos. No matter how shitty, blurry or badly exposed they are, you will 99% see that it is a natural pic, not a render.
Why? because the proportions in contrasts, colors, etc. are "correct / natural.
« Last Edit: 2020-04-22, 11:36:28 by Designerman77 »

2020-04-23, 04:09:31
Reply #13

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
This subject has surfaced over and over since Fstorm seems to produce great results out of the box.  I wanted to try.
I found a good Cornell box taken with a Nikon D70, I believe it was used to showcase Maxwell 1.0 back in the day.

So, this is how it looks when we compare it to Corona default settings:



I agree that it looks a little dull and uninspiring. Maybe even "not realistic" compared to what you get from a camera.

Now, this is what you get with Fstorm in the Default settings.


I used the exact same RGB values, maybe there is something I don't know about Fstorm, so I tweaked the RGB values to match a little bit closer.



That's better... so yea Conclusion: Fstorm produces more photographic images in the default settings.

However....

Look what happens when I enable LUT, choose Kim Amlan 02 and Enable Bloom and glare (it takes 1 second to do all of that)





Well, to me, that looks pretty photographic. I don't think it would be possible to correctly choose Fstorm or Corona in a blind test.

Now, if you like the original Fstorm look. Just increase Saturation to 0.2



Now, I'm not suggesting that just because it is easy it shouldn't be extremely easy.

My proposal is to include a "Magic" button (just like the iphone has) that automatically enables the most photographic settings. Maybe even enables a new pannel with limited options that all produce "ready to use" images.

I do not think you will be able to produce Johanes level images just like that. But I do remember how much my images improved when I discovered LUTs, and how long it took, because I didn't have that information in my brain, there was some friction.

I would also say make the photographic LUT the default, since everyone who doesn't want LUTs, know what they are and how to disable them, but people who WANT LUTs, might not know they exist.

But that's probably too much of a change.

Anyway, hope this helps!
 


2020-04-23, 10:59:21
Reply #14

Nejc Kilar

  • Corona Team
  • Active Users
  • ****
  • Posts: 1096
    • View Profile
    • My personal website
That is a very interesting observation and post Lolec. Thank you for your time :) I am a huge fan of those Dubcat's LUTs. They are amazing.

I suppose I'm wondering, could you share that Cornell box with us? I'd be really interesting in seeing how Octane handles it because they have a fairly specific "look" themselves.

Thanks again for your hard work, man!
« Last Edit: 2020-04-23, 12:44:38 by Nejc Kilar »
Nejc Kilar | chaos-corona.com
Educational Content Creator | contact us

2020-04-23, 14:48:50
Reply #15

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Regarding Lolek´s test: me personally I know those parameters in Corona and what effects they have, including LUTs. Using them all day long.
I guess 99,5% of us know those tricks.
However, a simple box or a living room with all furniture, etc... still seem to be two different things. :)
Who knows, maybe if you make all shaders super-duper physically correct, according to real life, it might get an extra kick of realism... ???
But who's time for such?

This morning I tried to look at good 3D images that stand out ( like those fron J. Lindqvist, Ferretti, etc. ) and have to say that a lot of the realism happens in the way they deal with the light
in therms of highlight peak, contrasts, etc. Those great images mostly have like a color layer that influences all colors... vs. too clean colors. And no crazy contrasts... since actually real world looks like this.
That's actually a thing one learns as an artistic painter at academies in the first semester. :)

2020-04-23, 17:31:45
Reply #16

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Right now I was a bit stuck on lighting a bathroom that has no windows and some creme colored wall tiles... not the nicest thing for the light mood.... I hate such rooms. :)

Long story short: by mistake I unticked "tone mapping" while the KimAmland LUT was full on ( value 1 )... and voilà, suddenly the freaking grayish light in the bathroom changed to a much more natural look,
just like I see now in Lolek´s posting.

Dudes... this setting which Maru mentioned (camera settings on default + Amland LUT)... seriously, this should be standard Corona setting and the first basic tutorial.
As other users mentioned: a pity to fiddle around for years in the cam settings, the light, materials, etc - just to realize that camera on more-less base setting plus a good LUT is the better way to go.

Glad that this topic of light mapping came up here!





2020-04-23, 17:46:46
Reply #17

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
It seems a lot of people believe there is a magic button that makes everything look photoreal but that won't happen. We're 90% there with Dubcat's LUTs. The only missing bit is a slightly better handling of highlights. That's it. Materials, composition and lighting play a huge part too. In fact, the main issue here is that with Dubcat's LUTs, you're stuck with embedded curves. So a good tone mapping tool would let access to toe, shoulder and medium parts of the image, in an elegant manner (probably in log space like ACES is doing during the display transform). But there is no universal settings... So I disagree to make those settings default.

2020-04-23, 18:28:31
Reply #18

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
I agree there is no "magic" setting. But Fstorm has a strong LUT as default and maybe that gives people the idea of it being so much better (it is not)

I think this is a separate discussion. After you've done every possible thing, Fstorm still handles tonemapping a little better. But I suspect what most people would notice is just enabling a photographic LUT.

That's why I think a single button with an attractive name which selects a photographic LUT enables bloom and maybe some other small tweaks would make a huge difference and unlock peoples brains to start exploring with those settings too.


2020-04-23, 19:53:54
Reply #19

agentdark45

  • Active Users
  • **
  • Posts: 581
    • View Profile
Thanks for taking the time to post up the Cornell box tests lolec.

While I agree that the Kim Amland LUT with minimal VFB fiddling produces a vastly super result than stock settings, the way Fstorm handles highlight peaks, the bottom curve of extreme darks and colours in general is still superior. It may not be as apparent in a simple Cornell box scene - start populating a real scene, with strong lighting certain areas and watch what happens to your highlights, strongly coloured materials and deep shadows; things start to break down and you end up chasing the dragon.

You can start to notice it in your Corona + Kim example 1; look at the exaggerated shadowing and black crush creeping in, as well as the front face of the small cube compared to the Fstorm tweaked example. Overall scene contrast is retained in the Fstorm tweaked example without exaggerated shadowing - it almost looks like light is behaving more realistically on the objects.
« Last Edit: 2020-04-23, 20:02:49 by agentdark45 »
Vray who?

2020-04-23, 20:48:35
Reply #20

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@lolec that is a really compelling example, although I am reluctant to generalize too much from such a well-defined scene. That being said, the comparison does drive home your point about photographic LUTs.

Thank you for sharing!

2020-04-23, 23:04:00
Reply #21

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Super important topic!
Maybe THE most important since long time. So thankful, all of you guys are participating with great input, ideas, opinions, etc.

I think many 3D artist would like to have a default setting that works like the human eye sees reality. Without lost details in highlight, nor with too blackish shadows, dull mid tones, etc.
And on top of this basic "eye-like" setting, more photo-typical behavior should be possible to - lets say - fade in with a slider. :))))

Of course, with enough grey cells, experience and try-and-error, one can always find a solution, also now.
But when clients demand great AND quick... well, then a render engine that simply does the magic out of the box would be the go to.

Today, for example, I lost hours with light setting... while a huge project has to be done as quickly as possible. And I know that my client will ask tomorrow for results.
So I'm sitting again now, at night and do work.
Costs nerves...


2020-04-24, 11:37:10
Reply #22

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Well, that's more related to how you market your work, to be honest. Making great images takes time. So great and quick is a bit wacky.

IMHO the topic is not to mimic the perceptual representation of a scene but a filmic representation of a scene that is probably more contrasty than what you would think due to dynamic range limitations. Terminology is really important.

Not to mention that pro photographers never use their images as is, they retouch them to look good, convey emotions, emphasizing some aspects of the scene, tweaking colors, bringing more light in some places etc etc.... So having a reference curve to lookdev is totally fine but that stops there for me. Hence the fact that a lot of us were pushing for ACES as it offers both a nice filmic preview AND a color management system that will allow you to deliver your work on a variety of media including HDR displays. That said, that's a fixed embedded curve on which you have no control (if you use default OCIO workflow) so it may disappoint a lot of people.

Not to mention that such filmic curves compress highlights so much that you will have to increase energy levels of your scene which increase realism but will also have an impact on render times (more overblown areas). Unreal engine implementation fakes it by applying a 1.45 gain before the post-process because they did not want artists to be bothered by the exposure loss (was a bad choice imho).

2020-04-24, 13:44:36
Reply #23

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website

Not to mention that such filmic curves compress highlights so much that you will have to increase energy levels of your scene which increase realism but will also have an impact on render times (more overblown areas). Unreal engine implementation fakes it by applying a 1.45 gain before the post-process because they did not want artists to be bothered by the exposure loss (was a bad choice imho).

This doesn't make sense. Exposure increase is identical to increasing light source intensity. One stop of Light is double the light (since one is exponential and other linear). That's not opinion, that's fact.
None of this affect render times negatively per se, that's the purpose of MSI to control.

These threads are just getting weirder and weirder.

(Great tests Lolec! btw).
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-04-24, 15:04:02
Reply #24

mantaskava

  • Active Users
  • **
  • Posts: 80
    • View Profile

Not to mention that such filmic curves compress highlights so much that you will have to increase energy levels of your scene which increase realism but will also have an impact on render times (more overblown areas). Unreal engine implementation fakes it by applying a 1.45 gain before the post-process because they did not want artists to be bothered by the exposure loss (was a bad choice imho).

This doesn't make sense. Exposure increase is identical to increasing light source intensity. One stop of Light is double the light (since one is exponential and other linear). That's not opinion, that's fact.
None of this affect render times negatively per se, that's the purpose of MSI to control.

These threads are just getting weirder and weirder.

(Great tests Lolec! btw).

So here's a situation/scene - a single room with a single light in it. Now there's two ways to brighten it up, either you brighten your light or increase your EV/Exposure. Do I understand correctly that there's totally no difference which method you choose? Even if you would brighten your light to some crazy values (like 20 000 of whatever), it would still be okay? By okay I mean materials would still respond to light as expected etc.

2020-04-24, 15:46:21
Reply #25

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile

This doesn't make sense. Exposure increase is identical to increasing light source intensity. One stop of Light is double the light (since one is exponential and other linear). That's not opinion, that's fact.


Did I said anything otherwise? What's the point here?

None of this affect render times negatively per se, that's the purpose of MSI to control.

Well, this may not apply to corona sampling method but definitely would with Vray bucket mode. Renderer has hard time sampling bright areas. Here are two renders with the exact same settings, just different display transform (so the linear render behind both images is exactly the same):

sRGB


ACES


If you wanna match ACES perceptual light intensities to the sRGB one, you'll have to increase the exposure. So the linear render will expose more overblown areas with which the renderer struggle to deal with. Of course we're not talking tremendous amounts but still may count in animations. That what I experienced in my tests.
But as I said earlier in this post, it may indeed not apply to corona as it is a different sampling method. Maybe I took a bit of a shortcut here, I admit. It still is not devoid of foundation. I'm pretty sure you did encounter that Juraj, as you were a Vray user. Remember the last bucket of death ?

source refering to the UE tonemapping gain : https://acescentral.com/t/creating-an-ocio-view-transform-to-match-default-unreal-4-aces-viewport-tonemapper/2437
                                                                  https://acescentral.com/t/acescg-for-animation-feature-and-further-questions/1545/21

"Keep in mind that Unreal Engine 4 ACES fitting applies a 1.45 gain, i.e., it is over-exposed compared to ACES reference implementation. This is an aesthetic choice done so that artists at Epic Games had an easier transition to ACES compared to a simple Gamma 2.2/sRGB OETF or a tonemapper keeping brightness in the ballpark."

So to conclude, ACES tonemapping allows you to bring more energy into your scene which makes quite a difference. But is it not the point of a tonemapper ?
« Last Edit: 2020-04-24, 17:07:56 by Fluss »

2020-04-25, 00:14:36
Reply #26

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
These threads are just getting weirder and weirder.

Hopefully you mean these *comments* are getting weirder and weirder - the thread is a pretty honest request for help on a diverse topic.

Those great images mostly have like a color layer that influences all colors... vs. too clean colors. And no crazy contrasts... since actually real world looks like this.

@Designerman77 can you elaborate on what is meant by "too clean colors"?

the way Fstorm handles highlight peaks, the bottom curve of extreme darks and colours in general is still superior.

F-storm always comes up as the comparison group on Corona threads when tonemapping is concerned, and what is described here^ is often the punchline of those conversations.

Is there really any disagreement about this, and if not, how is that reflected in the tonemapping rework that appears on the roadmap?
« Last Edit: 2020-04-25, 12:47:55 by cjwidd »

2020-04-25, 12:20:23
Reply #27

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Well if you guys consider those comments weird then I'm gonna stop posting from now. I always try to add valuable insights and I always try to document my posts with links and tests, as much as possible. And I spend time for it as english is not my mother language. I also often mention that all the stuff I write is the reflect of my personal opinion and interpretation of things. I'm fine with contradiction as long as there is some argumentation to go with it. So I'm gonna stay far from those technical threads from now, I'm tired of this TBH.

2020-04-25, 12:29:20
Reply #28

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@Fluss I want to be clear that my comment was not targeting you and I appreciate you taking the time to share your insights on the forum +1
« Last Edit: 2020-04-25, 12:45:32 by cjwidd »

2020-04-26, 00:29:02
Reply #29

Jpjapers

  • Active Users
  • **
  • Posts: 1507
    • View Profile
oh god this is going to turn into one of those threads isnt it haha

2020-04-26, 01:03:28
Reply #30

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
oh god this is going to turn into one of those threads isnt it haha

This discussion might one day catapult Corona into another level of realism.

2020-04-26, 01:10:35
Reply #31

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
Yes, and perhaps more to the point, why aren't there more tutorials from accomplished (professional) artists on this topic - is because so much is left up to taste?

I would pay for a tutorial / walkthrough from Juraj just to study his process. @Jpjapers pointed me to Vizguru which has some good stuff, but it's funny how realtime art is basically drowning in tutorials - for every little workflow or technique you can imagine - but archviz, not so much.

Maybe I'm just not looking in the right places? Obviously Youtube has some tutorials, but it's mostly a toss-up.

2020-04-26, 01:13:22
Reply #32

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
These threads are just getting weirder and weirder.

Hopefully you mean these *comments* are getting weirder and weirder - the thread is a pretty honest request for help on a diverse topic.

Those great images mostly have like a color layer that influences all colors... vs. too clean colors. And no crazy contrasts... since actually real world looks like this.

@Designerman77 can you elaborate on what is meant by "too clean colors"?

the way Fstorm handles highlight peaks, the bottom curve of extreme darks and colours in general is still superior.

F-storm always comes up as the comparison group on Corona threads when tonemapping is concerned, and what is described here^ is often the punchline of those conversations.

Is there really any disagreement about this, and if not, how is that reflected in the tonemapping rework that appears on the roadmap?




I mean "raw" looking colors... you know, that unnatural, artificial look. :)

2020-04-26, 02:32:01
Reply #33

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
I'm starting to believe that there are 2 simultaneous conversations happenings because there is a lot of overlap between the two, they look like a single conversation. I'm hoping this post would help see everything more clearly.

Last time I did some cornell box tests and some of you rightfully noted that it is not a sufficiently conclusive test as it is a simple setting with no dramatic light or materials. I decided to run a second test. My goal is NOT to say: Corona is as good or better than Fstorm, because I actually don't know.

My goal is to find out WHY people seem to think Fstorm is SO MUCH BETTER in terms of color management. And to find out if it is, maybe I should switch to Fstorm.

Disclaimer: I never used Fstorm before yesterday.

For that reason I set myself a few requirements:

1. The scene should be based in a Johannes L. Image, since that's what most people refer to when they say "Fstorm handles color better"
2. The least number of materials. Since I will have to replicate them in Fstorm and I'm a noob
3. Real World use. Something that a client would ask for.
4. Basic lighting

Ok, so I found this perfect scene by Johannes L. I set it up using just 4 materials. Corona Sun and Corona Sky. For Fstorm I used Fstorm Sun and Fstorm Sky. I left everything I could at default, only played with white balance to match colors on both engines.

Maybe it would be more "correct" to load the same HDR in both engines instead of using the stock SUN and SKY system. But I figured they are close enough, and an accurate representation of the available tools.


First, this is the image I'm trying to approximate:




This is what I get out of both engines by default:



Ok, sure Fstorm is in fact producing a more pleasing and "photographic" by default. But you can see it is nowhere NEAR what Johannes makes.

Next is to enable everything I usually enable in Corona. This is my default workflow and the only thing I needed to tweak to come closer to the original image was a greater contrast value (usually 3 or 4, this time 7) 



As I showed in my previous example, the image changes dramatically.

Here is a before and after:




Then I tried to match the Fstorm Default with as few settings as possible. To me, it looks like highlight compression plus contrast does a good job.




You can see how my workflow produces more "photographic" images than Fstorm default. I must say that I use these settings for 90% of my renders. And only tweak in specific cases. I usually do most of the project in my default settings and only tweak at the very end (Curves, contrast)




I wanted to see what would happen if I loaded Kim Amlan 02 LUT into Fstorm. I needed to tweak the "burn" value a little bit. I tried to move the least possible settings. I bet an experienced user would be able to get a better result, but this was very interesting to me.



Yes, of course, there are slight differences. Their SUN and SKY models are different, so maybe that's why colors are a little different. And the reflections don't match. I don't know how to properly use Fstorm materials so I did what I could. 

But after this experiment, I think it's clear that it would be very very difficult for anyone in a real-world scenario like this to say that Corona's tonemapping needs to be reworked from scratch because Fstorm produces incredibly more photographic images.

I simply don't believe that is true. I can tweak my contrast and color settings to make the images match a little better, but I wanted to see what a minimum workflow looks like, and if Fstorm is truly superior beyond a doubt.

I do believe that there is a very valid argument to be made about enabling LUT, highlight compression, and increasing contrast by default. Or make it a single button. Or in some way make it SUPER EASY to do, even by accident. My logic is that the pros would don't like it but know how and why to disable. And the rest of the world would enjoy much better images by default.

There is a different discussion, that only the top of the top of th professional users understand and really care about when you are pushing Corona to the limits you DO bump into the inherent limitations of the sRGB color space. But my guess is that 90 -95% of the time, when people complain about tonemapping, what ther are taliing about is what I'm showing in this experiment.

Another note. When a lot of people say something, even if you KNOW it is wrong, it is worth paying attention to. My guess is that because the experts know how to do what I'm showing intuitively and naturally, they dismiss conversations as nonesense. If there was only one or two people asking for "fstorm colors" then I would agree. But a LOT of people want that, maybe they don't know exactly what they want, but they want their images to look better.

So my hypothesis stands.

"fstorm colors" means 2 things.

1 is super complex, only a handful of people understand it and care about that. It requires a lot of work from the devs.

The other one is making certain settings, that are already there, more approachable, maybe the default.
« Last Edit: 2020-04-26, 02:41:09 by lolec »

2020-04-26, 05:46:34
Reply #34

agentdark45

  • Active Users
  • **
  • Posts: 581
    • View Profile
Regarding the "colour" issue, I'm talking specifically about saturated colours breaking down in unrealistic ways under certain lighting conditions. There was an example above that shows the difference with what happens to reds and blues in ACES vs SRGB colour spaces as the exposure/lighting intensity changes. I have noticed this first hand in my scenes (yes even with the fabled Kim LUT) that sometimes a material that is properly created, without wacky RGB values will just not work as it should in real life and will need heavy post editing - I have yet to come across the same issue in FStorm.

While I appreciate your latest tests lolec, you would need to redo this with exactly the same materials (not quickly converted ones) and an HDRI capture of the exact same exterior lighting to draw any meaningful conclusions. We would also need to see a histogram of the images to compare what is going on in more details. When I talk of Coronas tone mapping limitations, which cannot be fixed by simply cranking the contrast slider (which crushes the hell out of blacks) + applying a LUT; I'm speaking specifically about the well known limitations of Reinhard tonmapping (burnt, saturated highlights or murky greys when clamped). The interplay between highlights, mids and the falloff to extreme black levels without getting crushed blacks gets completely ruined by cranking Corona's contrast slider.

Lets take a look at your last example, I will bullet point the major differences:

1.  The highlights are clearly overly clamped on the Corona version - most noticeably on the exterior/backplate, look at that ugly Reinhard burn and grey murkyness surrounding it. Now compare to the Fstorm version, notice how the highlights have a much more photographic burn and a realistic falloff to the green mids on the plant (which has kept its vibrancy and contrast, vs being washed out in a shade of grey). In fact you can see Corona struggling to tame the highlights of the exterior and plant from the very beginning (look at that neon green/yellow, and it doesn't get any better until it goes grey...sigh.)

2. Wood tones have not desaturated or lightened realistically in the presence of light/exposure in the Corona version. Perfect example of what I'm talking about when it comes to colour issues (along with the neon green plant highlights). I've had wood go vibrant overly contrasted red on certain parts of a scene, when the same material is almost desaturated in other parts of the scene (which it should have all been fairly desaturated) - usually in the presence of strong directional spotlights with a realistic Kelvin value. A good test of this is warm spot lighting in wooden shelves, with the same material elsewhere in the scene. Give it a go in Fstorm and Corona and marvel at the difference (again in a full scene). This colour phenomenon also happens sometimes to materials with a strongly coloured reflection (gold, copper e.t.c - and no, the RGB values were not sky high).

I will sum this up: you can not simply stick a LUT on a photo taken with a phone camera from 2010, adjust the contrast and claim it's as good as starting with the photo taken with the DSLR. You might get 90% of the way there, maybe fool a few people who think "it's good enough" but after extensive testing of both engines myself in many different types of scenes (yes even with dubcats ACES emulation settings, Kims LUTS e.t.c) I can hands down tell you they are not at parity tonemapping wise.
« Last Edit: 2020-04-26, 06:33:18 by agentdark45 »
Vray who?

2020-04-26, 07:02:38
Reply #35

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@lolec Johannes has a Patreon page, Vizguru, where this scene can be downloaded for FStorm - is that where you obtained the scene file?

It sounded like you were recreating the scene from reference, but you could access the actual scene files from the Patreon page, if you are a member, of course.
« Last Edit: 2020-04-26, 07:15:05 by cjwidd »

2020-04-26, 07:33:19
Reply #36

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@agentdark45 just to be clear, the summary of your analysis is that Fstorm and Corona are not on parity in terms of tonemapping (?) and you are in disagreement with @lolec who is saying that, on balance, the tonemapping potential of both renderers, Corona and Fstorm, is basically comparable(?)


2020-04-26, 13:48:26
Reply #37

agentdark45

  • Active Users
  • **
  • Posts: 581
    • View Profile
@agentdark45 just to be clear, the summary of your analysis is that Fstorm and Corona are not on parity in terms of tonemapping (?) and you are in disagreement with @lolec who is saying that, on balance, the tonemapping potential of both renderers, Corona and Fstorm, is basically comparable(?)

Correct. One more example to show how badly Reinhard stacks up compared to a proper filmic/ACES implementation:



Again, this cannot be done just by sticking a LUT over an image (i.e the band-aid approach).
« Last Edit: 2020-04-26, 13:56:36 by agentdark45 »
Vray who?

2020-04-26, 14:20:37
Reply #38

Jpjapers

  • Active Users
  • **
  • Posts: 1507
    • View Profile
TLDR: Its all about the filmic response curve and coronas currently isnt how it should be for proper filmic tonemapping. Fstorm has a more DSLR-like tonemapping system and so produces slightly more photographic looking renders.

2020-04-26, 14:22:30
Reply #39

Jpjapers

  • Active Users
  • **
  • Posts: 1507
    • View Profile

This discussion might one day catapult Corona into another level of realism.

This issue has been discussed at length for the years on these forums in threads much much bigger than this as im sure agentdark remembers.

https://forum.corona-renderer.com/index.php?topic=14958.0

2020-04-26, 16:58:56
Reply #40

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
@Lolec, I guess all of us are very thankful for the work you have put into the tests!
Very cool, man!


@agentdark45: you, Sir have a brilliant eye and observation! The over clamping and grayish stuff behind that plant, as well as the tone of the wooden furniture, etc... exactly those are the things that also bug me in my renders. And I guess a lot of other users.
One could always say "aaah, that's only small differences". BUT our eye / brain instinctively notices exactly that stuff and realizes that it´s a fake pic.

As I superficially mentioned before: it´s all about the correct proportions of colors, contrasts, etc. that make images seem real.

Dudes, this discussion is by far the most awaited and defo my favorite one. I do believe that this is going to boost the quality of Corona in the next months, as the devs. brains surely start to work in the background. :)







2020-04-26, 17:02:29
Reply #41

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile

This discussion might one day catapult Corona into another level of realism.

This issue has been discussed at length for the years on these forums in threads much much bigger than this as im sure agentdark remembers.

https://forum.corona-renderer.com/index.php?topic=14958.0


I remember that this topic is not new.
Well, then it is high time to not only discuss but start programming. :)

2020-04-26, 20:05:44
Reply #42

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
No, my point is not that corona is as good at tonemapping. It is an objective truth that Fstorm uses a superior system. More realistic.

My point is that I have a feeling that when people, specially new users, complain about corona's tone mapping. What they are really talking about is that it is harder to get photographic images in corona, because the default settings are not meant to be photographic and the tweaks are not obvious.

That is why I think this topic comes up over and over and over. I believe there are two different subjects.

TRUE: Corona needs to move away from sRGB. It is the future, it is demonstrably better.

Also true: In order to notice the limits of sRGB you need to do some very extreme examples. In the day to day, most people would probably not hit that limit.

Maybe true: A lot of people, specially new users don't know the correct workflow and don't even get all of what Corona can offer. So they are comparing Corona default to Fstorm default. 

Fixing the first takes a lot of development  time apparently. but fixing the second might be easier.



2020-04-26, 23:36:08
Reply #43

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
No, my point is not that corona is as good at tonemapping. It is an objective truth that Fstorm uses a superior system. More realistic.

My point is that I have a feeling that when people, specially new users, complain about corona's tone mapping. What they are really talking about is that it is harder to get photographic images in corona, because the default settings are not meant to be photographic and the tweaks are not obvious.

That is why I think this topic comes up over and over and over. I believe there are two different subjects.

TRUE: Corona needs to move away from sRGB. It is the future, it is demonstrably better.

Also true: In order to notice the limits of sRGB you need to do some very extreme examples. In the day to day, most people would probably not hit that limit.

Maybe true: A lot of people, specially new users don't know the correct workflow and don't even get all of what Corona can offer. So they are comparing Corona default to Fstorm default. 

Fixing the first takes a lot of development  time apparently. but fixing the second might be easier.



Lolec, let´s be honest. Tweaking here, twisting there... all of us want colors & contrasts that look realistic as a starting point.
It is 2020... this should be the basic in any good render engine.




2020-04-27, 13:24:57
Reply #44

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
@agentdark45 just to be clear, the summary of your analysis is that Fstorm and Corona are not on parity in terms of tonemapping (?) and you are in disagreement with @lolec who is saying that, on balance, the tonemapping potential of both renderers, Corona and Fstorm, is basically comparable(?)

Correct. One more example to show how badly Reinhard stacks up compared to a proper filmic/ACES implementation:



Again, this cannot be done just by sticking a LUT over an image (i.e the band-aid approach).

The thing is, Reinhard is only compressing highlights. Filmic TM is compressing Highlights and applying a filmic curve to the render so that's a biased comparison. So basically, using dubcat LUT (the curve) and compressing the Highlights (Reinhard) already give pretty decent results. But, there is still that outdated sRGB EOTF in the process and that sucks. Other than that, the technical side of Corona is way more solid than Fstorm one from what I've seen.

But it's not all about Tonemapping, rendering space also have a significant impact. The wide gamut of ACEScg space offers quite a lot of extra colors that are beneficial to the render (colors are not compressed at the boundary of the gamut like in sRGB). Chris brejon made some interesting tests on that matter and you can clearly see it :

https://imgsli.com/MTQ2NzA

https://imgsli.com/MTQ2NzA/2/3

Corona is already using a wide gamut render space on which we do not know anything. How are handled the translations between sRGB and this space? Could totally be fine but i'd be curious to see how it compares to ACEScg one as we know the choice of the colorspace as a tremendous impact color saturation and can lead to bad results if not chosen carefully (otherwise everybody would be using XYZ as rendering space). All I know is that there is a reddish color cast on corona that is caused by that choice of colorspace(?).

edit : (?) mark on colorspace part
« Last Edit: 2020-04-28, 19:14:47 by Fluss »

2020-04-27, 14:08:47
Reply #45

romullus

  • Global Moderator
  • Active Users
  • ****
  • Posts: 7664
  • Let's move this topic, shall we?
    • View Profile
    • My Models
Fluss, can you elaborate about "reddish colour cast"? I know about slight warm tint when using white Corona light with neutral WB, but i think it has nothing to do with colour space.
I'm not Corona Team member. Everything i say, is my personal opinion only.
My Models | My Videos | My Pictures

2020-04-27, 14:40:30
Reply #46

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
While making comparisons some time ago, I've noticed some sort of magenta shift in corona compared to Vray. That's subtle tho but it's there. Will try to find those and post them.

What is strange is I did not notice any color shift using ACEScg in Vray, but I did not look for it specifically so that's something I have to check.

Maybe it's not related to colorspace tho but I'd be curious to know where does it come from if I'm mistaken. Any hint about that?

2020-04-27, 15:05:45
Reply #47

romullus

  • Global Moderator
  • Active Users
  • ****
  • Posts: 7664
  • Let's move this topic, shall we?
    • View Profile
    • My Models
I'm not Corona Team member. Everything i say, is my personal opinion only.
My Models | My Videos | My Pictures

2020-04-27, 23:17:35
Reply #48

Jpjapers

  • Active Users
  • **
  • Posts: 1507
    • View Profile
There were several discussions about that in the forum. You may want to read it in case you've missed it.

https://forum.corona-renderer.com/index.php?topic=19308.0
https://forum.corona-renderer.com/index.php?topic=20622.0
https://forum.corona-renderer.com/index.php?topic=27966.0

Im really surprised this is still a thing given its been reported for 3 years and is actually a fairly important thing that white is white.
Yeah 6500k could actually be closer to 6504 but if you set the same K in WB and light it should always be white no?



2020-04-28, 00:17:19
Reply #49

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@Jpjpapers You probably searched the links, but iirc the take home message is that there is some good internal reason why it is the way it is and it's best to set the light color directly if you mean to avoid the magenta tint(?)

2020-04-28, 10:13:28
Reply #50

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Thanks for the links romullus. Btw, what makes you think this can't be related to colorspace? As far as I understand it, blackbody temperature just refers to a given RGB triplet in the rendering space. The shift could totally be occurring in the translation from wide RGB to sRGB. It also looks like the wide RGB colorspace has a D50 illuminant. This might also have its bite in the color shift. And just to avoid confusion and to prevent people from crying out 'scandal', I'm not a colorist.

That said, that's not a critical issue to me and I can totally understand that this stay as is, that's just curiosity.
« Last Edit: 2020-04-28, 14:08:39 by Fluss »

2020-04-28, 10:20:39
Reply #51

romullus

  • Global Moderator
  • Active Users
  • ****
  • Posts: 7664
  • Let's move this topic, shall we?
    • View Profile
    • My Models
Btw, what makes you think this can't be related to colorspace?

Don't ask me :] All those talks about colour spaces etc. sounds like black magic to me. I'm happy to have enough brain power to somehow figure out linear workflow back in the days. I doubt i'll ever manage to handle all those fancy ace thingies you all talking about.
I'm not Corona Team member. Everything i say, is my personal opinion only.
My Models | My Videos | My Pictures

2020-04-28, 13:48:53
Reply #52

Jpjapers

  • Active Users
  • **
  • Posts: 1507
    • View Profile
Im not saying it should change im just confused as to why regardless of what tint it is, if you have your WB and lighting kelvin the same,  why they wouldnt always be white?

2020-04-28, 13:50:25
Reply #53

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Hey everyone,

as a byproduct of this discussion, I experimented a bit yesterday on the question "how is your raw render and how do you want it to look".

Conclusion: there ist not much escape from doing things in post at the moment.
However, simply saying "I want it to look good" will not do the trick. Instead, analyzing other (great) pics, an having a clear idea of what you want your pic to look like is a path one has to take.

Long story short, even after short and "minor" tweaks in PS, my images started to look more natural - with the side effect of losing that "dramatic" shadows, contrasts, color bleeding, etc.
But I have to admit that less artistic, dramatic & punchy looks "better", since it looks more natural.
The eye is not bombarded with crazy contrasts and colors - similarly to music recordings that do not violate your ears with artificially cranked up frequencies.

What I also noticed years ago: if you are not satisfied with color grading, etc., desaturate the pic to BW and tweak the adjusters ( highlights, contrast, blacks, etc. etc. ) until your eye is pleased.
When returning the saturation-slider back to default, the pic will - in most cases - look way better also in color.
I got this idea from the observation that well lit & color graded pictures somehow always look good, no matter how much you darken your screen, or desaturate them, how blurry a pic is, etc.
« Last Edit: 2020-04-28, 13:55:51 by Designerman77 »

2020-04-28, 16:03:33
Reply #54

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
These two comments, in my opinion, summarize the issue pretty well.
But -and I'm far from an expert in raytracing- can we discard that the realism in Fstorm vs Corona comes from sampling and raytracing?


1.  The highlights are clearly overly clamped on the Corona version - most noticeably on the exterior/backplate, look at that ugly Reinhard burn and grey murkyness surrounding it. Now compare to the Fstorm version, notice how the highlights have a much more photographic burn and a realistic falloff to the green mids on the plant (which has kept its vibrancy and contrast, vs being washed out in a shade of grey). In fact you can see Corona struggling to tame the highlights of the exterior and plant from the very beginning (look at that neon green/yellow, and it doesn't get any better until it goes grey...sigh.)

2. Wood tones have not desaturated or lightened realistically in the presence of light/exposure in the Corona version. Perfect example of what I'm talking about when it comes to colour issues (along with the neon green plant highlights). I've had wood go vibrant overly contrasted red on certain parts of a scene, when the same material is almost desaturated in other parts of the scene (which it should have all been fairly desaturated) - usually in the presence of strong directional spotlights with a realistic Kelvin value. A good test of this is warm spot lighting in wooden shelves, with the same material elsewhere in the scene. Give it a go in Fstorm and Corona and marvel at the difference (again in a full scene). This colour phenomenon also happens sometimes to materials with a strongly coloured reflection (gold, copper e.t.c - and no, the RGB values were not sky high).

I will sum this up: you can not simply stick a LUT on a photo taken with a phone camera from 2010, adjust the contrast and claim it's as good as starting with the photo taken with the DSLR. You might get 90% of the way there, maybe fool a few people who think "it's good enough" but after extensive testing of both engines myself in many different types of scenes (yes even with dubcats ACES emulation settings, Kims LUTS e.t.c) I can hands down tell you they are not at parity tonemapping wise.

My point is that I have a feeling that when people, specially new users, complain about corona's tone mapping. What they are really talking about is that it is harder to get photographic images in corona, because the default settings are not meant to be photographic and the tweaks are not obvious.

That is why I think this topic comes up over and over and over. I believe there are two different subjects.

TRUE: Corona needs to move away from sRGB. It is the future, it is demonstrably better.

2020-04-28, 19:05:36
Reply #55

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
I would pay for a tutorial / walkthrough from Juraj just to study his process.

:-) No need to pay me, I posted my approach almost every year, every time someone asked. It's true I didn't make a concise tutorial out of it, but that's because (and this answers why there in general aren't many tutorials on topic) it's not easy to concentrate and doing this properly. Some people open Twitch and ramble random stuff for 40 minutes and others, like me, struggle to make good script and follow that. I rather not produce anything that produce something that's mediocre. My CGArchitect article took me week to write..and I don't want to spend so much of my free time to CGI.

Disclaimer: I also absolutely want both: color management (OCIO/ACES is great) and better tonemapping (ACES Filmic curve is great). I've requested these years ago. I will be super happy once we eventually get them. And we will, I've never seen any devs more receptive that Corona guys.

But.. these things aren't as radical as some people might hope for. Out of fun I googled ACES Filmic curve and google shows on first image page an experiment I did with Dubcat in 2017. He took my scene to Fusion (pure linear 32bit output) and applied ACES Filmic curve (with final sRGB output, sRGB is color space only btw, it has nothing to do with tonemapping whatsover). The other image is Corona HC8 (I guess was chosen to maximize highlights compression) plus a simple S-Curve that mainly focuses on boosting midtones (the midtones look bit more shitty because the curve was applied to clamped image in Photoshop... Corona didn't have curves in VFB yet)
The results are fairly similar. Of course, the scene lacks strong colors where the saturation would fall apart, but even this can be easily fixed in interim by using Vibrance to adjust saturation levels.

This is currently not reproducible in Framebuffer with LUT because it would requite exact translation to LOG first and applied right after in that order { Linear --> LOG ----> LUT }

I said these threads are getting weirder because there is odd militancy in responses. Lot of strange stuff is taken as fact, it's becoming quite haven for misinformation.

But for you out of interest: My current settings are sub <2 HC, Midtones boosting curve (with peak more towards highlight to avoid flatness) without toe as I like to adjust black with Contrast, which is 4-5. Then..still comes a lot of selective post-production, but this identical to my photography from that point, creative changes, not fixes. So, luminance masks, vibrance adjustment, micro-contrast enhancement, painting glares on (I don't like the current bloom&glare much and I want to boost this artificially).

Last: Great posts from Lolec!



« Last Edit: 2020-04-28, 19:14:53 by Juraj Talcik »
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-04-28, 20:11:12
Reply #56

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
What I also noticed years ago: if you are not satisfied with color grading, etc., desaturate the pic to BW and tweak the adjusters ( highlights, contrast, blacks, etc. etc. ) until your eye is pleased.
When returning the saturation-slider back to default, the pic will - in most cases - look way better also in color.

That's an interesting trick, I'll have to try that sometime - thanks for sharing!

2020-04-28, 20:17:18
Reply #57

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
Some people open Twitch and ramble random stuff for 40 minutes[...]

I wouldn't be opposed to listening to you ramble for 40min *about your work* - open a project and just walkthrough what you did and why - that would be very informative for someone like me and presumably others. Of course, this is basically what Johannes Lindqvist does on his Patreon and it has been very well received. If you don't have time because of work or other reasons, well, I understand that too.

I've seen / read many of the write-ups that you've shared to the forums, including the ACES experiments posted above - they were very interesting!

2020-04-28, 22:30:31
Reply #58

Jpjapers

  • Active Users
  • **
  • Posts: 1507
    • View Profile
I agree with Juraj. Theres alot of misinformation and lots of language around colour spaces gets used interchangeably or incorrectly and it just clouds the waters and makes it difficult to find the right information.

I think it would be nice for a 'once source of the truth' approach eventually. Perhaps if a proper filmic approach turns up in corona in the future there might be a nice little video explaining it all with examples of why its (evidently) a better approach for the way people use renderers to clear up the cloud of rambly threads that this kind of discussion always manages to turn into.

« Last Edit: 2020-04-29, 00:50:09 by Jpjapers »

2020-04-28, 22:37:46
Reply #59

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
A Trump meme already 😀Boy this escalated quickly.
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-04-28, 22:59:10
Reply #60

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
Isn't there something futile about discussing color management at all with respect to Corona Renderer for 3ds Max? As I understand it, Max is not a color managed workspace, and if it's taken Autodesk like 7 years just to add a new spinner to the chamfer modifier, why should we even begin to expect that color management is going to appear(?)

2020-04-28, 23:22:37
Reply #61

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
Waiting for Autodesk might indeed be futile but renderer can supply lot of native features like Corona shows 🙂
As long as its the main rendering engine you use it can even be better integrated solution (selection in framebuffer, etc.).

Vray already went this path so it makes sense Corona will opt so too.
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-04-29, 00:20:44
Reply #62

Jpjapers

  • Active Users
  • **
  • Posts: 1507
    • View Profile
A Trump meme already 😀Boy this escalated quickly.

Points for guessing the kelvin of his face.

2020-04-29, 08:36:37
Reply #63

niljut

  • Active Users
  • **
  • Posts: 16
    • View Profile
Isn't there something futile about discussing color management at all with respect to Corona Renderer for 3ds Max? As I understand it, Max is not a color managed workspace, and if it's taken Autodesk like 7 years just to add a new spinner to the chamfer modifier, why should we even begin to expect that color management is going to appear(?)

https://makeanything.autodesk.com/3DSMAX/public-roadmap-25B1-201DB.html#rendering-future

It is definitely on their minds at least.

2020-04-29, 09:21:43
Reply #64

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
oh wow, thanks for sharing that^ I had no idea they were even thinking about it

2020-04-29, 09:23:53
Reply #65

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Isn't there something futile about discussing color management at all with respect to Corona Renderer for 3ds Max? As I understand it, Max is not a color managed workspace, and if it's taken Autodesk like 7 years just to add a new spinner to the chamfer modifier, why should we even begin to expect that color management is going to appear(?)

https://makeanything.autodesk.com/3DSMAX/public-roadmap-25B1-201DB.html#rendering-future

It is definitely on their minds at least.

Well it was added 3years ago so wouldn't capitalize on this 😁: https://forums.autodesk.com/t5/3ds-max-ideas/opencolorio-ocio-support/idi-p/7028915

2020-04-29, 09:55:06
Reply #66

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Also, for those who missed it, you really should read Chris brejon online book: https://chrisbrejon.com/cg-cinematography/

He is a lighting artist and he worked in some of the biggest studios. There is a whole chapter about ACES and a lot more useful tips. Check it out!

2020-04-29, 10:05:58
Reply #67

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
yeah this is a great [FREE] resource, I've had it bookmarked for some time now :/

2020-04-29, 10:38:25
Reply #68

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
did you check it from time to time? he is adding new stuff, recently in the ACES chapter

2020-04-29, 10:49:51
Reply #69

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
Holy shit that resource has become serious wiki-bible. Now it's even more daunting :- D

Seriously if I were to start with CGI today, and came across this I would immediately find other career.
I need to bite into it.

The first section is just mind-boggling. The whole CGI community spent 10 years deciding on question of "What can change the nature of gamma?" and this guy just outlines the whole matrix.
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-04-29, 10:57:22
Reply #70

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
Seriously if I were to start with CGI today, and came across this I would immediately find other career.

lmao

did you check it from time to time? he is adding new stuff, recently in the ACES chapter

Nah, I haven't checked it since it was first recommended to me some time ago - it was already very thorough then. I'll have to see what chapters have been updated.


2020-04-29, 12:08:48
Reply #71

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Holy shit that resource has become serious wiki-bible. Now it's even more daunting :- D

Seriously if I were to start with CGI today, and came across this I would immediately find other career.
I need to bite into it.

The first section is just mind-boggling. The whole CGI community spent 10 years deciding on question of "What can change the nature of gamma?" and this guy just outlines the whole matrix.

hehe, if I came across this at the beginning, I'd be a way better artist today!

2020-04-30, 08:18:04
Reply #72

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
When I first posted this thread I was a little concerned there would be some self-righteous indignation about how I was basically asking for a magic button to cure my bad images or something like that. Of course, that was not what I was asking, but I was - more or less - implying that a significant proportion of what constitutes a good image is tonemapping (or lighting).

I recently subscribed to Johannes Lindqvist's Patreon to get a behind the scenes look at his approach to archviz / image creation, and he shared an anecdote in response to a Corona forum thread that *WAS* asking for a magic trick (based on Betrand Benoit's Varenna project).

He said,

Quote
If you take a photo in really bad lighting, the image looks photo realisitic. It may look like shit, but still realistic. It doesn’t matter what lighting circumstances you take your photo in, it still look realistic, since it’s obviously a real photograph. And the same principle actually applies to 3D as well, even if you have bad, ugly lighting, that alone is not the reason to why your image doesn’t look realistic.

I guess I did not have the common sense or clarity of mind to arrive at this really obvious truth before reading it, but it really struck me.

2020-04-30, 10:14:36
Reply #73

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
While I totally agree with what Johaness said about real photography, I'm also a bit disappointed by the CGI part. That statement is true when everything is done right, not only tone mapping. There are a lot of phenomena hidden behind the hood. Caustics, light polarization, diffraction etc etc, to name a very few. The thing is while we are calling render engines unbiased, we're still faking it or mimicking it. Tonemapping really is important but with the wrong shaders, wrong light energy, not enough details, perfect surfaces, it won't look realistic at all. That's a balance between a whole load of stuff. A good tone mapping is crucial for achieving photorealism, so as the rest. So yes we need it but it won't revolutionize the whole industry like some people seems to say.

If you compare a sony and a fuji camera output, pictures do not look the same. Colors are different, tone mapping is different but it still looks realistic. Simply because you capture the right scene-referred data (Everything is behaving as it should, how we use to see it). So if you don't feed the rendrer with the right data, the tone mapping curve won't make it look good.

So photorealism really isn't that simple to define as there is a whole lot of intricacies. The key is to use references and observe the world around you. I often look at light interactions around me and a lot of these observations are making me think "If I'd see this in one of my renders, i'd probably think this is a bug".

It would be really interesting to make a thread with great images and review them to highlight why it works well.

Also, what's make Fstorm appealing on the first sight, to me at least, is that the tone mapping curve is not driven by the LUT so you can use the LUT slot for other purposes. If you look at Johaness workflow, he is using LUTs to stylize his renders (desaturate colors, shifting black etc etc..)
« Last Edit: 2020-04-30, 10:27:14 by Fluss »

2020-04-30, 10:39:45
Reply #74

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
Quote from: Fluss
That statement is true when everything is done right, not only tone mapping. There are a lot of phenomena hidden behind the hood. Caustics, light polarization, diffraction etc etc, to name a very few.

Hah I started writing response before you and then you write 90perc. of what I wanted ! Absolutely agree with the above.


I would argue that "ugly" lighting in CGI is easier to pull of photorealism, best example being flesh light. High contrast, small & directional, creating super strong highlights, placing focus on shapes instead of surfaces.
And vice versa, high-end studio photography sets look very CGI like, similarly like retouched people look almost plastic.

This is where I would caution against jumping to conclusion there is something this has to do with "tonemapping" necessarily. Your cell-phone is not doing any advanced tonemapping that wood look better than Reinhard. Every digital sensor captures data linearly, those linear data when extracted in that way look exactly like Corona with everything at 0 (and HC=1). Yet it looks photoreal because of the amount of detail.

The opposite is also true to some degree, that very good tonemapping & grading can make simple & lazy CGI set look photoreal...but only in low-res and only for a moment.

So both of those aspects are important, something Lolec and few others brought out. It can't be tied to tonemapping.

Also...the mythical "DSLR-like" tonemapping. It's not imho what people think it is. What your camera does, it only applies gamma curve onto linear data, then raw convertor applies S-Curve (either its own, or the one described in raw file, which is usually based on JPEG post-production by camera maker) and their own bit of tonemapping which you have available in the settings on front page of ACR for example. At that point, camera hasn't done anything, the Sony sensor (whether sold to you by Sony, Nikkon, Fuji, etc..) has only provided linear data. No tonemapping whatsoever. They do attach associated s-curve into the file (which the raw converter only interprets! the curve is only identical if you save directly to jpeg from camera).

So if we want "DSLR-like" tonemapping, there would have to be way to use any raw converter tools. But there's a catch, they don't expect the amount of information provided by CGI. Raw formats are barely 16bit at best, already with gamma curve.
If you force-open 32bit file in Adobe Camera Raw, it will do random equilizing, which can look almost bizzare depending on dynamic range of your scene. This is the only "magic" it does to photos as well... but the algorithm obviously gets very confused by the kind of data it gets with rendering.
You can do the old trick of zero-ing everything in older process setting, then you can use the default state of rendering and only apply tonemapping that AdobeCameraRaw offers. It only affects highlights.

Quote from: Fluss
So photorealism really isn't that simple to define as there is a whole lot of intricacies.

Yeah, very much this.

There is interesting research done by Paul Devebec & Co (or John Hable? I am not sure, I came across it on FilmicWorlds which is his blog) where ultra-realistic models (like the most high-end Hollywood VFX stuff we can do) still looked wrong to our eyes. Then they flipped them above to fool our brain finally. The evolutionary amount of sensory recognition is so well trained, that we perceive almost minutae differences between real humans and CGI humans. Even if we are talking super rendered, super lit, super post-produced photo-scans with perfectly shaped-in emotion.

I believe this effect manifests to much higher degree in architecture as well as we would like to admit. After all, how much time humans spend inside some kind of architecture? Same with nature. It always looks somewhat wrong. Opposed example would be cars, recent innovation, rather simple in shapes. I think almost nobody can tell difference between CGI or real car... even not exactly great renderings.

Every time people congratulate something for looking very photorealistic in forum, there is some crutch that strongly helps it. Super massive DOF, odd angle focusing on detail, etc. Lot of time these projects have 4-5 very "real" looking images...until the last one, with flat lighting, wide angle showing the whole room that client actually requested. And it doesn't look real at all. With no hack to help it, the same grading & tonemapping suddenly doesn't help at all.

Can most artists be sure when the issue can be down to tonemapping & grading, and when it's the detail? I would say this ratio is lot different than people think.

talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-04-30, 11:55:54
Reply #75

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
[...] best example being flesh flash light.

*edited for general audiences 😄

👀
« Last Edit: 2020-04-30, 12:42:04 by cjwidd »

2020-04-30, 12:03:14
Reply #76

romullus

  • Global Moderator
  • Active Users
  • ****
  • Posts: 7664
  • Let's move this topic, shall we?
    • View Profile
    • My Models
LOL
I'm not Corona Team member. Everything i say, is my personal opinion only.
My Models | My Videos | My Pictures

2020-04-30, 12:05:01
Reply #77

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@JurajTalcik Are you referring to this article about the uncanny valley?
« Last Edit: 2020-04-30, 12:10:35 by cjwidd »

2020-04-30, 12:37:53
Reply #78

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
False security of grammar checker 😀
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-04-30, 17:12:07
Reply #79

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Also, what's make Fstorm appealing on the first sight, to me at least, is that the tone mapping curve is not driven by the LUT so you can use the LUT slot for other purposes. If you look at Johaness workflow, he is using LUTs to stylize his renders (desaturate colors, shifting black etc etc..)

Just wanna add that I know curves are there but I find them uber-sensitive and it's impossible to fine-tune, I never use them.

2020-04-30, 17:40:52
Reply #80

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
Guys, I really think that the evidence is clear: Renderings done with Corona are more easily recognizable as CGI than those made with Fstorm.

Really, tonemapping is not IT. Even if the Corona team re-works the tonemapping area, it may get better, but it's not it.

I was just at the Facebook pages of Fstorm and Corona.

It took me a while to find images in the Fstorm group, even by what it seemed the most inexperienced memebers, that didn't look like a photo.
Then I went to the Corona Facebook page. The fist image I found was obviously a rendering done by a not-so-experienced user and there was no way to tell my brain that it may be a photo.

See images below. I tried to take the worst images in the Facebook pages. I think the evidence is obvious: It's easier to make photorealism with Fstorm than with Corona. And I cannot believe that the discussion is about improving tonemapping to get the photorealism that you get with Fstorm by default. (Very clear to me with the image of the laundry room, made with Fstorm. It looks like a toy, but there's a realistic quality to it, with seemingly no effort)

It's like when we were having this conversation in the Vray forum just when Corona was released back in 2013 or so. The issue is not tonemapping.
IMO the issue has to do with the level of bias-ness: Both Fstorm and Corona are biased render engines. But the way Corona does the bias by default is not as good as what Fstorm does.








Even in this image, obviously with a lot of detail, I can tell that boat is fake.



« Last Edit: 2020-04-30, 17:47:26 by lupaz »

2020-04-30, 17:46:14
Reply #81

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
I don't know if image compression ruin all of these but to be honest, the last render looks the best of all of those to my eye

2020-04-30, 17:48:28
Reply #82

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
I don't know if image compression ruin all of these but to be honest, the last render looks the best of all of those to my eye

Yes, I added that render there to show that even an excellent work with a lot of detail, in Corona, still doesn't look like a photo.

2020-04-30, 17:54:01
Reply #83

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
Even on this image: Very simple scene by what it seems to be a not very experienced artist:
Except for the window, my brain may believe this is a photo.


2020-04-30, 18:04:20
Reply #84

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
Since I never tested fstorm I cant tell what causes this but I noriced this on Bertrands scene he sell. The Fstorm has much stronger direct light. Why is it? Clamped GI? Is it accidental or intentional. It gives the space more readable light in way that I doubt is due to tonemapping.

Someone would have to do a deep dive becayse the features have odd parity in behavior. Fstorm HDRI has two modes (weird option to cut off direct light?), their portal light effectively becomes new area light (like one option in Vray).

Very hard to pick out what exactly is the cause with so many variables.

With Dubcat, we suspected the shader. I wonder if something lije OrenNayar wouldnt cause more highlights from durect light?
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-04-30, 19:04:54
Reply #85

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
There you are, the effect on this one is more what we are referring to when talking about Fstorm TM. I wouldn't even call this image realistic, there is a bunch of stuff that are off but there is something in it that feels natural indeed.

2020-04-30, 19:25:32
Reply #86

romullus

  • Global Moderator
  • Active Users
  • ****
  • Posts: 7664
  • Let's move this topic, shall we?
    • View Profile
    • My Models
Those comparisons are meaningless when viewer is even more biased than renderer...
I'm not Corona Team member. Everything i say, is my personal opinion only.
My Models | My Videos | My Pictures

2020-04-30, 19:39:01
Reply #87

agentdark45

  • Active Users
  • **
  • Posts: 581
    • View Profile
There you are, the effect on this one is more what we are referring to when talking about Fstorm TM. I wouldn't even call this image realistic, there is a bunch of stuff that are off but there is something in it that feels natural indeed.

The Fstorm TM tonemapping magic imo is mostly to do with the way it handles highlights/burnouts, the falloff to extreme blacks (and colour behaviour in general) and the ability to produce a nicely contrasted image without sacrificing the previously mentioned elements. The Corona VFB crushes the hell out highlights/blacks when trying to achieve a contrasty photographic image, either via a combo of stock HC/contrast/LUTs (see the previous comparison examples by lolec and my detailed observation notes). The way Fstorm clamps highlights and the gentle roll off to true black is beautiful, it isn't in Corona, period.

Again, I cannot stress enough that the handling of highlights and blacks cannot simply be fixed by arbitrary curves/LUTs inside of Corona's VFB - sure we might be able to get close (read: 90% as close) by rendering out a linear 32bit image and fooling around in photoshop - but goddamn, what a painful workaround.

I'm also starting to think part of the Fstorm magic is the internal colourspace, take a look at the examples on Chris Brejon's ACES page (at the very bottom and click left and right): https://chrisbrejon.com/cg-cinematography/chapter-1-5-academy-color-encoding-system-aces/ - notice that even GI gets affected along with colour bounce, most apparent in the red ball.

Again I should stress that I am nitpicking here, Corona has decent tonemapping options but when compared to FStorm it simply isn't on the same level after using both engines extensively. Sure, I can produce nice images in Corona (and I do for all interior work), but if I wasn't as limited with Fstorms materials + VRAM I would use Fstorm everytime. For high end product shots I use Fstorm exclusively - I have tried to match a lot of my scenes 1:1 in Corona and something is always off in comparison (overly clamped highlights/highlights too burned/colours being weird unless manual silly tweaking/blacks being crushed when trying to achieve more contrast...the list goes on).

« Last Edit: 2020-04-30, 19:45:10 by agentdark45 »
Vray who?

2020-04-30, 19:49:14
Reply #88

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
Those comparisons are meaningless when viewer is even more biased than renderer...

Not sure what that means.


For further comparison some images below.

Could DOF have something to do with it?
The hotspots in Fstorm are very good. ACES?

This is the what Vlado says about biased and unbiased rendering https://www.chaosgroup.com/blog/the-truth-about-unbiased-rendering
I believe that how Corona and Fstorm choose to be unbiased is what ultimately makes the difference in realism.

CORONA:







FSTORM:















2020-04-30, 21:06:02
Reply #89

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
I really did not create this thread with the intention of comparing FStorm and Corona Renderer...

2020-05-01, 17:08:01
Reply #90

agentdark45

  • Active Users
  • **
  • Posts: 581
    • View Profile
I really did not create this thread with the intention of comparing FStorm and Corona Renderer...

I think people are making the comparison because Fstorm has the best tonemapping algorithm out of all common rendering engines at the moment.
Vray who?

2020-05-01, 17:57:32
Reply #91

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
I really did not create this thread with the intention of comparing FStorm and Corona Renderer...

I think people are making the comparison because Fstorm has the best tonemapping algorithm out of all common rendering engines at the moment.

The thing is, as Juraj pointed out, we're not even sure the "Fstorm look" can be attributed to tone mapping only. So further investigations need to be conducted before making such a claim.

2020-05-01, 21:02:09
Reply #92

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
Well, regardless of the comparisons, there are a lot of interesting comments about *tonemapping* that are emerging from those comparisons, so that's good 👍

2020-05-01, 22:11:29
Reply #93

agentdark45

  • Active Users
  • **
  • Posts: 581
    • View Profile
I really did not create this thread with the intention of comparing FStorm and Corona Renderer...

I think people are making the comparison because Fstorm has the best tonemapping algorithm out of all common rendering engines at the moment.

The thing is, as Juraj pointed out, we're not even sure the "Fstorm look" can be attributed to tone mapping only. So further investigations need to be conducted before making such a claim.

Maybe not attributed solely to tonemapping, but I would hazard a guess it's at least 50% of the reason. Please look at the highlights and roll-off to extreme black in the examples posted earlier by lolec - that is one aspect that is definitely down to tone mapping. Colour behaviour differences could be due to the internal colour space/custom matrix, I agree.

I would urge anyone who has doubts about the highlight/black crush issue to test it for themselves in 1:1 comparisons as Fstorm is fully functional (with watermarks). You will not get the same level of pop/photographic contrast in Corona (without external post processing), without sacrificing either blacks or highlights - I would almost liken this to differing dynamic ranges of cameras as an IRL analogy.
Vray who?

2020-05-02, 00:47:05
Reply #94

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
There is no way 50% of the realism coming from the Fstorm community is because of the tone mapping. Install Fstorm and try it. See if your images look 50% more realistic just like that. They won't.

My guess based on the actual tests I did and trying to figure out the Fstorm phenomenon: 

1. It is easier to get a baseline good result because of the default settings, easier to take that to photoshop than what corona gives as default. Less settings to fiddle with, less to learn in order to get a good result (Great and excellent results are just as hard IMHO)
Other than the LUT default examples I provided earlier, here is another one for bokeh:


You can argue it is stupidly simple to just move center bias to 0.5 instead of 0. But it is SIMPLER to do nothing.

2. Thanks to gaming, people already have powerful GPUs installed, so while a 7k USD PC might provide little speed distinction between a threadripper build vs multiple 2080Ti build, new artists are more likely to already own the right hardware to take advantage from Fstorm, thus learning it faster. In the early days of learning, when the curve is steeper, 50% faster results means you are learning and iterating twice as fast. My theory is that you can become GOOD much faster with Fstrom with the hardware you already own. 

3. Totally subjective based on MY experience and appreciation, but settings seem to be "dumbed down" example: White balance is based on rgb, so you are forced to tweak by feel instead of by Kelvin value. This promotes visceral interaction instead of technical interaction, which I believe attracts artists wich tend to work by feel more than by scientific technique.



PS. I want to make absolutely clear that I DO believe tonemapping should be reworked. I Know reworking tonemapping would help get better images, that is just the truth, my vote for most wanted feature is set to that and I'm sad we don't have it yet. Bokeh is actually one of the areas you would notice the change, as overblown highlights would look white in Fstorm and Solid Color in Corona, as it was discussed in my thread in 2018:
https://forum.corona-renderer.com/index.php?topic=22219.msg136777#msg136777

But the fact that so many people think that so much of Fstorm's realism is because of tonemapping, does not make sense. The fact that you estimate that Fstorm is more realistic 50% thanks to tonemapping is evidence of that.

We''ve had Ultra realistic renders for decades, even without all of the technologies we have today on materials, ray tracing and tonemapping. New tools just make it easier.

TLDR: I would like Corona to rework tonemapping to achieve what it is now IMPOSSIBLE to do in corona, but there is a lot they can do to make what is already possible much accessible. I think that would clear much of the misconception that Fstorm is so much better at "realism"

2020-05-02, 01:50:00
Reply #95

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
I'm not sure it's a meaningful or empirical question to ask what percentage of an image is attributable to tonemapping - or shaders, lighting, modeling, etc.
« Last Edit: 2020-05-02, 03:45:06 by cjwidd »

2020-05-02, 03:42:37
Reply #96

agentdark45

  • Active Users
  • **
  • Posts: 581
    • View Profile
There is no way 50% of the realism coming from the Fstorm community is because of the tone mapping. Install Fstorm and try it. See if your images look 50% more realistic just like that. They won't.

But the fact that so many people think that so much of Fstorm's realism is because of tonemapping, does not make sense. The fact that you estimate that Fstorm is more realistic 50% thanks to tonemapping is evidence of that.

You yourself claimed you are not an experienced Fstorm user - I am a very experienced dual user of both Corona and Fstorm (along with other engines). As mentioned in my last post I exclusively use Fstorm for high end product vis. Why? Because due to testing both engines extensively on many scenes back and forth, some even with very basic materials, the result is always poorer in Corona (and no, this isn't down in part to DOF settings, material conversion or a lack of knowledge of non "out of the box" Corona settings). It is down to tonemapping and how highlights and blacks are handled by each engine. I have tried for days in some circumstances to match the image in Corona and the same ultra-photographic result simply was not possible within Corona's VFB no matter any combo of curve adjustments, contrast tweaking, HC tweaking, Kim Luts, various other LUTS, Dubcats settings e.t.c. I'm simply not willing to go "yeah that looks pretty close", which might suffice for some.

Again, look at your own example (attached). Highlights have been visibly ruined and the wood tones are off and not reacting correctly as they should IRL - an observation I went into in detail previously. The average viewer may not spot this but in detailed viewing it is clearly apparent. If you don't think this affects overall photorealism, or certain edge cases in visuals in any meaningful way (would you concede to some degree?) then I don't know what to say to you.

Why am I bringing up Fstorm? Because out of all of the render engines I've used it simply produces the most true-to-life images possible right now (and yes, it requires settings tweaking just like every other engine, so it's not a case of n00bish "hurr muh default settings" user error). The tonal balance of the image one is able to get out of Fstorm is always superior, and I'm no Fstorm fanboy either - I wouldn't use it if I didn't have to, I would love to use Corona for everything due to it's superior feature set.

If not for tonemapping, what else do you think contributes to Fstorms superior photorealism? I've already mentioned several times on the forums better bitmap filtering/sharpness, bump mapping implementation, colour handling, pleasing noise patterns, and better DOF speed/sampling - however most of these can be worked around within Corona, Tonemapping however cannot be as it stands. Hence my prior statement of 50% (which was a fast and loose approximation, not verbatim gospel) - put a highly experienced user of both engines to work on the same scene in both engines, work around the quirks of both Corona and Fstorm, and the Fstorm image will be more photographic and pleasing to look at, fact. Case in point: look at anything Bertrand Benoit has produced (who I think most would regard as one of the best in the game that uses many different rendering engines), and compare to his Fstorm renders, and this was with his first trial of the engine: https://bertrand-benoit.com/blog/westkaai-via-fstorm/ there is always something "off" about his photo realistic Corona and Vray renders (a minor uncanny valley effect, i.e lacking in brain fooling true photorealism).

Side note: Fstorm as a rendering engine in general is terrible in some real world use cases; interiors with a lot of bounced light where it's annoying portal system will not save you. And not to mention VRAM limitations. It's lack of material and plugin support is also shocking (still doesn't have proper composite maps and multi-layered materials) - another reason I don't use it for anything more complex than product vis.
« Last Edit: 2020-05-02, 04:30:53 by agentdark45 »
Vray who?

2020-05-02, 04:51:50
Reply #97

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
Can you try to state what you think my point is? I think I've repeated myself and it's not working.

Do you think my point is:  Corona's tonemapping is fine, lets leave it like that ?  Because I've said the exact opposite.

Not sure what you are arguing with, so please state my point so I can continue the conversation with that as a starting point.

2020-05-02, 06:11:36
Reply #98

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
Regarding that image of my earlier test, as others noted, I would have to use the exact same HDRI and closest as possible materials, not a simple conversion.

So I repeated the test. Same HDRI, Same LUT (KimAmlan02), Matched white balance. very simple materials.

Do you really see a 50% bump in realism in the Fstorm one? Maybe I don't have your eyes, it is entirely possible that I actually don't have the perceptual capacity.



Anyway, now that I've conducted the test in a more proper manner, my point stands. I will write it in bold so you don't miss it and you stop arguing with something I'm not saying:

Fstorm has better tonemapping.   I wish Corona had a DSLR like tonemapping

However, I don't think that is the main reason for the Fstorm community producing seemingly more realistic images more consistently than the corona render community.
I think there is a small number of users (Agentdark45 included) who have superior understanding and perceptual abilities and actually notice the super slight differences between the two renderers, which are real and measurable.

But I believe what MOST people notice is a vibrant community that attracts great artists who are learning faster and feeling comfortable sooner with Fstorm, and thus, becoming good enough that, overall, the Fstorm community is producing better renders, more artistic, more realistic.

Just to be super clear, FSTORM HAS BETTER TONEMAPPING, 100% agree. And it has a real impact in realism! I wish Corona implements this soon
ps. I also wish corona becomes a lot more beginner and artist friendly by making it MUCH EASIER to produce the best possible images it already can, which, BTW would inevitably be WORST than Fstorm regarding tonemapping. At least until corona implements a better tonemapping, which I will continue to push and wish for.

 

2020-05-02, 09:22:15
Reply #99

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
I know Dubcat posted some F-storm LUTs a while back, what are you feelings about using those in Corona? I know it's already been said that no amount of LUTs can make up the difference, just curioous.

2020-05-02, 16:11:54
Reply #100

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
I think the whole discussion would be over if we had the option of taking a raw rendering from corona ,absolutely linearly, like a raw photo from a camera and do tests outside of Corona. It's my understanding that this is not possible currently. Is it?




EDIT: This is hilarious:
https://www.dpreview.com/forums/thread/4119248
Quote
Nothing scientific, but ince the early 2000's I've always felt the look of Canon RAW images had a more plastic-like look (that's not negative - just different) than Nikon images, which seems more realistic.
« Last Edit: 2020-05-02, 16:22:40 by lupaz »

2020-05-02, 16:27:46
Reply #101

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
I think the whole discussion would be over if we had the option of taking a raw rendering from corona ,absolutely linearly, like a raw photo from a camera and do tests outside of Corona. It's my understanding that this is not possible currently. Is it?

Did you mean F-Storm?

From Corona linear is easy, it's everything default (off & HC=1). Not sure how straight-forward it is from F-Storm.

That quote from DPReview.. photography "experts" are almost like homeopathy believers. "More depth", "More 3D", "Color science", etc.. are the most common nonsense they throw around constantly and discuss endlessly.

talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-05-02, 16:31:45
Reply #102

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
I think the whole discussion would be over if we had the option of taking a raw rendering from corona ,absolutely linearly, like a raw photo from a camera and do tests outside of Corona. It's my understanding that this is not possible currently. Is it?

Did you mean F-Storm?

From Corona linear is easy, it's everything default (off & HC=1). Not sure how straight-forward it is from F-Storm.

No. I meant Corona.
So pixels collected to form the rendering are not filtered at all?
If this is the case, renderings taken from Corona to Photoshop would look just as realistic as fstorm. That doesn't happen.

2020-05-02, 16:48:14
Reply #103

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
I think the whole discussion would be over if we had the option of taking a raw rendering from corona ,absolutely linearly, like a raw photo from a camera and do tests outside of Corona. It's my understanding that this is not possible currently. Is it?

Did you mean F-Storm?

From Corona linear is easy, it's everything default (off & HC=1). Not sure how straight-forward it is from F-Storm.

No. I meant Corona.
So pixels collected to form the rendering are not filtered at all?
If this is the case, renderings taken from Corona to Photoshop would look just as realistic as fstorm. That doesn't happen.

Can you elaborate? I don't really follow.
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-05-02, 17:26:56
Reply #104

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
Sure.
Some are saying that what Corona needs is better tone mapping in order to get the same realism that you get from Fstorm.
Others are saying that the reason why Fstorm renderings feel more realistic is not tone mapping, but something else (to be determined).


Wouldn't it be a matter of taking a raw rendering from Corona and do the tone mapping in Photoshop?

This is what many of us do, and I feel that, even doing the tone mapping in photoshop, renderings from Corona don't look absolutely real, regardless of the scene, materials, and lighting. As someone said before in this thread, a photo looks real no matter the materials or lighting (unless you purposely try to make it appear not real).

So, again as someone said before, we are talking about 2 different things.

1- Tone mapping
2- realism.

To me to talk about tone mapping as the tool to make a rendering realistic doesn't make sense. And I think the confusion came from the beginning of this thread, when the OP wrote:

Quote
I'm either approaching lighting or tone mapping totally incorrectly; it is absolute guesswork every time, and rarely is it 'photorealistic'.

May be he didn't mean to use the word 'photrealistic'?

I do prefer to talk about making corona as realistic as possible, regardless tone mapping.

2020-05-02, 17:38:23
Reply #105

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
I see now, thank you. Yes that would point against tonemapping.

talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-05-02, 18:08:30
Reply #106

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Right now I'm totally struggling with an empty bathroom scene my client wants rendered.
Mo matter what I do, it all the time looks unrealistic due to lighting, colors, etc.
When I turn up the contrasts, the shadow tones start to get too colorful, etc. Sure, one can "fix" it in post...

But as I mentioned before: the "proportions" of light/shadow, colors, etc. simply do look not realistic.
And from such a "soup" it is hard to do miracles in Photoshop.


Annoying... one loses hours and hours instead of simply getting jobs done.
« Last Edit: 2020-05-02, 18:11:56 by Designerman77 »

2020-05-02, 21:31:24
Reply #107

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Here is another misconception: Most of the time, if your render looks too saturated after tone mapping, it's because your diffuse (or any other colormap it is) was set too saturated on the first sight.
« Last Edit: 2020-05-02, 21:38:46 by Fluss »

2020-05-02, 22:19:03
Reply #108

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Here is another misconception: Most of the time, if your render looks too saturated after tone mapping, it's because your diffuse (or any other colormap it is) was set too saturated on the first sight.


Thanks for your reply, Fluss.
Neither the original JPG (texture) was too saturated, nor did I not additionally desaturate it in C4D. I do take care for this detail and almost always desaturate bitmaps.

Let´s be honest... one cannot talk away the fact that Corona is not so easy in the topic of tone mapping, etc.
I see this in hundreds and hundreds of renders, even from super good artists like Angelo Ferretti.
I bought some scenes from him to see how others work. Same thing with colors in many of his scenes, too.






« Last Edit: 2020-05-02, 22:22:59 by Designerman77 »

2020-05-03, 00:18:13
Reply #109

twoheads

  • Active Users
  • **
  • Posts: 261
    • View Profile
If I may add my two cents, this discussion can be boiled down to braking bad's quote: "after all, how pure can pure be?"
For most people 96% purity is more than enough and for some people difference between 96 and 99 is not just noticeable it is like tremendous gulf.


2020-05-03, 00:52:31
Reply #110

bluebox

  • Active Users
  • **
  • Posts: 268
    • View Profile
Since I never tested fstorm I cant tell what causes this but I noriced this on Bertrands scene he sell. The Fstorm has much stronger direct light. Why is it? Clamped GI? Is it accidental or intentional. It gives the space more readable light in way that I doubt is due to tonemapping.

Someone would have to do a deep dive becayse the features have odd parity in behavior. Fstorm HDRI has two modes (weird option to cut off direct light?), their portal light effectively becomes new area light (like one option in Vray).

Very hard to pick out what exactly is the cause with so many variables.

With Dubcat, we suspected the shader. I wonder if something lije OrenNayar wouldnt cause more highlights from durect light?

I was trying to pinpoint what is the difference between the two engines and came to the same conclusion as you have Juraj that in Fstorm light seems to be more directional.

Flash-light renders looked totaly photorealistic back in the vray 3 days even with crappy shaders. This thogether with superior tonemapping is imho the key to ultimate realism.

Can't say tho whether the directionality comes from the shader as you guys with Dubcat suspected or form the GI algorithm.

Asked about Oren-nayar shader model in the daily builds but the devs haven't picked up the conversation. Take a look:



It instantly looks 10 times more realistic. The devs tho are busy with according to trello "key new features" like faster colour picker....

2020-05-03, 07:56:58
Reply #111

noldo

  • Users
  • *
  • Posts: 2
    • View Profile
Since I never tested fstorm I cant tell what causes this but I noriced this on Bertrands scene he sell. The Fstorm has much stronger direct light. Why is it? Clamped GI? Is it accidental or intentional. It gives the space more readable light in way that I doubt is due to tonemapping.

Someone would have to do a deep dive becayse the features have odd parity in behavior. Fstorm HDRI has two modes (weird option to cut off direct light?), their portal light effectively becomes new area light (like one option in Vray).

Very hard to pick out what exactly is the cause with so many variables.

With Dubcat, we suspected the shader. I wonder if something lije OrenNayar wouldnt cause more highlights from durect light?

I was trying to pinpoint what is the difference between the two engines and came to the same conclusion as you have Juraj that in Fstorm light seems to be more directional.

Flash-light renders looked totaly photorealistic back in the vray 3 days even with crappy shaders. This thogether with superior tonemapping is imho the key to ultimate realism.

Can't say tho whether the directionality comes from the shader as you guys with Dubcat suspected or form the GI algorithm.

Asked about Oren-nayar shader model in the daily builds but the devs haven't picked up the conversation. Take a look:



It instantly looks 10 times more realistic. The devs tho are busy with according to trello "key new features" like faster colour picker....

Absolutely, but fstorm also manages the diffuse roughness better, with much softer brightness variations as the roughness increases

2020-05-03, 08:49:45
Reply #112

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
I really don't think Oren-Nayar is going to span the gulf between expectations about realism and what is currently achievable with Corona or any other renderer for that matter; you won't even notice Oren-Nayar unless the surface is very rough.

2020-05-03, 10:53:01
Reply #113

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
If I may add my two cents, this discussion can be boiled down to braking bad's quote: "after all, how pure can pure be?"
For most people 96% purity is more than enough and for some people difference between 96 and 99 is not just noticeable it is like tremendous gulf.



In my yesterdays "trivial bathroom"-scene, I again could see that it is not about "96% or 99%". :)
It was simply impossible to get a decent lighting and material look with the on-board tools in Corona.
Tweaking in PS Cam Raw was necessary, in order to get something out off the dull, raw render.

One can of course argue that it is the same with real RAW pics out of pro cameras.
BUT: sorry, 3D render engine users who need to meet client deadlines would have an easier life without massive post production.
In most cases, it is okay what comes directly out of Corona engine.
However, in this case of shitty, small room with sand colored tiles, etc. Corona clearly seems to meet its limits.
And yes, I also tried artificial lights inside the room... looked even more unnatural, despite trying different tones - from white to whatever, IES and non IES lights, etc.








 

2020-05-03, 14:35:59
Reply #114

ynotsop

  • Active Users
  • **
  • Posts: 39
    • View Profile
I love Corona but here is a nice example where Fstorm really shines:

Photo by Eric Floberg captured with Canon EOS-1D X MARK II (I think):
http://tiny.cc/l93aoz

Kitchen interior from Johannes:
https://viz.guru/wp-content/uploads/2019/11/Test8.jpg
« Last Edit: 2020-05-03, 14:43:50 by ynotsop »

2020-05-03, 15:41:26
Reply #115

twoheads

  • Active Users
  • **
  • Posts: 261
    • View Profile
I love Corona but here is a nice example where Fstorm really shines:

Photo by Eric Floberg captured with Canon EOS-1D X MARK II (I think):
http://tiny.cc/l93aoz

Kitchen interior from Johannes:
https://viz.guru/wp-content/uploads/2019/11/Test8.jpg
In my opinion this comparison proves absolutely nothing. Never tried fstorm so I can't tell for sure if its tonemapping is superior to corona's but yeah it looks like it is.

Have you seen fstorm's gallery on their website? (this one is for all of us) For my taste it shows variety of good and average shots both architectural and product design, exactly the same as corona's gallery and so what?
My point is that corona is a tool which in right hands can produce as good results as fstorm, maybe not as fast and easy but it definitely can. Poor models, low texture quality and so on can make the best renderer in the world just average. I don't think the way fstorm handles tonemapping is the answer here. We take Johannes's work as an example but he's one of the best in the business and even his work really stands off.

Lastly, "that horrible reinhard's burn in the background is unacceptable" type of clients are the worst, they really kill creativity :)

side note: it's just my personal opinion, no offence.





2020-05-03, 15:51:15
Reply #116

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
My point is that corona is a tool which in right hands can produce as good results as fstorm, maybe not as fast and easy but it definitely can.

Not saying this isn't true, but can you post a couple of images from any artist you want, made with corona, that tricks your brain to believe it's a photo?

edit: or maybe we should do a "photo or corona" game, to be fair to corona.

2020-05-03, 17:20:41
Reply #117

ynotsop

  • Active Users
  • **
  • Posts: 39
    • View Profile
We take Johannes's work as an example but he's one of the best in the business and even his work really stands off.

If we are trying to evaluate render engine realism it's only natural to compare real photos with the work of the best artists in the industry. The strange thing is that I saw the real kitchen photo I posted many months ago and it suddenly popped into my head while I was looking at Johaness' work. That's why I posted his work but in all fairness I should have posted another similar interior from (insert your favourite CGI artist) who uses Corona. If I take the best Corona image I can find and compare it to the two images I posted I can clearly see that Fstorm's render looks more photorealistic. Even when comparing it with the latest work from Jakub Cech which is drop dead gorgeous.

With regards to Fstorm the only way I can explain it is as if the materials have more "depth" along with what was already said about better tonemapping. In my opinion Fstorm users have the advantage of creating better looking materials because of the built-in tonemapping which gives them the ability to see materials how a camera would see it. In Corona you have to fiddle around with the sliders and LUTs until you feel satisfied and go on from there. Some Corona users get better at this, some master it, some don't and some never will.

I'm not trying to bash Corona because it's a great render engine and it got me excited about CGI again but as with all things there is always room for improvement.
« Last Edit: 2020-05-03, 17:46:04 by ynotsop »

2020-05-03, 17:35:17
Reply #118

Giona

  • Primary Certified Instructor
  • Active Users
  • ***
  • Posts: 150
    • View Profile
    • Giona Andreani 3D Visualization

edit: or maybe we should do a "photo or corona" game, to be fair to corona.

We did that one year ago, and the survey results ended with more people saying that the photo was the render :)

2020-05-03, 17:45:40
Reply #119

twoheads

  • Active Users
  • **
  • Posts: 261
    • View Profile
In my opinion Fstorm users have the advantage of creating better looking materials because of the built-in tonemapping which gives them the ability to see materials how a camera would see it. In Corona you have to fiddle around with the sliders until you feel satisfied and go on from there. Some Corona users get better at this, some master it, some don't and some never will.

Agree. Built-in tonemapping really? I need to give fstorm a try and fiddle with it.

there is always room for improvement.
As above!

To be honest I like to make shaders. Tweaking it, changing parameters here and there and see results in IR. It gives me satisfaction and peace of mind :)

2020-05-03, 17:53:19
Reply #120

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile

Agree. Built-in tonemapping really? I need to give fstorm a try and fiddle with it.



This is what I've been saying. Corona should move the defaults to be tonemapped and more photographic instead of the plain meh look it has now as default.

2020-05-03, 18:08:47
Reply #121

twoheads

  • Active Users
  • **
  • Posts: 261
    • View Profile
We did that one year ago, and the survey results ended with more people saying that the photo was the render :)

I like to check these "real or fake" surveys out of curiosity and for fun, but unfortunately most of them make no sense at all. Most even lame renders look "real" in small resolution. In most cases to find out if you're dealing with CGI or photo all you need to do is zoom in.

What was the resolution of compared pictures, were they the same size? It's not an accusation, just asking.

2020-05-03, 19:00:13
Reply #122

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
We did that one year ago, and the survey results ended with more people saying that the photo was the render :)

I like to check these "real or fake" surveys out of curiosity and for fun, but unfortunately most of them make no sense at all. Most even lame renders look "real" in small resolution. In most cases to find out if you're dealing with CGI or photo all you need to do is zoom in.

What was the resolution of compared pictures, were they the same size? It's not an accusation, just asking.


You are right, yes and no... :) Pics with incorrect light calculation, color mapping, etc. look fake even in veeeery small size. :)))


2020-05-03, 19:06:29
Reply #123

Giona

  • Primary Certified Instructor
  • Active Users
  • ***
  • Posts: 150
    • View Profile
    • Giona Andreani 3D Visualization
We did that one year ago, and the survey results ended with more people saying that the photo was the render :)

I like to check these "real or fake" surveys out of curiosity and for fun, but unfortunately most of them make no sense at all. Most even lame renders look "real" in small resolution. In most cases to find out if you're dealing with CGI or photo all you need to do is zoom in.

What was the resolution of compared pictures, were they the same size? It's not an accusation, just asking.

In the comments there were the two images in higher resolution.
I'm attaching them here.
But keep in mind that the survey was made for fun. I made the image in order to test and showcase the new caustics feature when it was released.
If you look at the images in high res, is quite obvious which one is the photo :)

2020-05-03, 19:36:40
Reply #124

BardhylM

  • Active Users
  • **
  • Posts: 31
    • View Profile
    • Behance
That photo looks so unrealistic, cameras should start using Fstorm tone mapping. I mean we do not need to have neutral results from render engine and apply our artistic choices.
We want Corona to do our job, completely. What's the point on putting materials and tweak the lights...etc... Why cant we have a "make my work" button.
And in the end, Corona is such a sweetheart, not trying to bash or anything... heehee.

- That is how some of you sound here, if you have not figure it out still.
I've been lurking here and don't get engaged in topics, mostly because they are always helpful and it's held up by really good members that keep on giving.
But this "Fstorm gud, others bed" has no place here i think. Yes it's been discussed here about tone mapping and ACES and everything, in a more educated approach.
Not in a "I think this looks better than this" approach, when there are biased views.

To everyone sorry for sounding so hostile, and making the forum a little bitter. Not my intention, I just had to...back in my cave again.


2020-05-03, 19:40:01
Reply #126

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
That photo looks so unrealistic, cameras should start using Fstorm tone mapping. I mean we do not need to have neutral results from render engine and apply our artistic choices.
We want Corona to do our job, completely. What's the point on putting materials and tweak the lights...etc... Why cant we have a "make my work" button.
And in the end, Corona is such a sweetheart, not trying to bash or anything... heehee.

- That is how some of you sound here, if you have not figure it out still.
I've been lurking here and don't get engaged in topics, mostly because they are always helpful and it's held up by really good members that keep on giving.
But this "Fstorm gud, others bed" has no place here i think. Yes it's been discussed here about tone mapping and ACES and everything, in a more educated approach.
Not in a "I think this looks better than this" approach, when there are biased views.

To everyone sorry for sounding so hostile, and making the forum a little bitter. Not my intention, I just had to...back in my cave again.



Understanding your irony... however, I guess no-one of us here asks for a "do my work" button, only a "give me a base that works as expected". :)


2020-05-03, 19:47:21
Reply #127

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
We did that one year ago, and the survey results ended with more people saying that the photo was the render :)

I like to check these "real or fake" surveys out of curiosity and for fun, but unfortunately most of them make no sense at all. Most even lame renders look "real" in small resolution. In most cases to find out if you're dealing with CGI or photo all you need to do is zoom in.

What was the resolution of compared pictures, were they the same size? It's not an accusation, just asking.

In the comments there were the two images in higher resolution.
I'm attaching them here.
But keep in mind that the survey was made for fun. I made the image in order to test and showcase the new caustics feature when it was released.
If you look at the images in high res, is quite obvious which one is the photo :)




Without clicking on the images I would say that the first one is 3D.
Super hard shadows on the doors, light is too cold despite the warm wooden tone of the floor, too extreme caustic... colors somehow do not harmonize.

2020-05-03, 19:58:35
Reply #128

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile

2020-05-03, 20:10:21
Reply #129

twoheads

  • Active Users
  • **
  • Posts: 261
    • View Profile

2020-05-03, 20:46:53
Reply #130

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile


10,11,13,14,15 look definitely better and more photorealistic than others. To be honest I thought they were photos.

I won ;)
[/quote]


Interesting... to me 10 looked super fake... the reflection on the floor slates, the tone of that little wooden furniture, the over saturation of the picture frame on the right,
the dull color tones... :)))))

Number 11 indeed looks nice... however it has visibly massive post work in it.

Look at 8 & 9... that smooth light changes without losing contrasts too much, without burning red tones in the woods, compression with less lost details, etc...
Typical characteristics of what I saw from FStorm until now.
And the image looks like having only minimal post work.
The eye can literally rest in that contrasts & tone mapping... THE typical difference I usually notice between photos and 3D with incorrect tone mapping, contrasts, etc.





2020-05-03, 20:59:07
Reply #131

PROH

  • Active Users
  • **
  • Posts: 1189
    • View Profile
Hmmm.... To me number 8 & 9 are super dull - clamped highlights and grey'ish.

I think nr. 10 definately looks better and more (a)live.

Are we talking taste and style?

2020-05-03, 21:17:59
Reply #132

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
Are we talking taste and style?
No. Just realism.

To me too, #10 and #11 are the best overall, UNTIL I start looking in detail and I realize that nothing is really in focus or has the detail and sharpness of the real thing.
While #8 and #9 overall look kind of 3d-ish (especially the walls), but when I start to look in detail my eyes can focus and the sharpness is there, at least in one point of the image.

Does anyone else agree that with corona is hard to get sharp details? As if the sampling doesn't get deep in small interstices. With fstorm, even if the overall image is not better than corona, the details are superb.

2020-05-03, 21:24:54
Reply #133

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Are we talking taste and style?
No. Just realism.

To me too, #10 and #11 are the best overall, UNTIL I start looking in detail and I realize that nothing is really in focus or has the detail and sharpness of the real thing.
While #8 and #9 overall look kind of 3d-ish (especially the walls), but when I start to look in detail my eyes can focus and the sharpness is there, at least in one point of the image.

Does anyone else agree that with corona is hard to get sharp details? As if the sampling doesn't get deep in small interstices. With fstorm, even if the overall image is not better than corona, the details are superb.



Yes, I agree with the sharpness point you mention.

2020-05-03, 21:34:32
Reply #134

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Are we talking taste and style?
No. Just realism.

To me too, #10 and #11 are the best overall, UNTIL I start looking in detail and I realize that nothing is really in focus or has the detail and sharpness of the real thing.
While #8 and #9 overall look kind of 3d-ish (especially the walls), but when I start to look in detail my eyes can focus and the sharpness is there, at least in one point of the image.

Does anyone else agree that with corona is hard to get sharp details? As if the sampling doesn't get deep in small interstices. With fstorm, even if the overall image is not better than corona, the details are superb.



#10 ???
So the fake plastikish shader on the floor (okay, that can be fixed), the strange, muddy wood color on that little furniture, the washed out contrasts on the walls or the visible GI-splotches on the walls don't bother your eye? I also find the displacement on the stone wall looks flat... not 3-dimensional.
Also, lack of detail on that tall brown door... (with the door handle that's too low) :)

Interesting that to my eye this is the weakest picture of all.





2020-05-03, 21:35:08
Reply #135

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile

2020-05-03, 21:38:54
Reply #136

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Hmmm.... To me number 8 & 9 are super dull - clamped highlights and grey'ish.

I think nr. 10 definately looks better and more (a)live.

Are we talking taste and style?

They are dull, yes... but the proportions in the contrasts and colors look like in photography! They look correct, as one would expect it in a photo.
Cranking up contrasts or highlights according to your taste shouldn't be a problem... especially when images (like this one) have so much (optical) information in it.



2020-05-03, 21:55:28
Reply #137

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
And here there's a comparison of the same scene, just for the fun of the experiment. Unfortunately the lighting is slightly different, even if the HDRI is the same.




2020-05-03, 22:14:36
Reply #138

twoheads

  • Active Users
  • **
  • Posts: 261
    • View Profile
I know few professional photographers and they use PS extensively. I'm not talking about developing process but pure postproduction. They're like michelin star chefs with butter. They put tons of this s..t everywhere. It takes a lot of time to make photo look great for publication believe me.

My reply about "which render looks better"  wasn't of course 100% serious. Now Imagine fstorm dev watching this topic laughing at us. "what the hell are they trying to achieve here?" :)

As for westkaai, we are discussing work we all know quite well (most of us) and I don't think we are talking about realism exclusively (it's not possible at this point). It would be fair to test both engines in some kind of blind test. Let's say 30 selected images we all have never seen before:  10 for corona, 10 for fstorm and 10 real photos and then we could debate for days and days and finally get conclusion "fstorm is 18% more photorealistic" who cares?

I'd really like to know what devs think about improvements we are talking about, I don't see any waving hands...

2020-05-03, 23:38:06
Reply #139

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
I know few professional photographers and they use PS extensively. I'm not talking about developing process but pure postproduction. They're like michelin star chefs with butter. They put tons of this s..t everywhere. It takes a lot of time to make photo look great for publication believe me.

My reply about "which render looks better"  wasn't of course 100% serious. Now Imagine fstorm dev watching this topic laughing at us. "what the hell are they trying to achieve here?" :)

As for westkaai, we are discussing work we all know quite well (most of us) and I don't think we are talking about realism exclusively (it's not possible at this point). It would be fair to test both engines in some kind of blind test. Let's say 30 selected images we all have never seen before:  10 for corona, 10 for fstorm and 10 real photos and then we could debate for days and days and finally get conclusion "fstorm is 18% more photorealistic" who cares?

I'd really like to know what devs think about improvements we are talking about, I don't see any waving hands...



Well, me for example, I do care... :)
And however... in my opinion, fiddling for hours in PS on an image cannot be the way to go.
3D software devs cannot expect that from users, nor do our clients want to wait for ages or pay a fortune for a single pic.
« Last Edit: 2020-05-03, 23:43:26 by Designerman77 »

2020-05-03, 23:39:52
Reply #140

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
And here there's a comparison of the same scene, just for the fun of the experiment. Unfortunately the lighting is slightly different, even if the HDRI is the same.






Thumbs up.
As I mentioned in some post before... you often see the better tone mapping once you desaturate an image to BW.
Cool test you did again!

Clearly visible differences...
« Last Edit: 2020-05-03, 23:44:10 by Designerman77 »

2020-05-03, 23:55:41
Reply #141

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
I would love to see the test above continued if you'll find time. Isn't the difference coming from mirrored HDRi? Every single loader, every engine, has this standardized differently (including starting angle!).

The region, top left corner is the most interesting difference.
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-05-04, 00:31:19
Reply #142

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation

2020-05-04, 02:28:04
Reply #143

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
I'm keeping noise in the images so we don't pay too much attention to shaders.

They're pretty much the same in my opinion.

What it is important, I think, and may be the reason why people think the problem is tone mapping, is that the shadows on the corners are stronger in Fstorm, even when the overall image has less contrast.
With the corona image, even after applying a strong post production, the corners will not have that nice shadow that you have in the fstorm image. May be using AO?
So I agree with the person that said that highlights and shadows are stronger in fstorm.

Also I notice that the geometry gets defined better (or sooner?) in fstorm. Look at the knob on the door, or the charger in the post on the counter

Just a note: I'm not saying that Fstorm is a better engine than Corona. Just comparing.





2020-05-04, 04:05:35
Reply #144

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
I notice you didn't enable LUT or modify Hightlight compression. Any reason why? Fstorm applies a LUT by default, so this is not a oranges to oranges compression, just wondering if you have a reason?

2020-05-04, 07:38:06
Reply #145

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Every single loader, every engine, has this standardized differently (including starting angle!).

This what hurt the most in those comparisons... Fstorm loader is indeed mirrored.. It seems like lupaz corrected it in the last render, not saying much about it.. Guys, checking the technical side is crucial in those kinds of comparisons, so as explaining the process... Everybody was taking a conclusion on renders with different lighting.... That sucks.

A lot of posts above were complete garbage.. comparing real photos to renders with completely different subjects, no controlled environment etc etc.. No need to expand more, this is useless. Let's not make this thread a complete junk pile.

Please lupaz, keep conducting your current test, it's interesting. But please make it fair and accurate. First thing would be to make them 1:1. So no missing materials/maps (if something do not work in an engine, should be changed by something that both supports in both scenes), same lighting (seems to be a little bit off still, based on the highlights upstairs), and a matching exposure. This would be a good starting point on which you can iterate.

Also, Fstorm as no caching solution, it's BF+BF. Not saying you have to change this one but this might explain some really fine details discrepancy some are talking about. Fstorm was remapping the glossiness range from [0.4;1.0] to [0.0-1.0], is it still the case? This could have a huge impact so keep that in mind.
« Last Edit: 2020-05-04, 07:53:39 by Fluss »

2020-05-04, 08:01:06
Reply #146

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
^thank you @Fluss

2020-05-04, 10:21:19
Reply #147

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Every single loader, every engine, has this standardized differently (including starting angle!).

This what hurt the most in those comparisons... Fstorm loader is indeed mirrored.. It seems like lupaz corrected it in the last render, not saying much about it.. Guys, checking the technical side is crucial in those kinds of comparisons, so as explaining the process... Everybody was taking a conclusion on renders with different lighting.... That sucks.

A lot of posts above were complete garbage.. comparing real photos to renders with completely different subjects, no controlled environment etc etc.. No need to expand more, this is useless. Let's not make this thread a complete junk pile.

Please lupaz, keep conducting your current test, it's interesting. But please make it fair and accurate. First thing would be to make them 1:1. So no missing materials/maps (if something do not work in an engine, should be changed by something that both supports in both scenes), same lighting (seems to be a little bit off still, based on the highlights upstairs), and a matching exposure. This would be a good starting point on which you can iterate.

Also, Fstorm as no caching solution, it's BF+BF. Not saying you have to change this one but this might explain some really fine details discrepancy some are talking about. Fstorm was remapping the glossiness range from [0.4;1.0] to [0.0-1.0], is it still the case? This could have a huge impact so keep that in mind.


May I ask what exactly is your point? You claim that Corona is a perfect render engine and devs. could go home and stop optimizing it?
It is NOT perfect and still needs a good amount of work until it can be called really photo-realistic. PERIOD !

Calling other peoples` thoughts "garbage" is an offense to all the guys here that take their precious time, energy and experience, willing to
make this great software even better.

I am sure the devs at Corona are not offended that users make suggestions. Actually, the nowadays quality of this software is based on the open minded attitude of the Corona dev team, not on an attitude of ignorance and "defending the existing limitations".

2020-05-04, 10:45:59
Reply #148

jms.lwly

  • Active Users
  • **
  • Posts: 163
    • View Profile
    • jms.lwly studio
it is absolute guesswork every time

Revisiting cjwidd's original point (things seem to have deviated from his original question - although brought about an interesting new conversation...) - every project I have the same back and forth in my setup:-

Use a LUT? Don't use a LUT, just use HC and Filmic...
Do less in VFB and more in Photoshop? Try to squeeze the very best output from VFB...
Use just HDRI for environment lighting? Use Corona Sun/Sky for lighting...

I realise with the all of above there is no correct answer - but I think this was the original question, which I fundamentally struggle with too.

2020-05-04, 10:54:00
Reply #149

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
May I ask what exactly is your point? You claim that Corona is a perfect render engine and devs. could go home and stop optimizing it?
It is NOT perfect and still needs a good amount of work until it can be called really photo-realistic. PERIOD !

Calling other peoples` thoughts "garbage" is an offense to all the guys here that take their precious time, energy and experience, willing to
make this great software even better.

I am sure the devs at Corona are not offended that users make suggestions. Actually, the nowadays quality of this software is based on the open minded attitude of the Corona dev team, not on an attitude of ignorance and "defending the existing limitations".

I'm just asking for a pragmatic approach. If you follow my post history, you'll actually see that I'm trying to push for some improvements as well. But come on, comparing a photograph of a kitchen with a render of a completely different scene, Making a poll with random renders... what's the point of this? This is useless as this proves nothing -> garbage. I'm not defending anything, you do.

I'm sorry for you but @BardhylM was right.

disclaimer : just to clarify, when I said no controlled environment, I was talking about the process of comparing real photos to renders. That part was not about lupaz test which is great even if it has some flaws I pointed out. So your test lupaz was not included in the garbage stuff.
« Last Edit: 2020-05-04, 11:27:49 by Fluss »

2020-05-04, 11:32:31
Reply #150

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
May I ask what exactly is your point? You claim that Corona is a perfect render engine and devs. could go home and stop optimizing it?
It is NOT perfect and still needs a good amount of work until it can be called really photo-realistic. PERIOD !

Calling other peoples` thoughts "garbage" is an offense to all the guys here that take their precious time, energy and experience, willing to
make this great software even better.

I am sure the devs at Corona are not offended that users make suggestions. Actually, the nowadays quality of this software is based on the open minded attitude of the Corona dev team, not on an attitude of ignorance and "defending the existing limitations".

I'm just asking for a pragmatic approach. If you follow my post history, you'll actually see that I'm trying to push for some improvements as well. But come on, comparing a photograph of a kitchen with a render of a completely different scene, Making a poll with random renders... what's the point of this? This is useless as this proves nothing -> garbage. I'm not defending anything, you do.

I'm sorry for you but @BardhylM was right.

disclaimer : just to clarify, when I said no controlled environment, I was talking about the process of comparing real photos to renders. That part was not about lupaz test which is great even if it has some flaws I pointed out. So your test lupaz was not included in the garbage stuff.


You split hairs in a problem that is visible even to a half blind. MY opinion.
The goal of this software ist to work photo real. Nothing less.

Improvements are not done by neglecting the flaws.
Do you also react like this if your clients ask for improvements in your renders? Like "shut up with your garbage, they are perfect" ???

2020-05-04, 11:43:25
Reply #151

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
I did not ask anyone to shut up.

I'm not neglecting anything.

We are trying to answer one question. Not "did you see it ?", not "Do you like it ?", but "Why?". That's the point.
« Last Edit: 2020-05-04, 11:47:26 by Fluss »

2020-05-04, 11:46:23
Reply #152

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
Eh, I feel like some of these comments are unnecessarily hostile and counterproductive...

2020-05-04, 12:02:03
Reply #153

ynotsop

  • Active Users
  • **
  • Posts: 39
    • View Profile
That's fine, I'll take the heat for posting those two pictures, but to me it illustrates how Fstorms renders mimic a camera. And I don't have to have a render based on a real photo to see it. The subjects are different but the overall materialism is very similar. To me the pictures look like they were taken with the same camera which goes back to my first point. Maybe I'm the only one seeing it and that's fine too.

The only thing garbage here is the attitude some of you have, and according to some only the ultra knowledgeable should be allowed to participate in discussions and the others can observe from their caves. If you don't agree with someone you can at least express it in a polite manner and we can keep the conversation on a level. I think it's fair to say that everyone participating in these forums are doing so with good intentions.

Eh, I feel like some of these comments are unnecessarily hostile and counterproductive...

I agree.

2020-05-04, 13:05:05
Reply #154

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
The thing is plenty of threads like this exist on the forum and they're all the same -> "Look how it's good", "Are you blind", "We want the same". Basically people moaning they want photo-real results without doing their part.
People keep posting example pictures from the internet, which they have no clue how they were made, how much post-production there is, how much effort the artist put in to look like that etc etc... Basically attributing the fact that the image looks good because of the renderer and not because of the artist.

So I tried to orient the discussion on a break it down approach. If it is so simple and Fstorm is so magical, show us examples you made and compare the results with what you can produce in Corona, 1:1. It's easy to make claims, it's less to prove them. The side by side comparison I've seen so far are not as mindblowing as some of you claim it to be. I never denied there is something common to some of the Fstorm renders we've seen, I'm just inclined to understand why rather than asking for something blindly...

Everybody is allowed to express an opinion, including me! So if I find these ever returning comments useless, I am allowed to express it. The thing is those are just making the thread going in circles like the 99 others before. Now if I offended some of you by calling those garbage, I apologize for that. I'm just tired of seeing the same stuff over and over again.

« Last Edit: 2020-05-04, 14:04:52 by Fluss »

2020-05-04, 13:22:56
Reply #155

bluebox

  • Active Users
  • **
  • Posts: 268
    • View Profile
I think we should wrap this up at this point to a conclusion that everyone is trying to contribute towards the same goal - improvements in the area that as we can all see is really really wanted by many.

Current tone mapping is not working as it should in many areas that were discussed back and forth many times.
Original contrast destroys blacks, oversaturates colours to a point that you are no longer able to see the original texture used which is really annoying when working with product design or anything that requires more than just a punchy-candy picture.
LUTs are also not the way to go, they should be used to get a certain degree of mood imho, not the base acceptable picture.

What concerns me the most is the fact that this in one of many topics abut all those issues and we still can't get noone from the Corona team to participate.
We still get all those "key new features" like faster color picker and the important stuff as we can see and extrapolate from the most wanted features poll gets neglected.

Those discussions basicaly IMHO end up in a vacuum and we get this BS policy of a roadmap that is all but "no promisses".
If the community wants certain features and they end up not being developed and we can't get any promisses then I vote for bringing back the box licence so I can upgrade when and only when the features that community wants get implemented.

My few cents.

2020-05-04, 14:37:20
Reply #156

agentdark45

  • Active Users
  • **
  • Posts: 581
    • View Profile
Regarding that image of my earlier test, as others noted, I would have to use the exact same HDRI and closest as possible materials, not a simple conversion.

So I repeated the test. Same HDRI, Same LUT (KimAmlan02), Matched white balance. very simple materials.

Do you really see a 50% bump in realism in the Fstorm one? Maybe I don't have your eyes, it is entirely possible that I actually don't have the perceptual capacity.



Anyway, now that I've conducted the test in a more proper manner, my point stands. I will write it in bold so you don't miss it and you stop arguing with something I'm not saying:

Fstorm has better tonemapping.   I wish Corona had a DSLR like tonemapping

However, I don't think that is the main reason for the Fstorm community producing seemingly more realistic images more consistently than the corona render community.
I think there is a small number of users (Agentdark45 included) who have superior understanding and perceptual abilities and actually notice the super slight differences between the two renderers, which are real and measurable.

But I believe what MOST people notice is a vibrant community that attracts great artists who are learning faster and feeling comfortable sooner with Fstorm, and thus, becoming good enough that, overall, the Fstorm community is producing better renders, more artistic, more realistic.

Just to be super clear, FSTORM HAS BETTER TONEMAPPING, 100% agree. And it has a real impact in realism! I wish Corona implements this soon
ps. I also wish corona becomes a lot more beginner and artist friendly by making it MUCH EASIER to produce the best possible images it already can, which, BTW would inevitably be WORST than Fstorm regarding tonemapping. At least until corona implements a better tonemapping, which I will continue to push and wish for.

Apologies if I missed the spirit of your post, it seemed contradictory at the time of posting. I'm glad to see you are supporting the call for improved tonemapping in Corona.

My main contention here is that some posters are claiming:

1. "Everything is fine with Corona as is, zero perceptual difference in tone-mapping between Fstorm and Corona; the issue is noobish user error being fixated on stock settings".
2. While other posters are stating that there is a clear and obvious limitation with Corona's tonemapping (despite Bertrand Benoit levels of Corona mastery and tweaking), and we want improved colour space/ACES like tonemapping implemented as a priority over other seemingly trivial features that eat up precious dev time.

Regarding your recent comparison, yes there is a difference between the two (perhaps not 50%) but I can clearly see the black crush in the shadows issue I have mentioned before in trying to achieve a contrasty punchy image whilst retaining not overly clamped highlights (look at how much detail is lost in the lower right shadow on the floor in the Corona version, along with the shadow at the base of the exterior sphere, the soft shadow gradient on the left hand interior wall, and the tendancy for Corona to go to pure black in corner gaps - the dynamic range of the shadows have been butchered). Now scale this issue up to images that mostly deal with subtle tonal variations (i.e moody or brightly lit product shots).

As a side note outside of tonemapping, you can also see the filtering issue I mentioned previously in the Corona example. Look at how much detail has been smoothed out of the concrete walls and wood material in the Corona version.
« Last Edit: 2020-05-04, 14:55:43 by agentdark45 »
Vray who?

2020-05-04, 14:52:58
Reply #157

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
Less strawmen and attacks or devs will lock this epic thread 😀

Everyone really wants the same, there is just various level of allergy to hyperbolic expectations and conclusions.
This might be tough topic even for devs, I've only ever seen Pixar doing indepth reaserch into realism that goes beyond the technical part, esp. because so much of it strays into subjective.

I personally love comparisons, I know they are flawed but lit can be peered from them. Fluss is only suggesting to do this more thoroughly as it will help avoid reaching premature and incorrect conclusions.

That takes unfortunately unholy amount of effort & time, which everyone lacks. So all my respect to people with patience to fiddle with this. More tests, less assured conclusions.
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-05-04, 16:01:56
Reply #158

ynotsop

  • Active Users
  • **
  • Posts: 39
    • View Profile
I completely agree with the fact that in the end things have to be investigated in a practical manner. It's cool to see these comparisons lolec and lupaz are doing. It's also saddening to see that this issue has been already brought up way back in 2017 (or maybe even before that?).
« Last Edit: 2020-05-04, 18:04:08 by ynotsop »

2020-05-04, 16:34:52
Reply #159

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
@Fluss
You said it yourself:
So photorealism really isn't that simple to define as there is a whole lot of intricacies. The key is to use references and observe the world around you. I often look at light interactions around me and a lot of these observations are making me think "If I'd see this in one of my renders, i'd probably think this is a bug".

It would be really interesting to make a thread with great images and review them to highlight why it works well.

The reason for showing renderings and also photos is to observe.
You can do all the tests you want being methodical and practical, but in the end there's also a perception component that we can learn from looking at different images.
So I disagree. It's not garbage. It's learning. Isn't AI based on this?

If I may add my two cents, this discussion can be boiled down to braking bad's quote: "after all, how pure can pure be?"
For most people 96% purity is more than enough and for some people difference between 96 and 99 is not just noticeable it is like tremendous gulf.
and finally get conclusion "fstorm is 18% more photorealistic" who cares?

The whole discussion, I think, it's because we strive to make more compelling images. Images that sell.
Clients will "buy" faster (18%? :)) an image if they FEEL they like it.

Some Fstorm images have a feeling that is very attractive.
We're trying to pin-point it, right?


Original contrast destroys blacks, oversaturates colours to a point that you are no longer able to see the original texture used which is really annoying when working with product design or anything that requires more than just a punchy-candy picture.

How do you guys know that this is the cause? A bad tone mapping...
Could it be that the shadows are wrongly represented by the engine in the first place, making them more bland with a gradient that makes contrast to flatten them?

That's what I'm trying to do with the black and white noisy tests:
I want to discard tone mapping, sharpening, LUT, bloom, DOF, color, textures, contrast; and I want to compare the core of the engine: How it resolves light and shadows, antialiasing, BRDF and sampling of small objects and details.



2020-05-04, 17:11:38
Reply #160

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
I notice you didn't enable LUT or modify Hightlight compression. Any reason why? Fstorm applies a LUT by default, so this is not a oranges to oranges compression, just wondering if you have a reason?

What I tried to do was to remove from the Fstorm frame buffer all post production. For this test I didn't care about Fstorm defaults. If we remove post production stuff and end up with the same rendering as corona, that would mean that you can get an Fstorm render with little effort, with Corona's controls or other software.

What I want to do next is add DOF to the test and see what happens.

2020-05-04, 17:23:48
Reply #161

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
LMAO, you're quoting me totally out of context. If you still did not get my point then there is not so much I can do. Anyway, I'll leave it here as I don't want to pollute the thread even more. let's move on
« Last Edit: 2020-05-04, 17:31:45 by Fluss »

2020-05-04, 17:42:45
Reply #162

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
I did not ask anyone to shut up.

I'm not neglecting anything.

We are trying to answer one question. Not "did you see it ?", not "Do you like it ?", but "Why?". That's the point.


Then I'm totally with you. :)

2020-05-04, 17:54:14
Reply #163

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
I think we should wrap this up at this point to a conclusion that everyone is trying to contribute towards the same goal - improvements in the area that as we can all see is really really wanted by many.

Current tone mapping is not working as it should in many areas that were discussed back and forth many times.
Original contrast destroys blacks, oversaturates colours to a point that you are no longer able to see the original texture used which is really annoying when working with product design or anything that requires more than just a punchy-candy picture.
LUTs are also not the way to go, they should be used to get a certain degree of mood imho, not the base acceptable picture.

What concerns me the most is the fact that this in one of many topics abut all those issues and we still can't get noone from the Corona team to participate.
We still get all those "key new features" like faster color picker and the important stuff as we can see and extrapolate from the most wanted features poll gets neglected.

Those discussions basicaly IMHO end up in a vacuum and we get this BS policy of a roadmap that is all but "no promisses".
If the community wants certain features and they end up not being developed and we can't get any promisses then I vote for bringing back the box licence so I can upgrade when and only when the features that community wants get implemented.

My few cents.




With your observations about tone mapping misbalance, burning colors, exaggerated contrasts in blacks, etc. you 100% nailed it!
And that's also basically my point: there is no way to talk those facts away.

However, regarding your assumption that Corona devs "don't care"... hmmm... I am sure the do care and read this thread very carefully, since it
criticized the very core of Corona - in a pro active way. :)
About one year ago, some guys and me massively complained about the new denoiser messing up round edges & bump textures to a point that one could not work anymore.
Relatively quickly. the devs implemented the new AI denoiser that does a good job.

Wouldn't be surprised to see a new, advanced tone mapping in Cor 7.
Let´s hope.







2020-05-04, 18:05:08
Reply #164

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Regarding your recent comparison, yes there is a difference between the two (perhaps not 50%) but I can clearly see the black crush in the shadows issue I have mentioned before in trying to achieve a contrasty punchy image whilst retaining not overly clamped highlights (look at how much detail is lost in the lower right shadow on the floor in the Corona version, along with the shadow at the base of the exterior sphere, the soft shadow gradient on the left hand interior wall, and the tendancy for Corona to go to pure black in corner gaps - the dynamic range of the shadows have been butchered). Now scale this issue up to images that mostly deal with subtle tonal variations (i.e moody or brightly lit product shots).

As a side note outside of tonemapping, you can also see the filtering issue I mentioned previously in the Corona example. Look at how much detail has been smoothed out of the concrete walls and wood material in the Corona version.
[/quote]






Brilliant eye & observation, Sir!

Exactly those things ( black shadow areas, wrong dark-light transitions, etc.) make you literally feel like somehow not being abele to look at an image and get the information your eye looks for.
A bit like listening to a music recording that has distorted frequencies.



2020-05-04, 18:32:51
Reply #165

hldemi

  • Active Users
  • **
  • Posts: 55
    • View Profile
I notice you didn't enable LUT or modify Hightlight compression. Any reason why? Fstorm applies a LUT by default, so this is not a oranges to oranges compression, just wondering if you have a reason?

What I tried to do was to remove from the Fstorm frame buffer all post production. For this test I didn't care about Fstorm defaults. If we remove post production stuff and end up with the same rendering as corona, that would mean that you can get an Fstorm render with little effort, with Corona's controls or other software.

What I want to do next is add DOF to the test and see what happens.

There was similar thread on Fstorm FB group and I remember Johannes Lindqvist saying that They tested same scene with Fstorm and Corona and when they turned off Fstorm tonemapping the render was practically identical. Also Andrey said Fstorm uses Canon 300D LUT right now as default. SO to try that LUT in corona and compare images would be interesting.
« Last Edit: 2020-05-04, 18:42:46 by hldemi »

2020-05-04, 19:09:25
Reply #166

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
Regarding your recent comparison, yes there is a difference between the two (perhaps not 50%) but I can clearly see the black crush in the shadows issue I have mentioned before in trying to achieve a contrasty punchy image whilst retaining not overly clamped highlights (look at how much detail is lost in the lower right shadow on the floor in the Corona version, along with the shadow at the base of the exterior sphere, the soft shadow gradient on the left hand interior wall, and the tendancy for Corona to go to pure black in corner gaps - the dynamic range of the shadows have been butchered). Now scale this issue up to images that mostly deal with subtle tonal variations (i.e moody or brightly lit product shots).

As a side note outside of tonemapping, you can also see the filtering issue I mentioned previously in the Corona example. Look at how much detail has been smoothed out of the concrete walls and wood material in the Corona version.


Brilliant eye & observation, Sir!

Exactly those things ( black shadow areas, wrong dark-light transitions, etc.) make you literally feel like somehow not being abele to look at an image and get the information your eye looks for.
A bit like listening to a music recording that has distorted frequencies.

I hope you don't take this the wrong way, but I think the fairest way to do a blind test is when the participants are actually blind.

I swapped the Corona and Fstorm renders, I removed the watermark from the Fstorm Render and added a Fstorm Watermark to the Corona render one.

I don't know what that means, and I don't mean to say that there is no difference or that Fstorm doesn't have some areas where it is superior. But at least in this particular test, it seems like Corona actually exhibits the great properties you are looking for, as long as it has a Fstorm watermark on top of it.

Hope this information can help us all reach further conclusions :) Please don't punch me !

2020-05-04, 19:51:25
Reply #167

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
I hope you don't take this the wrong way, but I think the fairest way to do a blind test is when the participants are actually blind.

I swapped the Corona and Fstorm renders, I removed the watermark from the Fstorm Render and added a Fstorm Watermark to the Corona render one.

I don't know what that means, and I don't mean to say that there is no difference or that Fstorm doesn't have some areas where it is superior. But at least in this particular test, it seems like Corona actually exhibits the great properties you are looking for, as long as it has a Fstorm watermark on top of it.

Hope this information can help us all reach further conclusions :) Please don't punch me !
[/quote]



So you basically say that you tricked everyone here by swapping the FS-watermark and put it on Corona render?
Okay... however, those characteristics others mentioned ( too black shadows, etc. etc. ) actually do exist in Corona when you touch the contrast and I observe it day by day.

I don't know what you did with the FStorm render... but from the FStorm renders I have seen up to date, I cannot remember those burnt colors, black shadows with no detail, etc.
This is totally non typical for F-storm.

So?

Sorry, but this all is surely not a "mass hallucination". :)

2020-05-04, 20:10:02
Reply #168

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
If it wasn't a typical Fstorm render, wouldn't you say you would have been able to notice?

Not only you didn't immediately notice but you even complained about the same things you are complaining about corona. And not only you, so maybe there is something else to it than objective observation.

Not sure why you label it as "tricked"  we are talking about comparisons and every serious study requires the elimination of variables.

I'm sorry you didn't like the results of the experiment, I bet that if the result was different and you actually noticed, you would have felt really good about that fact, and confirm your previous notions. So I think just because you didn't get the expected result, there is no reason to say "I did something to the Fstorm version".


2020-05-04, 21:12:06
Reply #169

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
I created this thread with a much more pedestrian goal in mind, basically to ask how others use the tonemapping features. Think of the following as a series of T/F statements, For example:

In general...
  • I never use contrast < 1.0
  • I never use highlight compression <1.0
  • I never use highlight compression <3.0
  • I never use contrast > 1.0 in combination with a LUT
  • I never use filmic shadows in combination with a LUT
  • I never use LUTS
  • I never use LUTS at an opacity of 1.0
  • I never use the built-in vignetting
  • I never use color tint
  • I never use filmic highlights
  • I always use filmic highlights = 1.0
  • I always use filmic highlights = 1.0 when using a LUT
  • I always use Dubcat ACES emulation
  • I always increase / decrease saturation a little bit
  • I never adjust saturation
  • I never use green-magenta tint
  • I always use sharpening / blurring in the VFB - here are the settings
  • I always keep exposure = 0 and modify light sources instead of EV spinner
  • I never use bloom / glare
« Last Edit: 2020-05-04, 21:16:29 by cjwidd »

2020-05-04, 21:33:26
Reply #170

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
There was similar thread on Fstorm FB group and I remember Johannes Lindqvist saying that They tested same scene with Fstorm and Corona and when they turned off Fstorm tonemapping the render was practically identical. Also Andrey said Fstorm uses Canon 300D LUT right now as default. SO to try that LUT in corona and compare images would be interesting.

I don't really appreciate the F-storm vs. Corona Renderer discussion happening on this thread, but this post^ is a really interesting contribution. It sounds like Johannes (unsurprisingly) already did controlled experiments - pretty compelling.

2020-05-04, 21:46:06
Reply #171

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
If it wasn't a typical Fstorm render, wouldn't you say you would have been able to notice?

Not only you didn't immediately notice but you even complained about the same things you are complaining about corona. And not only you, so maybe there is something else to it than objective observation.

Not sure why you label it as "tricked"  we are talking about comparisons and every serious study requires the elimination of variables.

I'm sorry you didn't like the results of the experiment, I bet that if the result was different and you actually noticed, you would have felt really good about that fact, and confirm your previous notions. So I think just because you didn't get the expected result, there is no reason to say "I did something to the Fstorm version".


Why should I be "disappointed"???? I am very glad about it, actually, since I work 100% in Corona and very much appreciate this program.
But: it still remains a fact that - despite that funny inversion of pics - tons of people see what all of us discussed about.
I don´t think that it is because FStorm users are all geniuses and Corona users just dilettantes.

The question remains: do we (the users) want Corona to evolve as a render engine, or rather relativize and discuss those things away which could be improved?

I claim that a few test do not have the same meaningfulness like hundreds and thousands of observations, don't you think?



2020-05-04, 21:47:37
Reply #172

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
I don't really appreciate the F-storm vs. Corona Renderer discussion happening on this thread, but this post^ is a really interesting contribution. It sounds like Johannes (unsurprisingly) already did controlled experiments - pretty compelling.
[/quote]


I do not see it as "versus"... but rather "what can we learn from each other and improve the status quo".

2020-05-04, 23:03:46
Reply #173

romullus

  • Global Moderator
  • Active Users
  • ****
  • Posts: 7664
  • Let's move this topic, shall we?
    • View Profile
    • My Models
  • I never use contrast < 1.0
  • I never use highlight compression <1.0
  • I never use highlight compression <3.0
  • I never use contrast > 1.0 in combination with a LUT
  • I never use filmic shadows in combination with a LUT
  • I never use LUTS
  • I never use LUTS at an opacity of 1.0
  • I never use the built-in vignetting
  • I never use color tint
  • I never use filmic highlights
  • I always use filmic highlights = 1.0
  • I always use filmic highlights = 1.0 when using a LUT
  • I always use Dubcat ACES emulation
  • I always increase / decrease saturation a little bit
  • I never adjust saturation
  • I never use green-magenta tint
  • I always use sharpening / blurring in the VFB - here are the settings
  • I always keep exposure = 0 and modify light sources instead of EV spinner
  • I never use bloom / glare

False to all.
I'm not Corona Team member. Everything i say, is my personal opinion only.
My Models | My Videos | My Pictures

2020-05-04, 23:06:04
Reply #174

jms.lwly

  • Active Users
  • **
  • Posts: 163
    • View Profile
    • jms.lwly studio
In general...
  • I never use contrast < 1.0
  • I never use highlight compression <1.0
  • I never use highlight compression <3.0
  • I never use contrast > 1.0 in combination with a LUT
  • I never use filmic shadows in combination with a LUT
  • I never use LUTS
  • I never use LUTS at an opacity of 1.0
  • I never use the built-in vignetting
  • I never use color tint
  • I never use filmic highlights
  • I always use filmic highlights = 1.0
  • I always use filmic highlights = 1.0 when using a LUT
  • I always use Dubcat ACES emulation
  • I always increase / decrease saturation a little bit
  • I never adjust saturation
  • I never use green-magenta tint
  • I always use sharpening / blurring in the VFB - here are the settings
  • I always keep exposure = 0 and modify light sources instead of EV spinner
  • I never use bloom / glare

This list is exactly the back and forth that I go through during almost every new project - I wish I could decide on a definitive personal favourite - but I can’t. Would love to know everyone else’s approach to all of the above.


2020-05-04, 23:13:14
Reply #175

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
From the default, I only change exposure trying to keep it as close to zero as possible (I  heard that corona has it easier if that happens), I increase highlight compression to about 2 and contrast to about 3.
I usually don't use LUTs because I do the color adjustments in camera raw in photoshop.

I do use bloom just to soften the aliasing.

Vignetting for some exteriors.





I created this thread with a much more pedestrian goal in mind, basically to ask how others use the tonemapping features. Think of the following as a series of T/F statements, For example:

In general...
  • I never use contrast < 1.0
  • I never use highlight compression <1.0
  • I never use highlight compression <3.0
  • I never use contrast > 1.0 in combination with a LUT
  • I never use filmic shadows in combination with a LUT
  • I never use LUTS
  • I never use LUTS at an opacity of 1.0
  • I never use the built-in vignetting
  • I never use color tint
  • I never use filmic highlights
  • I always use filmic highlights = 1.0
  • I always use filmic highlights = 1.0 when using a LUT
  • I always use Dubcat ACES emulation
  • I always increase / decrease saturation a little bit
  • I never adjust saturation
  • I never use green-magenta tint
  • I always use sharpening / blurring in the VFB - here are the settings
  • I always keep exposure = 0 and modify light sources instead of EV spinner
  • I never use bloom / glare

2020-05-04, 23:14:57
Reply #176

bluebox

  • Active Users
  • **
  • Posts: 268
    • View Profile
In general...
  • I never use contrast < 1.0
  • I never use highlight compression <1.0
  • I never use highlight compression <3.0
  • I never use contrast > 1.0 in combination with a LUT
  • I never use filmic shadows in combination with a LUT
  • I never use LUTS
  • I never use LUTS at an opacity of 1.0
  • I never use the built-in vignetting
  • I never use color tint
  • I never use filmic highlights
  • I always use filmic highlights = 1.0
  • I always use filmic highlights = 1.0 when using a LUT
  • I always use Dubcat ACES emulation
  • I always increase / decrease saturation a little bit
  • I never adjust saturation
  • I never use green-magenta tint
  • I always use sharpening / blurring in the VFB - here are the settings
  • I always keep exposure = 0 and modify light sources instead of EV spinner
  • I never use bloom / glare

This list is exactly the back and forth that I go through during almost every new project - I wish I could decide on a definitive personal favourite - but I can’t. Would love to know everyone else’s approach to all of the above.



Because you just can't decide.
I mean you can find the tone mapping settings you like and stick to them. But then if you find yourself in a situation that you really need more contrast and crank it up - all of a sudden your shaders will mostly break apart, saturation will go crazy and all this nonsense.

Just tried Fstorm again - have the demo installed but its pain in the @ss on my gtx1060 - contrast does not affect saturation - awesome, reducing burn value does not make the image instantly dull, in fact it does not make the image dull at all - you can even go with negative value - again awesome.

Try for yourself.

Can't really wait what Nvidia shows this year :)

2020-05-04, 23:24:16
Reply #177

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
From the default, I only change exposure trying to keep it as close to zero as possible (I  heard that corona has it easier if that happens), I increase highlight compression to about 2 and contrast to about 3.
I usually don't use LUTs because I do the color adjustments in camera raw in photoshop.

I do use bloom just to soften the aliasing.

Vignetting for some exteriors.



HC generally kills your light dynamics. I try to keep it to 1 and rather use the Filmic Highlights.

One of the Corona devs gave the hint to use the Amland LUTs and leave the Cam settings as basic as possible.
I tried this and it became my new standard.
Works quite nicely.

HC 1
Contrast max. 3
Filmic Highlights 0 - 9
Filmic shadow... better keep it on 0 / max 0.2
« Last Edit: 2020-05-05, 11:39:32 by Designerman77 »

2020-05-04, 23:26:47
Reply #178

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@romullus In general though, like 90% of cases - obviously the terms 'never' and 'always' are sort of tautological

2020-05-04, 23:44:19
Reply #179

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
I created this thread with a much more pedestrian goal in mind, basically to ask how others use the tonemapping features. Think of the following as a series of T/F statements, For example:

In general...
  • I never use contrast < 1.0
  • I never use highlight compression <1.0
  • I never use highlight compression <3.0
  • I never use contrast > 1.0 in combination with a LUT
  • I never use filmic shadows in combination with a LUT
  • I never use LUTS
  • I never use LUTS at an opacity of 1.0
  • I never use the built-in vignetting
  • I never use color tint
  • I never use filmic highlights
  • I always use filmic highlights = 1.0
  • I always use filmic highlights = 1.0 when using a LUT
  • I always use Dubcat ACES emulation
  • I always increase / decrease saturation a little bit
  • I never adjust saturation
  • I never use green-magenta tint
  • I always use sharpening / blurring in the VFB - here are the settings
  • I always keep exposure = 0 and modify light sources instead of EV spinner
  • I never use bloom / glare

This is the most interesting thing for me. As this settings are completely different to what I use normally and I actually have NO IDEA on what are the best ones to use.

I edited my scene to use only Corona materials and no copyrighted assets. It means it looks boring but it serves the purpose.

I will post the scene here for anyone to download and try their preferred settings and normal workflow over it.  Please post an image resulting from your workflow as well as your preferred settings. I'll start! This might be very interesting!!

Ok, so this is corona Defaults (only changed EV otherwise its burned up)


I never use that as a starting point though. My default settings are this:


I rarely leave it at that though. So My workflow always involves some amount of tweaking. For this particular image, a quick tweaking ends up like this:


Not saying that is what I would consider a finished image. But as a quick summary of my normal workflow, and not changing anything about the scene, I would say that's close.

So, If you want to try:

Download the scene from here (created in a student version so none of it can be used for commercial purposes, only learning):https://share.getcloudapp.com/yAu20wZX
Do not change anything about the scene! Leave light and materials the same, only tweak tonemapping.
Use your usual settings and post the results. It doesn't have to be all inside corona, if you usually take the image outside of corona to finish retouching, go ahead! do that and let's see what a difference workflow makes.







2020-05-04, 23:57:21
Reply #180

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
Just tried Fstorm again - have the demo installed but its pain in the @ss on my gtx1060 - contrast does not affect saturation - awesome, reducing burn value does not make the image instantly dull, in fact it does not make the image dull at all - you can even go with negative value - again awesome.


This is true. I wish corona at least behaved as a real camera and I could take the rendering from corona as pure raw and do a camera raw in photoshop. I still couldn't figure how to do this.

2020-05-05, 02:10:56
Reply #181

agentdark45

  • Active Users
  • **
  • Posts: 581
    • View Profile
If it wasn't a typical Fstorm render, wouldn't you say you would have been able to notice?

Not only you didn't immediately notice but you even complained about the same things you are complaining about corona. And not only you, so maybe there is something else to it than objective observation.

Not sure why you label it as "tricked"  we are talking about comparisons and every serious study requires the elimination of variables.

I'm sorry you didn't like the results of the experiment, I bet that if the result was different and you actually noticed, you would have felt really good about that fact, and confirm your previous notions. So I think just because you didn't get the expected result, there is no reason to say "I did something to the Fstorm version".

Yeah, I'm not buying this. You're saying you went through the effort of purchasing a full Fstorm license, matching the watermark 1:1 in photoshop, and your "swapped" Fstorm render just happens to exhibit exactly the same grain pattern and concrete microdetail as your previous tests...
« Last Edit: 2020-05-05, 02:17:40 by agentdark45 »
Vray who?

2020-05-05, 02:19:10
Reply #182

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
Quote
contrast does not affect saturation

This one is bit unfair, Corona's Contrast is identical to Adobe's, it's basic contrast. It's not mistake that it doesn't operate on luminosity channel only, or offers selective smart saturation (Vibrance).
Some raw editors replaced Saturation with Vibrance (and call it Saturation again), but since we already have saturation, I would just add vibrance. The saturation is the weirdly integrated feature.. doesn't act like anything I know (affects luminosity&contrast way too much).

This is true. I wish corona at least behaved as a real camera and I could take the rendering from corona as pure raw and do a camera raw in photoshop. I still couldn't figure how to do this.

When people say they want Corona to behave like "camera", they actually want it to behave like raw converter. SOOC Jpegs from Camera are done by their own internal raw converter. The raw file itself is linear.

Corona's linear is also "raw", just 32bit floating point raw with massive dynamic range, which needs tonemapping. Even photos when merged into higher dynamic range require tonemapping, you can't just open them in Lightroom or CapturePro and use the base
Most raw converters don't allow these format at all, or they don't work correctly with them.

I once proposed devs what I consider could work as workaround. To allow Corona to save directly into .dng format. Adobe's DNG has both 16bit and 32bit floating point options. And I thought perhaps this would confuse raw editors less, effectively "trick" them.

But even right now it sort of "works". When you use "Open as" in Photoshop, you can open linear floating point (either 16bit or 32bit .exr/.hdr), and you will use absolutely the same workflow ACR would offer you if you loaded merged HDR photo. But you have to use the older process, or accept the odd behavior it can exhibit.
(I didn't test one more theory, if artificially clamping dynamic range before would make ACR more predictable).

The only thing Corona devs can do, is integrate different tonemapping curve (global tonemapper are always just curves). But it does work like camera, and you can export raw.


Whether Lolec is trolling or saying truth, it's genius post :- ).
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-05-05, 03:05:27
Reply #183

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
Yeah, I'm not buying this. You're saying you went through the effort of purchasing a full Fstorm license, matching the watermark 1:1 in photoshop, and your "swapped" Fstorm render just happens to exhibit exactly the same grain pattern and concrete microdetail as your previous tests...

No, that would be silly!

I rendered a gray box in front of the Fstorm camera to get the raw watermark pattern. I then adjusted the brightness and used color burn to essentially revert what Fstorm is doing. I needed 3 layers to achieve that, as the shadow areas didn't quite work the same as the light areas.  Then I used that same "raw" Fstorm pattern and overlayed it to the Corona render version.

Here is a video of me turning on layers in photoshop :)  https://share.getcloudapp.com/WnuGqzoO

You can download the original image I posted and zoom in to see how much of a terrible job I did!

I know it can be hard to be confronted with information you don't like. But I have nothing to gain out of convincing anyone that Corona is better or worst, I'm a corona user! I want corona to be the best possible render engine! Actually, corona render is my main source of income now that I think of it... so NO I don't want it to stop improving!  BUT! I think it can improve more drastically by being easier to use and easier to achieve it's potential, than to extend it's potential even further. LETS DO BOTH, but lets do the easy one first!



2020-05-05, 09:40:18
Reply #184

bluebox

  • Active Users
  • **
  • Posts: 268
    • View Profile
Quote
contrast does not affect saturation

This one is bit unfair, Corona's Contrast is identical to Adobe's, it's basic contrast. It's not mistake that it doesn't operate on luminosity channel only, or offers selective smart saturation (Vibrance).
Some raw editors replaced Saturation with Vibrance (and call it Saturation again), but since we already have saturation, I would just add vibrance. The saturation is the weirdly integrated feature.. doesn't act like anything I know (affects luminosity&contrast way too much).


How is this unfair Juraj ? I converted a simple Corona playground scene of mine and when boosting contrast to 5 out of normall 1 my wood shaders started a new after psychedelic like life. They went to the moon with saturation. When doing the same in Fstorm the image remains visually 90% the same, only gets punchier.

I know you optimize your workflow and for sure you have some sort of a shader library of your own. This works great. Lets you streamline your workflow and become super efficient especially when incorporating triplanar wherever possible.
Only problem is that if you prepare your shaders in a controlled dedicated scene (say contrast 1) and then bring them to a more contrasty scene all of a sudden the colours break apart. This should not be the case.

Now, whether you call it saturation, vibrance or other is not important. If we get both controlls- contrast and saturation, imho contrast should operate only on luminosity, saturation should be saturation. And even saturation here does not work as it should - it shifts hue.

I agree that people wanting magicall "make dis shiet photoreal" button will get dissapointed because it will not happen as this comes from a lot of variables - working with references, matching colours etc. but lets just make the tools that we have work as they should. and upgrade the ones that are just bad at the moment - like HC, contrast, etc.

2020-05-05, 10:05:52
Reply #185

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
Pretty curious how much of this conversation will even be relevant if the tonemapping options get overhauled for an upcoming version. Presumably we wouldn't be talking about contrast, highlight compression, etc. in the same way.

2020-05-05, 10:45:11
Reply #186

Juraj

  • Active Users
  • **
  • Posts: 4324
    • View Profile
    • studio website
Because Corona Contrast is not part of tonemapping, it's grading tool and works correctly. Contrast 5 means 5 100 contrast modifiers in Photoshop.

It works the same in Photoshop, Lightroom & ACR.

VFB right now has small subset of post tools and ugly tonrmapper. That doesnt mean those post tools are wrong.
We dont need to copy every arbitrary decision someone else did.

Tool names are important, so yes ut matters how we call them. Contrast has nithing to do with anyones toxic looking shaders.
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2020-05-05, 11:25:50
Reply #187

MattiasD

  • Active Users
  • **
  • Posts: 15
    • View Profile
    • Key Visuals
I really like these discussions. It`s maybe not always that constructive or correct, but it does always learn me new things.

@Juraj Talcik: Could you explain how to convert a linear (not tonemapped) image with aces? I downloaded OpenColorIO and Aces and tried to use this in photoshop, but I`m not sure what input and output i should choose from al the options and not even sure i should use photoshop at all for this :).

@lolec:
Here is my take on the scene:
Lut: Filmic Very High Contrast (log color space)


2020-05-05, 12:02:19
Reply #188

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
lol funny little trick you did there @lolec, and thanks for posting the scene it is an interesting challenge. I gave it a try in Vray keeping everything the same, textures, lighting, ev etc only swapped the camera for a physical but settings are same. Settings were as follows, default Vray, Vray + LUT Canon 300C, Vray/Aces. I think the LUT is actually pretty close to the Aces result but less range in the top end.
I think for Vray using the exposure control gives you a better result at EV 0 but for now we can ignore that.

Vray





Vray + LUT Canon 300C




Vray/Aces








LUT here:
https://cpn.canon-europe.com/content/news/canon_log_luts_now_available.do
BT709_CanonLog2-to-BT709_WideDR_65_FF_Ver.2.0.cube

2020-05-05, 12:08:32
Reply #189

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
James, did you convert your color textures / internal lights to ACEScg before rendering? If not your results are wrong. You need to put lin_srgb as the input (won't change that much except correcting color shifting)
« Last Edit: 2020-05-05, 12:12:55 by Fluss »

2020-05-05, 12:15:32
Reply #190

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
Yes you are correct Fluss. Textures are converted to aces cg space before rendering.



2020-05-05, 12:37:35
Reply #191

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@James Vella your post indicates the BT709_CanonLog2-to-BT709_WideDR_65_FF_Ver.2.0.cube LUT, but when I downloaded the LUT resources from the provided link, there are dozens of LUTS included. Why did you choose the BT709_CanonLog2-to-BT709_WideDR_65_FF_Ver.2.0 version?

2020-05-05, 12:41:50
Reply #192

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
\canon-lut-201911\3dlut\65grid-3dlut\full-to-full-range\BT709_CanonLog2-to-BT709_WideDR_65_FF_Ver.2.0.cube

It doesnt give me the same result in Corona as Vray however.

edit (sorry I missed part of your question):
I chose it because there is only 3 to choose from as I need a log to rec709 df65 to match the setup of the corona scene. (6500kelvin).
- BT709_CanonLog2-to-BT709_WideDR_65_FF_Ver.2.0.cube
- BT709_CanonLog3-to-BT709_WideDR_65_FF_Ver.2.0.cube
- BT709_CanonLog-to-BT709_WideDR_65_FF_Ver.2.0.cube

The first one didnt have a vignette, so used that to keep things as clear as possible without adding additional parameters. Its how I work normally with the VFB, the more you play with the tone mapping controls (highlight burn, contrast etc) the more you complicate things when trying to compare results.






« Last Edit: 2020-05-05, 12:53:50 by James Vella »

2020-05-05, 12:52:47
Reply #193

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Yes you are correct Fluss. Textures are converted to aces cg space before rendering.



And the sun and sky (internal Vray colorspace)?

2020-05-05, 13:00:39
Reply #194

agentdark45

  • Active Users
  • **
  • Posts: 581
    • View Profile

Here is a video of me turning on layers in photoshop :)  https://share.getcloudapp.com/WnuGqzoO

You can download the original image I posted and zoom in to see how much of a terrible job I did! [img width=1024 height=1187]https://i.imgur.com/SjeKjtJ.png/img]

I know it can be hard to be confronted with information you don't like. But I have nothing to gain out of convincing anyone that Corona is better or worst, I'm a corona user! I want corona to be the best possible render engine! Actually, corona render is my main source of income now that I think of it... so NO I don't want it to stop improving!  BUT! I think it can improve more drastically by being easier to use and easier to achieve it's potential, than to extend it's potential even further. LETS DO BOTH, but lets do the easy one first!

Haha fair play, I'll concede that I was indeed duped by this.
Vray who?

2020-05-05, 13:12:46
Reply #195

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
Yes you are correct Fluss. Textures are converted to aces cg space before rendering.



And the sun and sky (internal Vray colorspace)?

Using Raw to AcesCG (but it doesnt make any difference in this example)



2020-05-05, 13:17:31
Reply #196

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
For the sun and sky you do not need OCIO nodes, you need to change the Vray internal colorspace and it's only accessible by script.

Execute this in the maxscript listener :

Code: [Select]
renderers.current.options_rgbColorSpace = 2
If you don't do this, you are rendering ACEScg textures with sRGB sun and sky
« Last Edit: 2020-05-05, 13:21:07 by Fluss »

2020-05-05, 13:18:11
Reply #197

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
For the sun and sky you do not need OCIO nodes, you need to change the Vray internal colorspace and it's only accessible by script.

Execute this in the maxscript listener :

Code: [Select]
renderers.current.options_rgbColorSpace = 2

Thanks! Ill give it a shot

2020-05-05, 13:24:20
Reply #198

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
Great tip thanks - its not true what they say about you Fluss ;) (jk), fixes the blue color cast.



2020-05-05, 13:33:48
Reply #199

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Glad it helped. I'm testing Vray 5 beta and this will be easier!


2020-05-05, 13:44:52
Reply #200

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
Glad it helped. I'm testing Vray 5 beta and this will be easier!




This checkbox "auto RGB primaries etc..."

This does the utility rgb - acescg conversion for vraybitmaps without needing the OCIO node?

2020-05-05, 13:49:12
Reply #201

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Exactly! And it's way quicker than the OCIO conversion. It's based on the filename tho, so you'll still have some work to do.

2020-05-05, 13:52:19
Reply #202

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
Seems like there could be a better way. Using the tag inside the VrayHDRI for sRGB or Linear. Anyway Ill stop derailing this thread with minor topics.


2020-05-05, 13:55:09
Reply #203

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Seems like there could be a better way. Using the tag inside the VrayHDRI for sRGB or Linear. Anyway Ill stop derailing this thread with minor topics.

It's also planned

2020-05-05, 14:33:30
Reply #204

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
Test #2 with Fstorm.

For this section I kept tonemapping OFF, just to see the results of purely increasing the exposure (keep your eye on the exposure setting).








Now we reached the point of blow out. This is where it gets interesting. I turn tonemapping on (defaults)



Now decreasing the burn below 0 (which I expected not good results, but its actually ok). Not sure what to make of it to be honest.



« Last Edit: 2020-05-05, 14:37:48 by James Vella »

2020-05-05, 15:00:57
Reply #205

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
The way I look at it is, color space makes the difference in terms of lighting - its obvious when comparing default Vray with Aces Vray - both have default settings just adjusting exposure value (no tone mapping).

We can then see Fstorm renders without tone mapping, and the light behaves differently to exposure than say default corona/vray. Tone mapping is the last piece of the puzzle here - important yes but not as important as getting a 'photo real' result - might as well use a composite software if you arent happy with the tone mapping controls (not saying it shouldnt be improved but just saying its not going to make a world of difference which you use once you have the correct rendered result).

As others have noted about small details, bumps, materials etc these all help with the bigger picture. My 2c

2020-05-05, 15:06:17
Reply #206

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
I once proposed devs what I consider could work as workaround. To allow Corona to save directly into .dng format. Adobe's DNG has both 16bit and 32bit floating point options. And I thought perhaps this would confuse raw editors less, effectively "trick" them.

This would be great! Honestly, I think if we could treat Corona as a camera (with the .dng format) and use any post production tool we choose, the tone mapping conversation would get much easier to follow and understand.


2020-05-05, 15:13:54
Reply #207

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile

Download the scene from here (created in a student version so none of it can be used for commercial purposes, only learning):https://share.getcloudapp.com/yAu20wZX

It doesn't include the textures? There are only Max files in that zip file.

Edit: I see a problem with this test scene: It uses corona sun and sky. May be we should use an HDRI to make comparisons easier.
« Last Edit: 2020-05-05, 15:17:48 by lupaz »

2020-05-05, 16:29:22
Reply #208

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
Haha fair play, I'll concede that I was indeed duped by this.

Cool that you took it the right way :)

WOW! thanks for downloading the scene and playing around with it. I think there is a lot we can learn from each other.

My objective was to see other people's Corona workflow, so that's why I used Corona tools exclusively, It wasn't meant to compare different engines.

HDRs are loaded at a different starting position and materials would need to be manually converted, it is not a straightforward process.

However, it is very interesting to see different workflows even within Corona! Hope more of you can download and try the scenes, I will later compile all of the images and see what we can learn about it.

2020-05-05, 18:11:36
Reply #209

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
Not sure if anyone is still interested in this. If not, let me know.

Below I rendered a test with gray materials. I removed the glass. No DOF still.
I did Fstorm, Corona with cache and Corona brute force.
Corona with PT never cleared the noise, even after over 300 passes.


It seems that Corona with path tracing is closer to Fstorm in terms of depth of shadows than with cache. Probably expected though.
So may be if we want more depth like with fstorm we need to use path tracing exclusively.

Look at the shadows inside the profile:


With cache, the shadow inside that tube is not as defined as with brute force and fstorm. I believe this is in part why people think tone mapping is the problem. Because we try to get more contrast but there's just not enough detail to begin with.


Corona Cache:




Corona Path Tracing:



Fstorm:


2020-05-05, 18:16:07
Reply #210

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
What's impressive about Fstorm is that outside you can see all the detail. With Corona is completely blown out.

2020-05-05, 18:23:56
Reply #211

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
Not sure if anyone is still interested in this. If not, let me know.

Below I rendered a test with gray materials. I removed the glass. No DOF still.
I did Fstorm, Corona with cache and Corona brute force.
Corona with PT never cleared the noise, even after over 300 passes.



Can you post your Corona settings?

I find that increasing Highlight Compression would allow you to recover details outside, but others are posting different settings, curious about what settings you are using. 

Also, would you mind downloading the scene I uploaded and try it with the same Corona settings you are using for this ?


2020-05-05, 18:27:48
Reply #212

moriah

  • Active Users
  • **
  • Posts: 24
    • View Profile
Does Fstorm use GGX? Seems the BRDF is different compared to Corona, as there's more detail in the shadows.

2020-05-05, 18:36:24
Reply #213

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Thx for the test James, I'm dubitative with this conclusion tho. A wider colorspace only allows for a slightly better GI (corona already benefit from that). The lighting part you mention in Vray is the consequence of the tone mapping curve embedded in the ACES view transform.

Lupaz is right, the scene needs to use the same lighting (free hdri or constant color) and the same assets (I see the widows suddenly disappeared in the Fstorm scene). For those who are wondering, the textures are from the corona library. Also, matching the initial exposure is important to start on a similar basis imho.

I made a quick test in Vray with this HDRi : https://hdrihaven.com/hdri/?h=reichstag_1
camera : iso 1600 / F-number : 4 / shutter speed : 1/50s

Default sRGB :



ACES view trasnform with sRGB primaries (so no texture conversion) :



sRGB + Highlight compression + Handmade curve :





It took me 10s to build the curve by hand to my liking and it's really close to what ACES do and I still have control over the curve and the highlight compression.
« Last Edit: 2020-05-05, 22:38:59 by Fluss »

2020-05-05, 18:46:02
Reply #214

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Thx Lupaz, they are pretty close now. I guess tone mapping is activated Fstorm ?
« Last Edit: 2020-05-06, 14:47:49 by Fluss »

2020-05-05, 18:49:25
Reply #215

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
Thx for the test James, I'm dubitative with this conclusion tho. A wider colorspace only allows for a slightly better GI (corona already benefit from that). The lighting part you mention in Vray is the consequence of the tone mapping curve embedded in the ACES view transform.

Lupaz is right, the scene needs to use the same lighting (free hdri or constant color) and the same assets (I see the widows suddenly disappeared in the Fstorm scene). For those who are wondering, the textures are from the corona library. Also, matching the initial exposure is important to start on a similar basis imho.

Good to know, ill have to have a play with these settings tomorrow. Cheers!

2020-05-05, 18:55:49
Reply #216

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Thx for the test James, I'm dubitative with this conclusion tho. A wider colorspace only allows for a slightly better GI (corona already benefit from that). The lighting part you mention in Vray is the consequence of the tone mapping curve embedded in the ACES view transform.

Lupaz is right, the scene needs to use the same lighting (free hdri or constant color) and the same assets (I see the widows suddenly disappeared in the Fstorm scene). For those who are wondering, the textures are from the corona library. Also, matching the initial exposure is important to start on a similar basis imho.

Good to know, ill have to have a play with these settings tomorrow. Cheers!

I still find ACES a bit better (did not try to match the curve exactly tho) but that was just to show that the differences are not huge.

2020-05-05, 19:09:06
Reply #217

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
Can you post your Corona settings?

I find that increasing Highlight Compression would allow you to recover details outside, but others are posting different settings, curious about what settings you are using. 

Also, would you mind downloading the scene I uploaded and try it with the same Corona settings you are using for this ?

Yes, definitely with highlight compression I'd start seeing

These were the settings before:


And this is with highlight compression. The interior starts to get dull.




2020-05-05, 19:11:28
Reply #218

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Try to make an s-curve like this with corona curves:




2020-05-05, 19:19:20
Reply #219

MattiasD

  • Active Users
  • **
  • Posts: 15
    • View Profile
    • Key Visuals
Maybe that`s why a lot of people are asking for ACES implementation. It`s not the holy grail, but you wouldn`t be doing some custom curves, that just look close enough...
It would be amazing if the devs find time to implement it. I personally think it`s just a matter of time, since it`s getting, much like Dontnod`s PBR back in the day, more an industry standard.
And i`d really like to move forward and start complaining about other stuff :)

2020-05-05, 19:34:07
Reply #220

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile


Yes, definitely with highlight compression I'd start seeing


I always use HC in combination to contrast ( 3- 4 )  In a way, contrast is similar to making a S curve.

2020-05-05, 19:40:28
Reply #221

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile


Yes, definitely with highlight compression I'd start seeing


I always use HC in combination to contrast ( 3- 4 )  In a way, contrast is similar to making a S curve.

Yup. I do the exact same thing. HC at 2 and contrast at 3 is my default.

2020-05-05, 19:58:58
Reply #222

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Yup. I do the exact same thing. HC at 2 and contrast at 3 is my default.
[/quote]


HC very quickly starts to eat the light dynamic in images, so they become sizer dull and dark.
I almost never increase HC over 1.1 / 1.2.
Instead you can crank up Filmic Highlights until you see details in overexposed areas, without sacrificing the overall dynamics of your image.

HC 1
Contrast 2-3
Filmic 0-0.9

Works in 99%




2020-05-05, 20:06:47
Reply #223

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
Yup. I do the exact same thing. HC at 2 and contrast at 3 is my default.


HC very quickly starts to eat the light dynamic in images, so they become sizer dull and dark.
I almost never increase HC over 1.1 / 1.2.
Instead you can crank up Filmic Highlights until you see details in overexposed areas, without sacrificing the overall dynamics of your image.

HC 1
Contrast 2-3
Filmic 0-0.9

Works in 99%
[/quote]

To me, filmic highlights makes everything even duller than HC. I guess it's a matter of taste.

2020-05-05, 20:11:32
Reply #224

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
To me, filmic highlights makes everything even duller than HC. I guess it's a matter of taste.
[/quote]


Oh, that's really interesting.

In C4D / Corona, this behaves exactly the opposite.
I'm still using Corona 4.
Hopefully this HC / Filmic behavior I got used to is not gonna change in Cor. 5...

2020-05-05, 20:16:05
Reply #225

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
Regarding those three test pics ( VRay sRGB / VRay ACES / VRay curves ), in the first moment my eyes went to Vray + curves.
But after short comparison, one realizes that looking at the ACES-pic takes much less effort from your eyes.
This is exactly the effect I notice in renders with lighting / contrasts / colors that feel "right" and "natural".

2020-05-05, 20:44:28
Reply #226

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Yes, gradations are way smoother. I think sRGB EOTF  is fucking with light to shadow transition.
« Last Edit: 2020-05-05, 21:04:52 by Fluss »

2020-05-05, 21:24:31
Reply #227

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
I've been using <1.0 HC, as is described in Dubcat's ACES emulation settings, and it *feels* like it brings back dynamic range, i.e. the image looks more punchy, more depth

2020-05-05, 22:36:36
Reply #228

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
I've been using <1.0 HC, as is described in Dubcat's ACES emulation settings, and it *feels* like it brings back dynamic range, i.e. the image looks more punchy, more depth

Pretty cool! Tried it out right now in one of those sh**ty tiny bathroom scenes.
Indeed it punches up that boring, grayish overlay.

It burns the highlights, of course, but even after increasing filmic, one still has enough contrast.

Why haven't I tried this before? :))))


Is there a link to Dubcat´s ACES emulation you mentioned?

« Last Edit: 2020-05-05, 22:59:13 by Designerman77 »

2020-05-05, 22:53:43
Reply #229

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
I've been using <1.0 HC, as is described in Dubcat's ACES emulation settings, and it *feels* like it brings back dynamic range, i.e. the image looks more punchy, more depth

I never tried those settings. I have to admit they sound completely nonsense for me, on paper at least. Maybe the combination with filmic highlight counterbalances the 0.7 HC compression. I'll give it a go tho, as Dubcat always conducted is tests rigorously and always delivered good stuff. Settings are attached for those wondering.

2020-05-05, 23:11:39
Reply #230

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
I've been using <1.0 HC, as is described in Dubcat's ACES emulation settings, and it *feels* like it brings back dynamic range, i.e. the image looks more punchy, more depth

I never tried those settings. I have to admit they sound completely nonsense for me, on paper at least. Maybe the combination with filmic highlight counterbalances the 0.7 HC compression. I'll give it a go tho, as Dubcat always conducted is tests rigorously and always delivered good stuff. Settings are attached for those wondering.


Thanks Fluss! Had just googled it... sorry guys for stupid asking. :)

Have tried those settings. Filmic highlights and filmic shadows on 1... hmmm... makes the image look a flat, as we all know.
But who knows, maybe that's the proper base for an efficient work in post???

We should check it.

However... HC lower than 1 seems to be the magic stick ! :))))))

Thanks, cjwidd !
Now that's what I call a great discussion leading to great sharing of knowledge.

2020-05-05, 23:15:18
Reply #231

hldemi

  • Active Users
  • **
  • Posts: 55
    • View Profile
I've been using <1.0 HC, as is described in Dubcat's ACES emulation settings, and it *feels* like it brings back dynamic range, i.e. the image looks more punchy, more depth

I never tried those settings. I have to admit they sound completely nonsense for me, on paper at least. Maybe the combination with filmic highlight counterbalances the 0.7 HC compression. I'll give it a go tho, as Dubcat always conducted is tests rigorously and always delivered good stuff. Settings are attached for those wondering.

I have bought this Bertrand Benoit scene ( see attachment ) and many of his cameras have 0.5 HC. This makes them much more realistic.

2020-05-05, 23:20:10
Reply #232

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
I gave it a spin in Photoshop.

I have the impression ( have not double checked I´t yet ) that with Dubcat´s HC 0,6 , even in the shitty 8 bit JPG there was plenty of information in the "lost" highlights,
which could be brought back to life in PS.

Such effect I only know from my Fuji X100 camera which gives out JPGs that behave more like a RAW in post.

2020-05-05, 23:21:51
Reply #233

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
I have bought this Bertrand Benoit scene ( see attachment ) and many of his cameras have 0.5 HC. This makes them much more realistic.
[/quote]


Ahaaaaa ! :)
Now I understand where all that dynamic in many top renders might have come from. At least probably in a certain percentage.

2020-05-05, 23:26:31
Reply #234

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
Yeah, just to clarify about the HC < 1.0 suggestion:

There is no algorithmic approach for flmic highlights, in my opinion - by default I leave them at 0.

Dubcat ACES emulation calls for filmic shadows = 1.0 and filmic highlights = 1.0, but in my experience, this does not lead to a desirable look. Moreover, Dubcat has also suggested not using filmic shadows in combination with his filmic LUTs (i.e. wonky contrast), so filmic highlights and filmic shadows seem to be a sort of "season to taste" option, rather than an option I (personally) lean into, as opposed to contrast, or highlight compression, which I adjust strongly.

If anyone has other tips / suggestions like this, I would be happy to hear about them

2020-05-05, 23:27:24
Reply #235

hldemi

  • Active Users
  • **
  • Posts: 55
    • View Profile
I have bought this Bertrand Benoit scene ( see attachment ) and many of his cameras have 0.5 HC. This makes them much more realistic.


Ahaaaaa ! :)
Now I understand where all that dynamic in many top renders might have come from. At least probably in a certain percentage.
[/quote]

And Bertrand figured it out few years ago. Just compensate with lowering exposure and contrast since going with HC under 1 pushes those parameters up ( not on paper so it might bypass some other bad byproducts of those ). This might be a way to increase contrast without increasing contrast if you get me.

2020-05-05, 23:32:37
Reply #236

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
And Bertrand figured it out few years ago. Just compensate with lowering exposure and contrast since going with HC under 1 pushes those parameters up ( not on paper so it might bypass some other bad byproducts of those ). This might be a way to increase contrast without increasing contrast if you get me.
[/quote]


Yep, how to not understand that? :)
However, I kind of wonder that this is not a standard setting in render engines yet...


I'm working on a job in parallel while reading here and applied those settings to the scene.
Gonna keep that 0.6 / 0.7 HC as standard as often as possible.

2020-05-05, 23:38:10
Reply #237

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
I have bought this Bertrand Benoit scene ( see attachment ) and many of his cameras have 0.5 HC. This makes them much more realistic.

That's a rather dark image though. Not many clients would accept it like that. What if you had to brighten it up?

Does he use filmic highlight?

2020-05-05, 23:38:43
Reply #238

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@hldemi

FYI, you should begin typing your response outside of the [/quote] snippet for proper formatting

2020-05-05, 23:39:11
Reply #239

hldemi

  • Active Users
  • **
  • Posts: 55
    • View Profile

Take a look at this also made in Corona ( you probably seen it already)
https://www.behance.net/gallery/42186689/Apoco77-House

What is similar to Bertrands scene is extreme highlight burn from outside ( by going under 1 HC ) . This however gives this dynamic range inside that is IMO worth the trade. And this burn can be bypassed by corona tonemap control texture map.

2020-05-05, 23:41:09
Reply #240

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
^this scene seems to be really driven by the materials and the modeling, although the tonemapping is obviously driving home a pretty specific look. Great project, of course, pretty sure it was on the Corona Renderer installer splash screen for a while.

2020-05-05, 23:42:16
Reply #241

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile

I have bought this Bertrand Benoit scene ( see attachment ) and many of his cameras have 0.5 HC. This makes them much more realistic.

I think this would be something that I'd think  a client would accept in terms of overall brightness:
(credit: Danthree.com)




2020-05-05, 23:58:33
Reply #242

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile

Take a look at this also made in Corona ( you probably seen it already)
https://www.behance.net/gallery/42186689/Apoco77-House

What is similar to Bertrands scene is extreme highlight burn from outside ( by going under 1 HC ) . This however gives this dynamic range inside that is IMO worth the trade. And this burn can be bypassed by corona tonemap control texture map.



I agree. That's worth the trade!
Such a good base is worth spending some more minutes in post, too in order to get rid of the excessive burning.

2020-05-06, 08:33:04
Reply #243

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
@Fluss, I gave you settings a go and I found at EV0 you are correct, they look similar. Things start to go pear shaped when you hit -5 +3 (for the sRGB/Highlight Compression). Whats your take on this, I miss something?

Vray HDRI Aces





Vray HDRI sRGB + Curve/Highlight Compression




2020-05-06, 11:15:22
Reply #244

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
James, leave the VFB exposure at 0 and increase your render exposure by 3 EVs (so let's say from iso 1600 to iso 12800 in camera settings) and see what happens.

edit:  render exposure was limited by Max Ray intensity.
« Last Edit: 2020-05-07, 09:06:14 by Fluss »

2020-05-06, 11:47:47
Reply #245

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
I made it for you:

Vray + curve + HC - ISO 1600 :



Vray + curve + HC - ISO 12800 :



edit : 3EVs at render time should match 3EVs in VFB, it does not in Vray (Vray 5 beta here)
« Last Edit: 2020-05-06, 23:02:19 by Fluss »

2020-05-06, 11:54:22
Reply #246

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Now here is the interesting part. Made the same test with Corona - photographic exposure - same settings as Vray.

Corona + curve + HC - ISO 1600



Corona + curve + HC - ISO 12800



« Last Edit: 2020-05-06, 12:47:06 by Fluss »

2020-05-06, 12:44:36
Reply #247

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
hmm ok I understand what you mean, but this requires re-rendering. As with the ACES workflow you can adjust the exposure and get the range you need without blowing out areas, similar to how you would work with a photograph in lightroom.


2020-05-06, 12:51:14
Reply #248

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
hmm ok I understand what you mean, but this requires re-rendering. As with the ACES workflow you can adjust the exposure and get the range you need without blowing out areas, similar to how you would work with a photograph in lightroom.

No, this will happen with ACES too. Look at your ACES example, it's also washed out.  Vray and Corona renders are not exposed the same
« Last Edit: 2020-05-06, 22:56:26 by Fluss »

2020-05-06, 13:13:37
Reply #249

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
This is probably a terrible example as I dont do indoor photography (and the snow doesnt help) but this is almost the same expectation of results that I got with the aces test. Exposure stops, -5, -2, 0, +3, +5












2020-05-06, 14:32:52
Reply #250

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Here is what we will consider default exposure for this comparison :

Vray - ACES TM - ISO1600



Now, which of those two renders do you consider correct? Corona is correct

Vray - ACES TM - ISO1600 - +3EV VFB



Vray - ACES TM - ISO12800 (+3EV rendering) - +0EV VFB



I personally made my choice.

edit : after some more thoughts, removed some too early conclusions
edit2 : here is what has been removed. Decided to finally keep it to open the discussion, might be false assumptions :

Now if you look at corona 12800 ISO, it looks like it is not calculating real scene exposure but a base exposure which is then multiplied (like the results we get by increasing exposure in VFB). I'm not really sure what is happening here but that could be an explanation.
-> Vray exposure is limited by max ray intensity


Corona - CURVE + 8HC 12800ISO



« Last Edit: 2020-05-07, 09:47:53 by Fluss »

2020-05-06, 15:48:52
Reply #251

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
hmm, neither do to be honest.

I just took a photo of a room in the house as a raw photo and took it into lightroom:

Exposure: 0


Exposure: 1.5


Exposure: 3.0


The problem Im seeing is the light behaviour is actually quite similar to the fstorm render tests a couple pages back (potentially the default bloom was effecting the result). Where as I see the highlight areas in these last renders as dull in comparison.

I think your last Corona render is probably the best of all 3. I find this quite interesting, ill have to think about this a bit more for awhile.

2020-05-06, 16:23:03
Reply #252

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Of course it's dull, we're using ACES to compress highlights like crazy. Adding some quick crazy bloom and glare help. Still, there is something strange with the corona exposure in my opinion. ->it's good




« Last Edit: 2020-05-06, 22:43:15 by Fluss »

2020-05-06, 22:16:25
Reply #253

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Well, always though Vray and corona photographic exposure settings would match but after quickly looking at it a bit closer, it does not seems to be the case. I'll try to dig a bit further if I find time. The corona one seems to make more sense at first sight.

edit: confirmed, Vray exposure is wrong limited by MRI, corona is good. So pointless comparison from the start between corona and Vray, forget my posts

Guys, checking the technical side is crucial in those kinds of comparisons, so as explaining the process... Everybody was taking a conclusion on renders with different lighting.... That sucks.

If you are looking for me, I'm hiding. 😂
« Last Edit: 2020-05-07, 12:44:19 by Fluss »

2020-05-06, 23:05:20
Reply #254

lupaz

  • Active Users
  • **
  • Posts: 763
    • View Profile
Is it even possible to match every technical aspect between render engines anyway?

2020-05-06, 23:15:07
Reply #255

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
@lupaz definitely not, but emerging color standards are helping to increase parity between different renderers


2020-05-06, 23:16:19
Reply #256

cjwidd

  • Active Users
  • **
  • Posts: 1069
    • View Profile
    • Artstation
Cross-posting a thread about calculating Corona EV from 'I need help' because it's relevant to the current discussion:

EV vs. Corona EV
« Last Edit: 2020-05-06, 23:28:51 by cjwidd »

2020-05-06, 23:25:20
Reply #257

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
For such trivial things as exposure, I expected it to be more or less the same. The thing is that if I raise my ISO from 1600 to 12800 ISO, I should have +3EV.
In corona, a render at  1600ISO, +3EV in post looks more or less the same as a render at 12800 ISO. In Vray, at least in V5 beta, it's not the case.
-> In Vray, I was limited by Max sample intensity.
« Last Edit: 2020-05-07, 09:08:04 by Fluss »

2020-05-07, 07:24:38
Reply #258

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
Is it even possible to match every technical aspect between render engines anyway?

@lupaz I dont think so, nor is it entirely important - what I find important is parity between materials. You should be able to change lighting based on preference and if the materials are setup correctly (PBR for example) then you know you can adjust the lighting/tonemapping to taste without borking something else along the line.

Well, always though Vray and corona photographic exposure settings would match but after quickly looking at it a bit closer, it does not seems to be the case. I'll try to dig a bit further if I find time.

@Fluss I think your first thought was correct, so you are not going crazy :) I found that with some testing this morning 1600 - 12800 is exactly 3 stops. Thanks to @cjwidd and @maru for posting the thread on the vs EV thread I was able to use the excel sheet to do the conversion and it seems correct to me.

Vray 1600 ISO / 0 Exposure



Vray 1600 ISO / +3 Exposure



Vray 12800 ISO / 0 Exposure



Corona 1600 ISO / 0 Exposure



Corona 1600 ISO / +3 Exposure



Corona 12800 ISO / 0 Exposure



Just an FYI if you are trying to follow along in Corona, I had to switch this to a Physical Camera as the Corona Camera was giving me strange results. Not entirely sure why - maybe its just because I always use Physical Camera for vray/corona and used to the setup.

edit:
My conclusion is that I still think the ACES workflow is producing the result that is most similar to photographic workflow (lightroom). Where if you have a well exposed image with no crushed areas then you should be able to push the light either way without compromising your highlights somewhere (actually more so since we are working in 32bit float). Similar to how you can blow out the fstorm render and then pull it back with the burn setting, this is impossible with Corona's highlight compression.
« Last Edit: 2020-05-07, 07:41:44 by James Vella »

2020-05-07, 07:40:38
Reply #259

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
I was using renderer's native camera in both. We're definitely experiencing different results here.

2020-05-07, 08:05:12
Reply #260

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
Yes it took me a few moments to work out how to get parity between them at 1600 / 0 Exposure, thats because there was some limitations with the Physical Camera. My mistake, I used Vray Camera for Vray, Physical Camera for Corona (as I had issues with the Corona Camera overriding my global settings).

This should get you off on the right foot for the comparison.

Vray Settings (Vray Camera):



Corona Settings (Physical Camera):



3D Files (3dsmax 2019):

https://we.tl/t-K37vLOCyk0

For this setup I had to change the fstop (16) shutter speed (2000) for a good base point for adjusting the exposure 3 stops. This way you should have both on parity from this point onwards. Based on the excel document in Corona thats -2.813 for +3 exposure, and -5.813 for 0 exposure. Vray you just adjust the exposure without any need for a conversion.
« Last Edit: 2020-05-07, 08:23:34 by James Vella »

2020-05-07, 08:46:31
Reply #261

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
I downloaded your scene and it worked as expected. But if you try with my camera settings (f4, 1/50s, 1600/12800) and load the HDRi I used for the test (Vraymap, spherical mapping, everything else default), it doesn't work anymore. really strange. Is it the case on your side? (Also, no tone mapping, standard sRGB display)
(Note that with this setup, we're closer to image burn)


« Last Edit: 2020-05-07, 08:55:45 by Fluss »

2020-05-07, 08:56:25
Reply #262

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Found out that max ray intensity was the "culprit" in Vray. Disabling it solved the issue. I edit my previous messages, it's kind of a mess. Maybe it would be better to delete them
« Last Edit: 2020-05-07, 12:49:27 by Fluss »

2020-05-07, 09:32:32
Reply #263

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
Found out that max ray intensity was the culprit in Vray. Disabling it solved the issue.

Yep, that did the trick!

2020-05-07, 09:43:59
Reply #264

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
This makes me think the more "directional look" of Fstorm is due to heavy GI clamping.

2020-05-07, 11:00:23
Reply #265

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
Reinstalled Fstorm,  and damn! I don't like it! I forgot how it feels, well, completely different. In a bad way. It seems the glossiness is indeed still mapped from 0.4 to 1.0.
« Last Edit: 2020-05-07, 12:28:38 by Fluss »

2020-05-07, 12:25:38
Reply #266

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
deleted, reinstalled Fstorm,  and damn! I don't like it! I forgot how it feels, well, completely different. In a bad way. It seems the glossiness is indeed still mapped from 0.4 to 1.0.

Takes a few moments to work some things out. I really like how simple it is, just be aware all bitmaps come in at the correct gamma no need to change anything except HDRI's should be set to 2.2 (kind of opposite but first time user friendly in a way, same goes for render settings in tone mapping leave the gamma settings alone). One thing that I find difficult is there is no camera settings like shutter speed, fstop doesnt affect the lighting like a camera does (that I have found yet, I dont really use it either to be honest just tests every now and then). This will also affect these tests since we cant really bench it against each other. Exposure seems to be a bit arbitrary (or I havent found the recipe for it). But I think this is the charm of it, you dont need to be a technician to use it, you just adjust the exposure/tone mapping and the rest is taken care of without much thought - which is kind of where I see this conversation going, people want an out of the box photo real experience, not having to think about gamma settings or camera settings and get the photo real result. Technical people will probably have a melt down at first but theres probably more to it that Im unaware of, just going by what I read of peoples blog posts etc.

2020-05-07, 12:29:01
Reply #267

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
TBH I'm fed up with those comparisons.. I'm gonna stop there, the more I dig into it, the more I feel it pointless... I always come to the same conclusion, a renderer is a tool, and if used wisely, you can produce beautiful results with any of them.

2020-05-07, 12:56:24
Reply #268

James Vella

  • Active Users
  • **
  • Posts: 202
    • View Profile
lol thats fair enough, thanks for chiming in its been an interesting experiment.

... I always come to the same conclusion, a renderer is a tool, and if used wisely, you can produce beautiful results with any of them.

I completely agree with you.

For the experiments I posted above, I was interested in keeping focus on just one variable - exposure, and disregarding tone mapping entirely for the sake of trying to single out of this variable has any effect on the flexibility of the lighting control from the artists perspective after rendering. From this point onwards introducing one additional variable - highlight compression, and what I expect to be able to control once I have my rendered image with tone mapping.

I once proposed devs what I consider could work as workaround. To allow Corona to save directly into .dng format. Adobe's DNG has both 16bit and 32bit floating point options. And I thought perhaps this would confuse raw editors less, effectively "trick" them.

I think Juraj is on to something here. For people in the archviz field this would be a great option.

For those who work in film, they already work in 32bit space with linear images and thus have a solid pipeline for post production corrections.

Speaking on behalf of myself (and assuming others in this thread), we want that same control in our VFB, we want to render like photographers. Which is why I think fstorm is toted as having excellent tone mapping controls.

edit:
Instead of rendering like a photographer, what I mean to say is render as the human eye perceives things.
« Last Edit: 2020-05-07, 13:19:21 by James Vella »

2020-05-07, 15:13:09
Reply #269

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile
TBH I'm fed up with those comparisons.. I'm gonna stop there, the more I dig into it, the more I feel it pointless... I always come to the same conclusion, a renderer is a tool, and if used wisely, you can produce beautiful results with any of them.

Not with "any of them". Otherwise there would not be different render engines on the market.

Before I chose Corona as an renderer, I compared ( I guess ) hundreds of images from the net, coming from Octane, VRay, Arnold, etc. etc.
One could clearly see differences. Guess which difference was the most obvious to me? Lighting & color grading!

Some render engines create (in average) more artificial images than others. Corona was and is among the top three engines that have beautiful, realistic lighting and especially a pleasant atmosphere in images.

But as with everything in life, everything is great until you see something greater.
And sorry to say... but from FStorm I see images from at user levels - and it is immediately visible that the engine has a more photoreal way to calculate light, contrasts and colors.
We can make 1 Mio. tests here with trivial little chambers and a ball in the background. :))))








2020-05-07, 15:23:01
Reply #270

Designerman77

  • Active Users
  • **
  • Posts: 507
    • View Profile

edit:
Instead of rendering like a photographer, what I mean to say is render as the human eye perceives things.
[/quote]



Yes, bud... that's exactly what I suggested at the beginning of this long discussion.
Make only two modes: 1. behavior like a DSLR.   2. Behavior like a human eye (more less).

When we look at our natural surrounding, we rarely see burnt highlights, black shadows, etc... only in very extreme sunlight situations.
That's because our iris and the brain constantly compensate light dynamic changes in milliseconds.

We are in 2020. It shouldn't be impossible to write two codes that mimic a cam and an eye 99% correctly, instead of making 3D artist fiddle with curves.

2020-05-07, 15:45:20
Reply #271

Fluss

  • Active Users
  • **
  • Posts: 553
    • View Profile
We are in 2020. It shouldn't be impossible to write two codes that mimic a cam and an eye 99% correctly, instead of making 3D artist fiddle with curves.

I give you 2hrs, show me :) Just hope you manage to teleport a display capable of human eye dynamic range from 22th century.

2020-05-07, 16:10:43
Reply #272

lolec

  • Active Users
  • **
  • Posts: 175
    • View Profile
We are in 2020. It shouldn't be impossible to write two codes that mimic a cam and an eye 99% correctly, instead of making 3D artist fiddle with curves.

DSLR already try to approximate what our eyes see. That should be achievable.

But achieving 99% of what the eye sees is literally impossible with current tech and understanding of the world. We don’t “see” in the same sense a camera does, our brain builds an illusion and we make sense of it in our memory.

I’m curios what reference do you hace of renders or images that in your opinion represent what the eye can see?

Again, DSLR try to achieve that, so I’m not sure what is the difference between the two thibgs you are asking

2020-05-07, 17:10:24
Reply #273

dfcorona

  • Active Users
  • **
  • Posts: 240
    • View Profile

edit:
Instead of rendering like a photographer, what I mean to say is render as the human eye perceives things.




Yes, bud... that's exactly what I suggested at the beginning of this long discussion.
Make only two modes: 1. behavior like a DSLR.   2. Behavior like a human eye (more less).

When we look at our natural surrounding, we rarely see burnt highlights, black shadows, etc... only in very extreme sunlight situations.
That's because our iris and the brain constantly compensate light dynamic changes in milliseconds.

We are in 2020. It shouldn't be impossible to write two codes that mimic a cam and an eye 99% correctly, instead of making 3D artist fiddle with curves.
That's hilarious, I suggested such a thing years ago to Vlado with vray.  I even called it Vrayphysicaleye instead of camera.  I know the name is a little goofy, and everyone laughed at my idea.
« Last Edit: 2020-05-07, 17:25:09 by dfcorona »

2020-05-07, 17:13:38
Reply #274

Designerman77

  • Active Users
  • **