Author Topic: Tonemapping - Plz Halp  (Read 115807 times)

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

romullus

  • Global Moderator
  • Active Users
  • ****
  • Posts: 8779
  • 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: 8779
  • 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: 1644
    • 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: 1077
    • 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: 8779
  • 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: 1644
    • 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: 951
    • 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: 4743
    • 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 »
Please follow my new Instagram for latest projects, tips&tricks, short video tutorials and free models
Behance  Probably best updated portfolio of my work
lysfaere.com Please check the new stuff!

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

cjwidd

  • Active Users
  • **
  • Posts: 1077
    • 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: 1077
    • 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: 1644
    • 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: 4743
    • View Profile
    • studio website
A Trump meme already 😀Boy this escalated quickly.
Please follow my new Instagram for latest projects, tips&tricks, short video tutorials and free models
Behance  Probably best updated portfolio of my work
lysfaere.com Please check the new stuff!