Author Topic: New Corona Physical Mat (separating reflections from refractions)  (Read 6492 times)

2021-03-17, 20:04:13

SharpEars

  • Active Users
  • **
  • Posts: 103
    • View Profile
There appears to be no way, that I have been able to figure out, to adjust the brightness and blur of reflections of a transparent surface, without changing the transparency as well. I am not talking about breaking “Energy Conservation” here, quite the opposite actually.

I would just like the ability to _decrease_ the brightness and increase the roughness of reflections without completely screwing up the refraction/absorption settings in the process.

With the “old” material, you could adjust the reflection settings independently (both color and glossiness), making it as rough and dim as you want dependent on the lighting and other scene objects.

With the new physically based material, when creating a fully transparent material (or even mostly transparent with little roughness) with maybe slight amount of absorption, reflections, especially of bright lights, tend to be super bright and razor sharp, dominating the material’s appearance and not giving the refractive properties a chance to show.

Perhaps what I am asking for is “bias” and the new material is supposed to move in the unbiased direction, but consider a good camera lens. It is coated with an anti-glare material that actually makes more light go through the lens, while reducing reflections from the lens elements’ surfaces. Clearly this is possible in the physical world! With Corona’s new material, sure there is a coating feature, but it sits on top of the horrid unchangeable reflections already present in a transparent material and can only add to “obscure” things further.

I haven’t found any way to tone reflections down (as in less bright via say a gray color instead of pure white) or blur them, while maintaining a good clean sharp transparency for a material.

Any tips short of masking/compositing?

BTW, node material doesn’t help here, because reflections get pulled in from the Directly Visible slot and do not get overridden by the reflection settings of the material connected to the Reflection socket. I don’t know if this is a bug or not, but in any case the material itself should allow for reflection adjustments (in terms of both color/brightness and roughness/glossiness) for refractive materials.
« Last Edit: 2021-03-17, 20:08:17 by SharpEars »

2021-03-17, 22:25:25
Reply #1

SharpEars

  • Active Users
  • **
  • Posts: 103
    • View Profile
I have created a sample scene (latest daily build of Corona from Feb. 2021 was used) demonstrating the issue (attached). It may be obvious to most, but the spheres in the image are hollow (with proper normal orientations).
« Last Edit: 2021-03-17, 22:33:15 by SharpEars »

2021-03-18, 17:15:50
Reply #2

SharpEars

  • Active Users
  • **
  • Posts: 103
    • View Profile
The situation is much worse than I thought. There seems to be no way with the new physical material to create a completely opaque black diffuse (non-metallic) material that has a real world IOR (e.g., 1.4-1.6) value.

Try it!

Forget about refraction, turn it off completely. Just create a basic Physical non-metal material with Base layer color set to black and a real world IOR of 1.4-1.6. You don't want it to be reflective, so set Glossiness to 0.0% (or Roughness to 100.0%). This should result in a simple black diffuse material, but it is gray and there is no way to get rid of that gray without using a fake low unrealistic IOR value closer to 1.0.

That's just wrong in the real world!

On top of that, what about glossiness? Turn it up to 100% and you get a harsh white reflection from a completely 100% black colored object! What the??? I am not talking Fresnel effects at glancing angles, here - that would be fine. I am talking direct front facing bright white reflections from an object that is solid 100% black in color.

All of this stems from the fact that there is no way to set the color, glossiness/roughness, and IOR values of the reflection itself in a physical material and to do so completely independent of (i.e., without affecting) any other properties of the material. Of course, energy conservation should continue to be in effect - that's expected.

Don't get me wrong. I love all of the features added by the new physical material. But, without having direct independent control over the color, intensity (i.e., brightness), and blurriness of its reflection property, it is pretty much useless as an accurate representation of a vast amount of real word materials that do not subscribe to the "Reflection reigns supreme over all other material properties!" philosophy.


« Last Edit: 2021-03-18, 17:42:06 by SharpEars »

2021-03-19, 11:04:24
Reply #3

burnin

  • Active Users
  • **
  • Posts: 1532
    • View Profile
Lower IOR.



& Learn about Color Science ("Pixar in a box" from Khan Academy)

PS.
ahh the old forum code...
« Last Edit: 2021-03-19, 11:17:10 by burnin »

2021-03-19, 16:43:30
Reply #4

SharpEars

  • Active Users
  • **
  • Posts: 103
    • View Profile
You don't even need roughness if you fudge the IOR like that. That's my point - not physical reality, at all.
 
Do you know what dialectrics (i.e., non-metals) in this world have an IOR value close to 1.13? Liquid methane, with an IOR of 1.15! Air is 1.0. Water is 1.33 and most dielectric solids have IORs that fall squarely between 1.4 and 1.6.

The black Shader Ball in your scene, as great as it looks in a biased sort of way, it's made out of liquid methane, right? Because it looks more like black rubber to me (thanks to the roughness applied), which by the way has a real world IOR of 1.52 or approx. the value used for the gray Shader Ball on the right, proving my point, exactly (i.e., blacks appear gray under reasonable lighting conditions).

If the new Physical Material is supposed to be closer to the real world, it should behave like it. Your example of the black Shader Ball solved the "blackness" problem for rough materials, but it's a workaround (and it is a workaround, mind you) that can only be applied to opaque materials.

There is still no solution to the transparency/translucency issue I described in my original post which suffers from a similar IOR issue, even with a hack and even with layered materials. But, I would love if someone can find a workaround for that (i.e., getting transparency to be independent of reflection or at least figuring out a way to tone down and blur those reflections without obliterating clear transparency), because the new material does have some great features that unfortunately are blocked by this reflective "elephant in the room!" And, I am talking a material based workaround that doesn't resort to compositing/post-processing, render masks and render elements, or other non-material based solutions.

« Last Edit: 2021-03-19, 16:52:02 by SharpEars »

2021-03-19, 17:58:30
Reply #5

burnin

  • Active Users
  • **
  • Posts: 1532
    • View Profile
serve your imagination w/ Vertically Aligned NanoTube Arrays
and contemplate on what is "black", if not lack of...

2021-03-24, 16:11:20
Reply #6

SharpEars

  • Active Users
  • **
  • Posts: 103
    • View Profile
And on this very topic:

Note: While you can use Clearcoat to increase reflection intensity over a lower IOR Base material as above, you cannot reduce the reflection intensity of a high IOR Base material. This is an example of a constraint introduced by physical realism.

Quoted from: https://blog.corona-renderer.com/behind-the-scenes-the-physical-material/

Well, I beg to differ (see attachment 1). Plenty of blurriness (and darkness) in the reflections, in a completely transparent (i.e., nearly 0% roughness/100% glossiness) material with a reasonably high IOR of at least 1.53 or higher.

And with regard to coatings, see attachment 2, or did you guys completely forget that there are anti-reflective coatings that can reduce, dull, and convert (to transparency) the original material's reflective properties?? Which by the way would necessitate separate color, glossiness/roughness, and IOR controls for reflection (at least at glancing angles, due to Fresnel affect), while making the reflection of the original uncoated material get converted into refraction (well transmission, to be physically correct) when the object is viewed head on (i.e., directly along its normal vector).

« Last Edit: 2021-03-24, 16:26:50 by SharpEars »

2021-03-24, 16:41:50
Reply #7

romullus

  • Global Moderator
  • Active Users
  • ****
  • Posts: 8779
  • Let's move this topic, shall we?
    • View Profile
    • My Models
Well, I beg to differ (see attachment 1). Plenty of blurriness (and darkness) in the reflections, in a completely transparent (i.e., nearly 0% roughness/100% glossiness) material with a reasonably high IOR of at least 1.53 or higher.

I see no blurriness at all, the reflections are perfectly sharp. Definitely you should be able to recreate this with new physical material without resorting to unphysical hacks.
I'm not Corona Team member. Everything i say, is my personal opinion only.
My Models | My Videos | My Pictures

2021-03-24, 19:13:54
Reply #8

pokoy

  • Active Users
  • **
  • Posts: 1850
    • View Profile
I kinda understand why you're saying this but your examples are really not showing anything.
Studio glass photography is a hack from start to end, always. Plus some (or a lot of) retouching. Also, your example shows nicely what studio lighting is about - creating soft gradients and sharp contrasts with creative lighting. The material has no blurriness here, it's a matter of how the studio is set up.
I'm pretty sure the second example is all about retouching. The original photo probably didn't have any tinting in it, it's all post production.

With that said - have you tried to recreate the examples you are talking about with PBR mats within a Layer material? I think that in some cases it'll be the only way to achieve a few effects like tinted reflections etc with the new PBR material.

I've played some time with it and yes, some things are harder to produce and may require the use of Layer material. Working with reflection maps got harder now since we're mapping IOR and the current behavior is not the most convenient. On the other hand, the material produces really great results.

I don't want to be in the developers' position right now - creating a material that allows for the same functionality while not breaking physical rules is probably impossible, and to make the UI nice, logical and convenient at the same time... that's really hard, and maybe impossible without sacrifices in other places.

2021-03-24, 20:47:13
Reply #9

davemahi

  • Active Users
  • **
  • Posts: 192
    • View Profile
    • iamstatic
I just don't get this "dont break the physical rules"

It makes no sense for a rendering pipeline. You will always have to bend things on the shader to make it look real.
And to others points, photos and "real" world images get touched up so much. I think it will be a handcuff for Corona render personally.
« Last Edit: 2021-03-26, 15:48:57 by davemahi »

2021-03-25, 20:41:21
Reply #10

SharpEars

  • Active Users
  • **
  • Posts: 103
    • View Profile
I kinda understand why you're saying this but your examples are really not showing anything.
Studio glass photography is a hack from start to end, always. Plus some (or a lot of) retouching. Also, your example shows nicely what studio lighting is about - creating soft gradients and sharp contrasts with creative lighting. The material has no blurriness here, it's a matter of how the studio is set up.
I'm pretty sure the second example is all about retouching. The original photo probably didn't have any tinting in it, it's all post production.

With that said - have you tried to recreate the examples you are talking about with PBR mats within a Layer material? I think that in some cases it'll be the only way to achieve a few effects like tinted reflections etc with the new PBR material.

I've played some time with it and yes, some things are harder to produce and may require the use of Layer material. Working with reflection maps got harder now since we're mapping IOR and the current behavior is not the most convenient. On the other hand, the material produces really great results.

I don't want to be in the developers' position right now - creating a material that allows for the same functionality while not breaking physical rules is probably impossible, and to make the UI nice, logical and convenient at the same time... that's really hard, and maybe impossible without sacrifices in other places.

I hear what you are saying. Unfortunately, I do not see any way to use Layer material to “lessen, blur, or re-color” the harsh reflections of an almost perfectly smooth refractive material. I need something that basically subtracts from the material (at least for the case of re-color and dim/lessen, when it comes to reflection), not adds on top of (re: layer material), for which clear coating can suffice. If I diminish a refractive material in any way via the layered material, I will affect its refractive properties as well as its refractive (which I want to remain at full brightness and unaffected).

If you know of a way to darken, blur, or re-color _just_ the reflections of a material without resorting to masking/compositing, feel free to describe said method in greater detail. I haven’t been able to find any way to do that using materials (including Layer Material) and their properties, alone. If such a workaround already exists, and I just haven’t found it, then there will be nothing to complain about with regard to highly glossy transparent materials. I love the new physical material for most textures, it’s just this shortcoming for highly transparent materials that makes it unusable for some common real world glass/crystal/acrylic highly transparent objects, when hard lighting is present and visible in their reflections (and cannot be toned down/blurred without touching the lighting itself, which of course cascades to other properties of the material and other objects and their materials, that are illuminated directly/indirectly by the same light).

« Last Edit: 2021-03-25, 20:49:43 by SharpEars »

2021-03-26, 15:39:00
Reply #11

davetwo

  • Active Users
  • **
  • Posts: 298
    • View Profile
I must admit- i'm a little concerned by the new direction. Half the joy of doing product shots in 3d rather than photographing them is that you can cheat in-render rather than having to retouch  afterwards.

Good to have the physically correct values as a base. But they shouldn't be a constraint. Everything we do is creatively art directed.

2021-03-26, 17:28:59
Reply #12

Cinemike

  • Active Users
  • **
  • Posts: 1000
    • View Profile
I must admit- i'm a little concerned by the new direction. Half the joy of doing product shots in 3d rather than photographing them is that you can cheat in-render rather than having to retouch  afterwards.

Good to have the physically correct values as a base. But they shouldn't be a constraint. Everything we do is creatively art directed.

I was told the legacy material is here to stay, so ...

2021-03-29, 09:13:49
Reply #13

mmarcotic

  • Former Corona Team Member
  • Active Users
  • **
  • Posts: 544
  • Jan - C4D QA
    • View Profile
Yes, as Mike says.

Corona Legacy material is here to stay for now, if you prefer Legacy in some scenarios and Physical in other, feel free to do so!

Thanks for the feedback, we are closely monitoring it for future tweaks and updates.
Jan
« Last Edit: 2021-03-29, 17:40:29 by mmarcotic »
Learn how to report bugs for Corona in C4D here.

2021-03-29, 16:19:51
Reply #14

SharpEars

  • Active Users
  • **
  • Posts: 103
    • View Profile
Yes, as Mike says.

Corona Legacy material is here to stay, if you prefer Legacy in some scenarios and Physical in other, feel free to do so!

Thanks for the feedback, we are closely monitoring it for future tweaks and updates.
Jan

So for the record, does the Corona team _not_ want to allow for control of reflection {color,glossiness/roughness,IOR} independently of that of refractions in the Physical Material?

2021-03-29, 17:09:08
Reply #15

pokoy

  • Active Users
  • **
  • Posts: 1850
    • View Profile
I think there's an intent to keep the physical material definition pure - only a rough surface will lead to rough (more diffuse) transmission. You can't have a non-glossy 'internal matter' without a rough non-glossy surface per definition, and vice versa, you can't have a clear perfectly transmissive 'interior' with a rough surface. In other words, the surface finish defines what happens internally, it is the terminal that sends ray through the object. The ray enters and travels through the object depending on its surface properties.
That's why this definition needs a coating layer, which they provide. But it has to be an individual layer of a coating material and must not interfere with the aforementioned principle.

If I understand correctly, that is...

The current shading model doesn't account for other effects, such as varying dispersion, varying IOR within a medium etc, too, but we don't care because it's really a tiny fraction of the materials that Corona is used for in a typical situation.

Please continue to provide examples for something that you can't do with the physical material. I'm sure that it's going to help if it's a valid request.

2021-03-29, 17:43:19
Reply #16

mmarcotic

  • Former Corona Team Member
  • Active Users
  • **
  • Posts: 544
  • Jan - C4D QA
    • View Profile
Yes, as Mike says.

Corona Legacy material is here to stay, if you prefer Legacy in some scenarios and Physical in other, feel free to do so!

Thanks for the feedback, we are closely monitoring it for future tweaks and updates.
Jan

So for the record, does the Corona team _not_ want to allow for control of reflection {color,glossiness/roughness,IOR} independently of that of refractions in the Physical Material?

It was mentioned in another topic as well - we are working on improvements to Physical material (I'm not confirming that we'll add the ones you've mentioned) and while we try to keep it as realistic as it gets, we might make some adjustments and additions in time.

Please note that development of such scale takes time and we do not want to neglect other features, bugs, etc.

Thanks,
Jan
Learn how to report bugs for Corona in C4D here.

2021-03-29, 18:50:18
Reply #17

Cinemike

  • Active Users
  • **
  • Posts: 1000
    • View Profile
What I am missing is to subtly cheat with a specular map to dim highlights.
Besides that, I am content with the physical material in the majority of all cases.