Author Topic: New Corona Physical Material (PBR) playground!  (Read 84577 times)

2021-06-19, 10:36:51
Reply #210

maru

  • Corona Team
  • Active Users
  • ****
  • Posts: 12708
  • Marcin
    • View Profile
I am still interested in seeing examples of:
- materials that are using IOR textures
- how materials with IOR textures are exported from various commonly used apps - what format are they exported to, what is their bit depth
As for materials with reflection color or IOR textures, basically anything with varying reflection amount on its surface:

- wood
- any sort of stone and minerals (with the exception of those polished for high gloss maybe)
- concrete, especially when mixed with additional stones or gravel etc
- skin (tiny sweat droplets etc)
- metals - almost any metal that's been touched/used will feature oxidation, impurities or stains etc that will result in varying reflection levels
- manmade surfaces that have seen any usage

The more I think about it, the more I'm convinced that almost everything will display varying reflection level. Absolute perfection will look artificial... an exception would be a water surface, maybe :D

Thanks a lot. These examples are awesome. Some of them really opened my eyes. Just a few ideas, of course without discrediting your list:
- some of these effects could be achieved by using Layered Mtl with masks (mixing two or more different Physical Mtls)
- metal with oxidation - could be done with a metalness texture rather than IOR? (maybe?)
- some things (maybe skin with sweat) could be done with uniform IOR and a roughness map rather than IOR map

But even with that said, yes, I see the point.

The question now is:
If you were to create such materials with varying IOR - how would you do it? Procedural textures in 3ds Max? Some textures generated with 3rd party apps? If they were bitmaps, what file format and bit depth would you use?
What would be the simplest, most ideal solution for you?
Marcin Miodek | chaos-corona.com
3D Support Team Lead - Corona | contact us

2021-06-23, 15:32:14
Reply #211

MarkusB.

  • Active Users
  • **
  • Posts: 5
    • View Profile
Hi

i am wondering, why isnt the corona converter converting the 3dsmax Physical Material to a Corona Physical Material ? These materials get skipped.

Also is it possible to convert a Corona Physical Material back to a legacy one for backwards compatible reasong? That would be really good. I know you would loose some feature...

2021-06-23, 16:10:58
Reply #212

Mohammadreza Mohseni

  • Active Users
  • **
  • Posts: 152
    • View Profile
    • Instagram

2021-06-23, 17:30:06
Reply #213

GeorgeK

  • Corona Team
  • Active Users
  • ****
  • Posts: 838
  • George
    • View Profile

2021-06-23, 17:35:04
Reply #214

lupaz

  • Active Users
  • **
  • Posts: 951
    • View Profile
one of my tests on new physical material. awesome sheen layer.

https://forum.corona-renderer.com/index.php?topic=33608.msg186807#msg186807

Really look forward to using the sheen layer. Looks awesome.
If I was the corona team I'd have split the "new features" and added sheen layer as a new feature instead of putting all together as PBR material.

2021-06-24, 10:42:28
Reply #215

rowmanns

  • Corona Team
  • Active Users
  • ****
  • Posts: 1892
  • Corona for 3ds Max QA Team
    • View Profile
Hi

i am wondering, why isnt the corona converter converting the 3dsmax Physical Material to a Corona Physical Material ? These materials get skipped.

Also is it possible to convert a Corona Physical Material back to a legacy one for backwards compatible reasong? That would be really good. I know you would loose some feature...
Hi,

3ds Max physical material isn't supported by the converter yet, we are planning on adding support in the future.

We also don't support converting from CoronaPhysicalMaterial to LegacyMtl and it is not planned..

I hope this info helps.

Thanks,

Rowan
Please read this before reporting bugs: How to report issues to us!
Send me your scene!

2021-06-29, 02:29:22
Reply #216

n2graf

  • Active Users
  • **
  • Posts: 55
    • View Profile
Hello, how have you decided the wave values ​​for RGB that are put in the complex IOR that appear here https://coronarenderer.freshdesk.com/support/solutions/articles/12000079075-how-to-use-complex-ior-for -coronaphysicalmtl-? Shouldn't they be equivalent to the ends of the sRGB triangle where Red corresponds to 0.65nm for example? (see link of the image)

https://upload.wikimedia.org/wikipedia/commons/thumb/8/8f/CIExy1931_sRGB.svg/1200px-CIExy1931_sRGB.svg.png

Thanks!

2021-06-29, 02:31:31
Reply #217

n2graf

  • Active Users
  • **
  • Posts: 55
    • View Profile
I know that you can pute the values to www.refractiveindex.info but the question is: how do you decided that red belongs to 0,7. Thanks

2021-06-29, 10:36:32
Reply #218

GeorgeK

  • Corona Team
  • Active Users
  • ****
  • Posts: 838
  • George
    • View Profile
Hello, how have you decided the wave values ​​for RGB that are put in the complex IOR that appear here https://coronarenderer.freshdesk.com/support/solutions/articles/12000079075-how-to-use-complex-ior-for -coronaphysicalmtl-? Shouldn't they be equivalent to the ends of the sRGB triangle where Red corresponds to 0.65nm for example? (see link of the image)

https://upload.wikimedia.org/wikipedia/commons/thumb/8/8f/CIExy1931_sRGB.svg/1200px-CIExy1931_sRGB.svg.png

Thanks!

Hi n2graf, please note that Corona is using color space with a wider gamut than sRGB, you will also find that approximations around 0.7-0.55-0.45 can produce a more favoured outcome depending on how much oxidation is introduced on the surface, and the type of metal.

George Karampelas | chaos-corona.com
Chaos Corona QA Specialist | contact us

2021-06-29, 12:42:01
Reply #219

n2graf

  • Active Users
  • **
  • Posts: 55
    • View Profile
In your freshdesck website you have considered:

0.70 μm for Red, 0,55 μm for Green, 0.45 μm for Blue

But for wideRGB gamut like that:

https://upload.wikimedia.org/wikipedia/commons/1/1d/CIExy1931_AdobeWGRGB.png

Green belongs to 0,52 and blue 0,390. why is there this discrepancy?

Thanks

2021-06-29, 15:08:52
Reply #220

romullus

  • Global Moderator
  • Active Users
  • ****
  • Posts: 8778
  • Let's move this topic, shall we?
    • View Profile
    • My Models
Here's 3 different tables of visible light spectrum that i quickly found on google (there are gazillion of them). Every single of them has slightly different range of wavelength for each colour. As you can guess, there's no single correct value at which we can call the wavelength as "blue", or "red". Values that are chosen in helpdesk by Corona team, lies within widely accepted range. I think there's no need to overthink this.
I'm not Corona Team member. Everything i say, is my personal opinion only.
My Models | My Videos | My Pictures

2021-07-01, 12:19:05
Reply #221

GeorgeK

  • Corona Team
  • Active Users
  • ****
  • Posts: 838
  • George
    • View Profile
In your freshdesck website you have considered:

0.70 μm for Red, 0,55 μm for Green, 0.45 μm for Blue

But for wideRGB gamut like that:

https://upload.wikimedia.org/wikipedia/commons/1/1d/CIExy1931_AdobeWGRGB.png

Green belongs to 0,52 and blue 0,390. why is there this discrepancy?

Thanks

I don't believe input values can be absolute, most approximations around the limits of the visible light spectrum will work just fine depending on, as said previously, "preferred visual outcome and on how much oxidation is introduced on the metal's surface" (excluding metal alloys - impure metals - although oxidization implies impurity, but let's just leave it at that).

Let's take the example of Pure Gold (Au) -  Generally, its electrons move at relativistic speeds (especially on its outer surface). The outer surface is responsible for its chemical behaviour and a lot of physical properties, including color. The human eye spectrum varies from wavelengths - approximately 390 nm (blue) to 700 nm (red), gold absorbs a lot of the low wavelengths (blue). So we’re left with the opposite (yellow - towards red when it's pure).

Pure Gold (au) does not get oxidised, hence pure gold should not appear green-ish, it should mostly reflect yellow with a minuscule tint of red. The following comparison showcases exactly just that, Complex IOR of Gold with 0.7-0.55-0.45 against sRGB limited values of 0.61-0.56-0.47: https://corona-renderer.com/comparer/RRnUev (Both outcomes can be described as Gold, but when thinking of recreating Pure Gold we should compensate for corresponding wavelength inputs, in order to remove the greenish tint)

So, to summarize, any approximations between 0.70 μm for Red, 0,55 μm for Green, 0.45 μm for Blue (~ +- within reason) will work great for most pure metals, in some cases, you might want to experiment with different wavelengths. I hope this helps, and it actually makes sense!
« Last Edit: 2021-07-01, 13:18:20 by GeorgeK »
George Karampelas | chaos-corona.com
Chaos Corona QA Specialist | contact us

2021-07-08, 16:57:10
Reply #222

Duron

  • Active Users
  • **
  • Posts: 106
    • View Profile
    • Portfolio
Today I noticed for the first time a difference in reflections when I changed the roughness IOR from 0.00 to 0.01.

I am not sure if this is already known. So sorry if this is a double post.

Simple reflective plastic material is used with no clear coat.
Everything looks fine at IOR 0.00. Once it goes above 0.01, the reflections change. It reminds of switching the reflection mode in UE4 from nonRT to RT.
Normally this wouldn't bother me at all, but I very often use a value of 0.00 to get softer highlights like from the sun. So almost all my materials are at 0.01. But this example shows why I had problems when creating my physical materials.

Please focus on the BMW badge.

UPDATE:
Did more tests and I can narrow down the cause a bit. The material does reflect an HDR in Dome mode. When the mode is set to Spherical everything is fine.

In Dome mode it seems that the lower hemisphere is displayed as black for materials with a Roughness of >0.01

UPDATE 2:
Updated with new images showing the issue with dome mode.
« Last Edit: 2021-07-08, 17:15:37 by Duron »

2021-07-08, 19:05:40
Reply #223

pokoy

  • Active Users
  • **
  • Posts: 1850
    • View Profile
Reflection glossiness/roughness switches to a different algorithm when used - this was the case with legacy, too. Check and you'll probably see the same thing.

Dome mode - yes, long standing issue, known and reported long ago. The workaround is to add a plane under the model - it can be switched off from camera and reflection/refraction visibility - this should 'fix' the blocky artifacts.

2021-07-09, 10:12:51
Reply #224

GeorgeK

  • Corona Team
  • Active Users
  • ****
  • Posts: 838
  • George
    • View Profile
Today I noticed for the first time a difference in reflections when I changed the roughness IOR from 0.00 to 0.01.

I am not sure if this is already known. So sorry if this is a double post.

Simple reflective plastic material is used with no clear coat.
Everything looks fine at IOR 0.00. Once it goes above 0.01, the reflections change. It reminds of switching the reflection mode in UE4 from nonRT to RT.
Normally this wouldn't bother me at all, but I very often use a value of 0.00 to get softer highlights like from the sun. So almost all my materials are at 0.01. But this example shows why I had problems when creating my physical materials.

Please focus on the BMW badge.

UPDATE:
Did more tests and I can narrow down the cause a bit. The material does reflect an HDR in Dome mode. When the mode is set to Spherical everything is fine.

In Dome mode it seems that the lower hemisphere is displayed as black for materials with a Roughness of >0.01

UPDATE 2:
Updated with new images showing the issue with dome mode.

This was a known issue and I believe it's already addressed, unless I am missing something. Which version of Corona DB or RC you are using in these examples?

(Report ID=CRMAX-436)
Dome - (Internal id=311287120)
« Last Edit: 2021-07-09, 10:24:58 by GeorgeK »
George Karampelas | chaos-corona.com
Chaos Corona QA Specialist | contact us