Isotropix Forums

standartshader - adopting the Arnold 5 shader layout

So you'd like something that's not there? Just tell us what, we'll see what we can do.

standartshader - adopting the Arnold 5 shader layout

Unread postby bjoern » Sat Aug 12, 2017 10:54 am

I love the fact that I can use Clarrise a bit in tandem with Maya and arnold. In the sense that the shaders are the same. Or have been the same.
I used extensively Alshaders and the new standard surface of Arnold 5. And I find the new shader in A5 allot better. It would be super great if Clarrise
would mimic these parameters (with essentially the same tech from Alshaders under the hood). For a few reasons.
- The Metalness parameter is very convenient and Artist friendly (my favorite part)
- Other apps like substance or mari etc... will add the A5 shader to there list. And if Clarrise is mimicking that shader, mixed pipeline (Maya+A5+Clarrise) are easier to handle :)
bjoern
 
Posts: 128
Joined: Thu May 11, 2017 1:43 pm

Re: standartshader - adopting the Arnold 5 shader layout

Unread postby isoyann » Thu Aug 17, 2017 1:21 pm

Hi,

The purpose of the metalness concept is to get a very good compromise between simplicity and power: with a very limited number of attributes, you can approximate almost all materials.

Alshader aim a different goal: because the metalness approach, while very effective in real time and animated features (with stylized look) is often a bit limited for vfx work, it was designed using a layered approach, more flexible, more complex, but still intuitive for artists.

Mixing the two in the same shader add lot of confusion and complexity.. and pollute what was a very good multipurpose uber shader. The disney approach (pbr with metalness) is very good, but to get the most of this, it is best to have it as a separate shader.. and if you want more options, go to the original alshader layered approach.

Another issue is it is not opensource anymore.. while Alshader was a good move toward standardization, it is now a black box.
User avatar
isoyann
 
Posts: 602
Joined: Mon Jan 28, 2013 12:11 pm

Re: standartshader - adopting the Arnold 5 shader layout

Unread postby bjoern » Fri Sep 15, 2017 1:13 pm

I don't think that the Metallness parameter is limiting. Also Arnold is mostly targeting photo-realism. Houdini as well adopted 1to1 the PixarDisney shader as there main ubershader. I can't see how the argument in regards to "real time and animated features" apply's here.
bjoern
 
Posts: 128
Joined: Thu May 11, 2017 1:43 pm

Re: standartshader - adopting the Arnold 5 shader layout

Unread postby jensonite » Sun Sep 24, 2017 12:40 am

I couldn't agree more. I doubt Solid Angle would implement a shader as the new standard if it wasn't capable of producing photo-real results. It is in fact heavily inspired by the alSurface shader except that it actually streamlines some of the workflow a bit more.
Also, if standardization is the issue, it doesn't get more standardized than PBR metal / roughness imho.
jensonite
 
Posts: 7
Joined: Tue Feb 14, 2017 8:45 pm

Re: standartshader - adopting the Arnold 5 shader layout

Unread postby bjoern » Mon Nov 06, 2017 11:26 am

I had to go recently allot back and force between Redshift, Mantra, Arnold and Clarisse. And while doing so I discussed their shaders
with fellow VFXartist and educators. For most the Metalness workflow is the preferred and future method. Because it is the easiest way to achieve Photo-real shaders while not breaking physics. If one wants to look up the why, please have a read through this excellent written guide https://www.allegorithmic.com/pbr-guide

Since Clarisse is a Pipeline Back-end tool and allot mixed with other renderers within a pipeline. It would make clarisse even more pipeline friendly if it can mimic some of the most workflow relevant features of their shaders.
Thats why I would like to propose now a feature more similar to redshift's "Fresnel Type", which gives you simply the option to choose your preferred workflow: (https://docs.redshift3d.com/display/RSD ... Reflection)

From the Redshift docs
Fresnel Type
This allows you to select between different methods of controlling the reflection Fresnel effect;
>>"IOR (Advanced)" lets you set the Index of Refraction for each red, green and blue channel. This allows you to control the reflectivity color for both dielectric materials (i.e. non-metals) and metals, while remaining physically correct. This mode also lets you set the 'Absorption/k' values required for metals. 'Absorption/k' should be 0,0,0 for non-metals.
>>"Color+Edge Tint" is an artist friendly mode that allows you to control the facing reflectivity color and metal edge tint, without required complex IOR values. The 'Metal Edge Tint' should be black for non-metals.
>>"Metalness" is a special mode designed for 'PBR' workflow compatibility with 'Substance' and 'Unreal' materials.
>>"IOR" lets you set a single Index of Refraction value to control reflectivity and is ideal for simple dielectric materials. Dielectric materials typically have an Index of Refraction between 1.0 and 2.0, where 1.0 means no reflectivity and higher values mean stronger reflectivity.


What do you guys think?
bjoern
 
Posts: 128
Joined: Thu May 11, 2017 1:43 pm

Re: standartshader - adopting the Arnold 5 shader layout

Unread postby isoyann » Tue Nov 07, 2017 12:12 pm

I still think it is a bad idea to mix every possible approaches in one shader.. If one day we implement the Disney shader, let's implement it the right way (like they did in mantra, renderman and lot of other renderers), not mixed with something totally different.
I've added a feature request (#7676: New material based on Disney principled Shader), waiting for approval.

Thanks
User avatar
isoyann
 
Posts: 602
Joined: Mon Jan 28, 2013 12:11 pm

Re: standartshader - adopting the Arnold 5 shader layout

Unread postby bjoern » Tue Nov 07, 2017 2:33 pm

thank you isoyann,
Alshaders "was" until recently the "state of the art" shader, but it is not anymore. Almost all renders go in the direction of the PBR/metalness aproach
And I brought the example of redshift up simply to be able to accommodate both workflows. But a move towards the PixarDisney shader would be definitely amazing! The Mantra implementation of Pix/Dis is especially great.
bjoern
 
Posts: 128
Joined: Thu May 11, 2017 1:43 pm

Re: standartshader - adopting the Arnold 5 shader layout

Unread postby ccidrais » Thu Dec 07, 2017 8:52 pm

I'd like to add my vote to the Disney Shader being ported to Clarisse.
Incidentally...is there a place on the isotropix website where we can see all the feature requests and their statuses?
I tried searching for #7676: New material based on Disney principled Shader in the bug tracker but couldn't find anything so I presume it's not there.
Thanks!
ccidrais
 
Posts: 2
Joined: Thu Nov 16, 2017 8:06 pm

Re: standartshader - adopting the Arnold 5 shader layout

Unread postby gjennings » Fri Dec 08, 2017 12:26 pm

Thanks for your input - unfortunately we do not expose feature requests.

Cheers,
Greg
User avatar
gjennings
 
Posts: 21
Joined: Sun Feb 28, 2016 6:53 am


Return to Feature Requests