OpenPBR: rendering results changed between using MaterialX and Maya standard

I have rendered the OpenPBR samples available on MaterialX’s github and found that the results are different when transcribed to Maya’s standard OpenPBR and when using vrayMaterialX (as well as in USD).
It seems that the problem is occurring in Transmission and Subsurface, could you please investigate?
We are aware of the following two points
- The internal units of subsurface_radius may be different between the Maya standard and MaterialX.
- Transmission_depth is not received properly from MaterialX, and any value is treated as 0.

		**open\_pbr\_glass**
		![](http:\/\/forums.chaos.com\/download\/attachments\/276064624\/open_pbr_glass.png?version=2&modificationDate=1741750865147&api=v2)

​​


            cause unknown
open_pbr_honey

            ​
            Transmission_depth receipt is not working.
open_pbr_ketchup


            Units of subsurface_radius do not match?
open_pbr_pearl


            Units of subsurface_radius do not match?

​​ ​

We are aware of the limitations and have already made significant progress in our internal builds. In there, when exporting either of those OpenPBR surface presets to MaterialX (via the V-Ray > Tools > Export materials to MaterialX tool), I get identical results. Those fixes will go to the stables builds soon enough.

Thank you for your reply.
I have reviewed the content and have some points of concern, so I conducted further investigation. I would like to share these findings and ask some questions.
Regarding the point that the appearance matches by changing the colorSpace, I find it inconsistent that it matches in lin_rec709 for base_color only, or when coat_weight=1 and coat_darkening=0, but sRGB conversion is required only when coat_darkening=1.
The table below shows a comparison of the appearance between Arnold and V-Ray when the colorSpace is fixed to lin_rec709 and the MaterialX attributes are the same. It shows that the appearance differs only when coat_darkening is involved.
My questions are: Is this coat-related processing specific to V-Ray? Also, are there any plans to make the appearance closer to other renderers when using MaterialX and USD in the future?

## version
mayaUSD: 0.31.0
vray: 7.00.04
mtoa: 5.5.1
maya: 2026

Linking threads.

Okay, I did some further testing and indeed the result is wrong - a regular VRayMtl with OpenPBR mode and Coat Darkening = 1 gives similar results to Arnold. Also, when we export a MaterialX for this VRayMtl, the results are correct. Only the sample is wrong. I’ve logged the issue (internal tracker ID: VMAYA-13478 ) for dev investigation.