Material IDs not used when rendering.

[AdmiralKew]

I've got an object that has been voronoi fractured and set so that the chunks that intersect with the beam objects are deleted. The RGB-Y coloration is for testing. The red material is supposed to be a painted hull material, green would be bare metal, and blue would be a molten metal material. 

The material IDs appear to be applied correctly in viewport, but when rendering using Arnold, all tyflow particles render as Yellow, which I used to represent 'error.'




This is the flow, currently:

What should I be doing to fix this?

[tyFlow]

Arnold doesn't have an open SDK, so the tyFlow implementation in Arnold was handled by the Arnold devs. Because of that fact, you'll have to contact them about any rendering issues.

[AdmiralKew]

Arnold doesn't have an open SDK, so the tyFlow implementation in Arnold was handled by the Arnold devs. Because of that fact, you'll have to contact them about any rendering issues.

Thanks for letting me know. Cheers!

[MadsD]

Arnold doesn't have an open SDK, so the tyFlow implementation in Arnold was handled by the Arnold devs. Because of that fact, you'll have to contact them about any rendering issues.

Just a quick note, I monitored the thread on arnold answers.

Arnold can render any and all IDs through Tyflow Particles.
I wrote a unified particle shader in OSL that can control Material ID, Element ID, Object IDs

I also note that he resolved this through the original path by using a mesher.

   

[patfish]

OMG!!!... I have the exact same problem and was trying to fix it the hole day :/ ...it got me crazy because in the viewport i saw the hole time what I want to have in the final rendering   ... so is there no simple option to work with Multi/Sub Materials (different MaterialIDs) in Tyflow  with Arnold? ...in my case I use a Multi/Sub Material to combine a camera map (Frame of a Video) and normal Arnold Surface Shader for the inside Material of my debris. 

thx,
Patrick

[MadsD]

Well, upload a file with an example you cant get working and I can analyse your approach and see if we can get it going.
It really is not advised to use the mesher.

You can potentially get my shader but work differently. Its sorta like Forest Packs Forest color map, just better.

[patfish]

Well, upload a file with an example you cant get working and I can analyse your approach and see if we can get it going.
It really is not advised to use the mesher.

You can potentially get my shader but work differently. Its sorta like Forest Packs Forest color map, just better.

the thing is, i wanted to show some students of my course (they are mostly Maya + Arnold users) some basic distruction stuff in tyflow (with their exported alembic animations) ...but if there is no normal "standard solution" to make the texturing work with Arnold -  it makes no sense to show them all this stuff ;-) 

It looks like Vray is at the moment the better option to use for tyflow? 

regards,
patrick

[MadsD]

You cant bind it up on a single missing fetch to a legacy node and say that vray is better option, since this is already working you just need to code a bit, that may not be for everyone.

I will try see if I can get the multisub working with Arnold, I have not tried at all, I just know that Material IDs works perfect with Arnold and Tyflow through the tools I have written.

---

Here is another way to get it working right now.
You just generate a float channel in DCM and parse it to an array map, it returns a construct that is ready to be shaded with Arnold.

I also added a max file so you can see how to go about it.

Since the Multisub seems broken, I just confirmed this, it is the multisub that is broken, it is not "working with material IDs" that is broken - Ill file a note on the developer slack channel so they know this is an issue and that users seeks to be able to use the old way of doing things.

   

[MadsD]

I got a comment from dev, it's known, they are looking into it.
Should be resolved in one of the upcoming builds.