Dipping a toe in to Vantage

[thomas_bussey]

Long time Vray for 3dsMax user, never used Vantage, but now I finally have an RTX card I'm thinking of using it for a project I'm working on.

I'm wondering if anyone can provide comment/clarity on a few potential pitfalls?

Live link - You can't render a sequence from this right? It's just like an IPR by the looks of it.

Decals - AFAIK decals aren't currently supported? Worst case I guess I could bake to texture or use a 2nd map channel, assuming vantage supports map channels more than 1?

Deforming Meshes - AFAIK these aren't supported either? The project is an ArchViz one so obviously the building is stationery, but I have a large crowd sim made with Tyflow. I'm planning on exporting this to one big Vray Proxy, but whether these will be fully rendered or just silhouettes is still up for debate.

Render elements. I was planning on rendering with cryptomatte and Z depth using GPU originally. Cyptomatte is out of the question I understand, but looking at the docs I can render a Z-depth pass which would be used in post for DoF. Is some kind of material ID pass possible?

Depth of Field - rendering raytraced DoF would be great if possible - removing the need to mess around with different passes for cameras that pass refractive surfaces at close proximity. I guess it will take some testing to know how well this works for my purpose.

Would it be an idea to export the scene is chucks, into different vrscenes? this might make it less of a problem if I need to update one of them later. Aside for the crowd, the only moving element is the cameras, so I assume it would be most efficient to export the buildings as [multiple] STATIC vrscenes and then the cameras as an animated one?


Appreciate any insight in to these things, it may well save me a lot of head scratching.

[npg]

Live link - You can't render a sequence from this right? It's just like an IPR by the looks of it.

You can. There is a button for this right next to the button for initiating live link in Max. Moreover, rendering animations this way enables some features which do not work yet when loading a vrscene file into Vantage 1.8, like deforming meshes.

Decals - AFAIK decals aren't currently supported? Worst case I guess I could bake to texture or use a 2nd map channel, assuming vantage supports map channels more than 1?

No luck here. Decals are not supported and we still only support 1 map channel (we plan to support more soon).

Deforming Meshes - AFAIK these aren't supported either? The project is an ArchViz one so obviously the building is stationery, but I have a large crowd sim made with Tyflow. I'm planning on exporting this to one big Vray Proxy, but whether these will be fully rendered or just silhouettes is still up for debate.

Deforming meshes work over Live Link. Instanced crowds should work this way. Animated vray proxy is not supported in 1.8.

Render elements. I was planning on rendering with cryptomatte and Z depth using GPU originally. Cyptomatte is out of the question I understand, but looking at the docs I can render a Z-depth pass which would be used in post for DoF. Is some kind of material ID pass possible?

Indeed, we don't support cryptomatte. We do output Z-depth in EXR files and other compositing elements (back to beauty) will be supported soon.

Depth of Field - rendering raytraced DoF would be great if possible - removing the need to mess around with different passes for cameras that pass refractive surfaces at close proximity. I guess it will take some testing to know how well this works for my purpose.

We do have raytaced depth of field. Our motion blur is not raytraced, it is a post-processing pass.

Would it be an idea to export the scene is chucks, into different vrscenes? this might make it less of a problem if I need to update one of them later. Aside for the crowd, the only moving element is the cameras, so I assume it would be most efficient to export the buildings as [multiple] STATIC vrscenes and then the cameras as an animated one?

We don't support VRayScene objects, but this feature is in development. You could manually create a main vrscene file that uses the include directive to combine several other files. You can also drag-and-drop vrscene files in Vantage and choose to merge/place them and then save a .vantage file with this configuration - this is probably more convenient than manual #include. In either case you'd have to reload the main file in Vantage when one of the pieces is updated.

[thomas_bussey]

Thanks a lot for your comprehensive reply!

I'll have a play with it if I get the chance, but it sounds a bit like it's not quite suitable for my needs just yet (without extra hassle of baking textures etc)

Another unsupported feature I realised I've used quite extensively in my scene is VRDistTex. Can't see a way around that at the moment.