Nuke<>Mari Environment and Matte Painting Bridge Released

[Burritoh]

Something else worth mentioning is...

The baking delay bug on the Windows version is now gone! It had been around since the early days and was something of a nuisance.

1.3v2 fixes that, and we no longer have those random but frequent hang-ups just before baking occurs.

[elfufu]

i think my only concern with this process is the dependency on UVs always, in the example shown, the fact that he needed to use a 16 k map to paint on, EVEN though it was reduced later in the projector scares me. would be nice if the projectors in mari could have the option to act the same way projection cameras work in nuke, where its never baked onto the mesh and not relying on UVS.

[forelle]

i think my only concern with this process is the dependency on UVs always, in the example shown, the fact that he needed to use a 16 k map to paint on, EVEN though it was reduced later in the projector scares me. would be nice if the projectors in mari could have the option to act the same way projection cameras work in nuke, where its never baked onto the mesh and not relying on UVS.

Hi

As Mari supports ptex (uv less workflow) you don't need uvs at all. When you pull over geo you can choose either ptex or uvd workflow. Without uvs you can just send back new projections.

We still have to bake to be able to paint on the surface, but ptex means you're free to work as you like, and also to dynamically add resolution where you need it, on a per face basis.

We think the combination is pretty powerful.

J

[oglu]

is nuke able to read ptex files..?

[forelle]

is nuke able to read ptex files..?

Not at the moment although I'm pushing for this.

If working in a PTEX workflow you can send back new projections from Mari. By sending back projections (Mari rendered plates and project3D nodes) Nuke doesn't need to know anything about PTex. The ideal solution would be Nuke supporting PTex as a native format, but a projection based workflow is pretty efficient.

J

[molgamus]

So at the moment the Nuke<->Mari bridge isn't really ideal unless you have UV's? However I do see the benefit of projecting plates onto geometry with ptex and then have the geometries rendered in V-Ray.

I have not tried this workflow out since Mari isn't available on OSX (will it be?). But the renders from V-Ray could then be projected onto the geometry once again for use in Nuke. A bit of round tripping, yes, but Mari<->Maya could probably be scripted with python. Until there is ptex for Nuke that is.

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.