Maya production workflow?

[srbarnes]

Hello, I have a couple of questions about Maya's production workflow for modeling, animating, and rendering a character based upon the following scenario:

I modeled an animal (a rabbit) using polygons and mapped out the model's UVs. Next, I brought the model into Mudbox and sculpted a displacement map for the model and painted a color texture for it. Afterwards, I went back into Maya made a rig for the animal and subsequently painted weights (and otherwise prepped the animal for animation). Since the model was now ready to be animated I selected the mesh and created a Subdiv Proxy (using the Open Subdiv Catmull-Clark option) to serve as the high-resolution renderable version of my animal for my final renders. Lastly, I increased the subdivision level of my high-resolution mesh in preparation for rendering. At this point everything was working fine and my rigged low-resolution mesh drove my high-resolution mesh (like a WRAP deformer characteristically does).

1) If I intend to render my high-resolution mesh (i.e., Subdiv Proxy) via Mental Ray how do I deal with/apply the displacement map (.exr file) that I generated with Mudbox since Open Subdiv allows me to increase the resolution of my Subdiv Proxy high-resolution at will? (I realize that I probably wouldn't need to use the Mental Ray Approximation Editor because the Open Subdiv option allows me to increase my mesh's resolution almost as high as the model that I sculpted on in Mudbox).

2) If I wanted to apply fur either via Maya Fur or XGen to the my high-resolution model (i.e., Subdiv Proxy) is it possible? Will the fur move or correspond to my high-resolution mesh's surface? If not, when should I have done this?

3) Lastly, what is an ideal workflow for my entire endeavor? That is, how does this work in industry in places like Pixar or ILM?

Thank you for your time and consideration in responding to my inquiry.

[santd]

Hello,

Thank you for posting to The Area Forums.

1) You should be able to either use the Subdiv Catmull-Clark settings or create an mental ray subdivision approximation. You may want to run tests to see if one renders faster than they other, to me that would be the deciding factor as to which one I would use.

2) Adding fur or Xgen hair can be done after the rig is done and the mesh is bound to the joints or right before the finished model is bound to the joints usually done right before binding the joints and creating the rig. So at this point you should be able to add the fur or xGen hair to your high poly mesh and have it respond.

3) Large animation studios have a lot of internal tools and pipelines they use expecially for the size of the productions they create. When creating one character the process that you are going through should work well. Model, BlendShapes, UV\Texture, Hair\Fur, Rig, Animate. Normally I create a proxy rig of the mesh cut into different peices and then parent the mesh sections to the joints instead of binding the mesh. This makes for quick calculations when animating. Then I have the high poly mesh that is bound to the joints so that mesh can be hidden in a Layer and then shown come render time. There was no mention of blendshapes, but if you are using them, this is one thing that will want to create prior to binding the skin as oder of operation will have an effect on this and you will see double transforms on vertices. There are ways to work around this if by accident you bind first and then create the blendshapes, like re-odering the inputs.

I hope this informations is helpful.

Cheers,