studioGPU

What kind of 3D workflow do you use for CG animation?

I ‘ve been working on a story about 3D work flows and I would love to get some feedback on what kind of work flows other 3D artists have set up and are using for the 3D CG animation work.

Obviously there will be as many ways to set up a production pipeline as there are 3D artists, but from what I can tell, it boils down to only 2-3 unique variations.

The most common work flow is to model and then hand off to one or more lighters to render “in camera” within an applications like Maya or XSI, using a 3rd party rendering solution like VRay or Mental Ray. The pluses are that it feels like one environment.  The drawbacks are that the work flow is linear so you can get way downstream, doing a lot of test renders before you see problems or come up with a good creative. Also the rendering settings can be very complex to setup and if you use multiple lighters, you can get multiple results. Getting a great result can be time consuming, especially for animation. (See 3D Workflow Seminar from Assembly 2008.)

The second type of work flow relies much more on compositing in After Effects, Flame or similar software.  The animation scene is rendered in layers and compositors work with the rendered images in the same manner as the 3D, working with separate passes for things like specularity, diffuse, fill, reflection, etc. (See Professional Production Pipeline.)

Of course, MachStudio Pro offers yet a third way to work letting you work non-linearly “In Camera” to do both previs as well as final rendering. One artist can handle all aspects of production in a scene. Output can be a full composite or in layers for subsequent compositing.

I would like to get any feedback on other 3D work flows that CG artists are using and why? Where are you pain points?  At what points do you need to hand off your work to other people and when can you retain creative control yourself?  How does your work flow impact your ability to work with other creatives like the technical director?  Where would you want to see acceleration to speed up the inevitable multiple iterations? Do you use render farms or do you try to do it all in house? Really I am looking for anything that helps me better understand common 3D work flow pluses and minuses.