Moved from “General Forums > Blender and CG Discussions” to “Support > Basics & Interface”
When it comes to modeling and texturing, Blender plays nice with others; it has reasonably good bidirectional support. That is, you can round-trip a model in and out of Blender using exchange formats (OBJ, FBX, Collada) without losing too much in the process. There are a couple caveats, but those usually apply to models intended for games rather than animations.
For steps further down the pipe, Blender might not be ideal unless your plan is to go into Blender from another program and stay there for the rest of the process. That is, Blender’s support for typical animation exchange formats (FBX, Collada, point cache) can be a bit hit-or-miss. So if you want to animate in Blender, plan on rigging from it. Likewise, an animation-friendly rig in Blender might not export cleanly for use in another program, so it probably won’t be wise to plan on rigging in Blender and animating elsewhere.
This also means that you’ll probably need to manage rendering from Blender (you’re not limited to BI or Cycles, but if you want to use a different renderer, you’ll need to make sure there’s exporter support for it from Blender). There’s rudimentary point cache support, but I don’t think it’s mature enough yet to assume you can bake a point cache from Blender and render elsewhere. Point caches from outside sources might be able to be pulled into Blender, but they’re currently a bit difficult to manage.
For the most part, yes. The tools are there for creating high quality visuals (I can reference a few projects if you haven’t already researched this). As with anything, it’s the final polish touches that may prove to be frustrating. A lot of them involve physics simulation. For example, hair dynamics are difficult to control and cloth dynamics can occasionally blow up on you if you’re trying to mix them with a traditional rig. In planning, I’d either make sure to allocate time to wrestling with dynamics or assume that you’ll need to rig and animate them by hand.
Blender does have basic support for pulling in mocap data (BVH files, mostly). However, the toolset for managing motion capture data is a little bit sparse. Retargeting is a difficult task (but not impossible).
I know it sounds like I’m being rough on Blender here, but I’m really not. I’m giving you my take as another animation producer who also happens to do art. I love Blender and it’s a major (the major?) part of my pipeline. But I’m also aware of its current shortcomings and I’ve always planned my productions accordingly. Granted, I’ve never had the size budget you’re talking, but it boils down to the same things in planning:
- Plan on working around/with Blender’s limitations
- Have a developer or suitably talented technical director capable of writing scripts (or patches) to help alleviate these limitations
As a producer, the advantage you get with Blender is the fact that you have a fully customizable in-house tool and quite good community support (both in terms of development and art technique). You also save on license fees, but that really won’t factor in until you’re farming your render (and you smile realizing you don’t have to pay a for a license per processor). The monetary cost for artist seats is negligible and that cost should be weighed against the cost of additional production time if you don’t adequately plan on doing the two things mentioned above.
Hope that helps.