I’m currently writing a proposal for an animation module at uni and was wondering if I could ask a few questions. Its on the research into character anatomy and physiology before rigging, and whether this necessarily leads to more convincing character performances and believability.
I’m confident that the obvious answer is “of course more research will lead to better rigging and animation, duh!” but obviously it’s a bit more nuanced than that and may differ from theory when practical concerns like cost and deadlines are taken into consideration.
So I guess what I’m asking the industry professionals here is:
[ul]
[li]how much research into character anatomy, bone structure, etc. would you perform before rigging and what form would this take?
[/li][li]To what extent would you try to model this anatomy accurately in a rig?
[/li][li]Do you think that this led to more believable animations?
[/li][/ul]
Research is important!
I spent a few days on the equestrian estate looking at horses before rigging Maximus for Tangled.
That combined with a huge reference binder of anatomical and other references for horses artists have combined over the years at the studio was incredibly useful!
I’ve done a fair amount of work for animal and prehistoric documentaries where this is fairly important but budgets and time are tight.
We generally have image reference and measurements of fossil skeletons for doing the modelling. We also refer to books and Google for anatomy and make best-guesses about how similar various muscle groups are. If it is a prehistoric animal with a modern analog, then we study reference of those animals.
For animals with specific characteristics or if the shot is highlighting something specific, I pay special attention to mimicking the best known reference that we have and get feedback from the art director and the clients if they have specific expertise. (paleontologists for example.) Sometimes their advice contradicts itself and we just give it a try and fix it in production if it is an issue.
But with that as a foundational guideline, I place the bones and controls with the goal of creating an animator-friendly interface. Usually there is no contradiction, but I can imagine an example where the spine is close to the rear of the torso, but to get better bending and twisting deformation, I’ll put it a bit closer to the center of the torso.
Another example, if a toe or finger bone has a weird bend or angle to it, it might be necessary to fudge the bone positions and orientation just to make the control chain have better IK/FK behavior. In these cases, this type of careful precision is usually superfluous.
[QUOTE=ShadowM8;14460]Research is important!
I spent a few days on the equestrian estate looking at horses before rigging Maximus for Tangled.[/QUOTE]
This sounds very similar to the research conducted by Disney’s animation team for their older features. WHat kind of activities did your research there involve? It’d be nice to have a connection between the research used during the birth of the animated feature and its modern equivalent.
[QUOTE=clesage;14461]I’ve done a fair amount of work for animal and prehistoric documentaries where this is fairly important but budgets and time are tight.
We generally have image reference and measurements of fossil skeletons for doing the modelling. We also refer to books and Google for anatomy and make best-guesses about how similar various muscle groups are. If it is a prehistoric animal with a modern analog, then we study reference of those animals.
For animals with specific characteristics or if the shot is highlighting something specific, I pay special attention to mimicking the best known reference that we have and get feedback from the art director and the clients if they have specific expertise. (paleontologists for example.) Sometimes their advice contradicts itself and we just give it a try and fix it in production if it is an issue.
But with that as a foundational guideline, I place the bones and controls with the goal of creating an animator-friendly interface. Usually there is no contradiction, but I can imagine an example where the spine is close to the rear of the torso, but to get better bending and twisting deformation, I’ll put it a bit closer to the center of the torso.
Another example, if a toe or finger bone has a weird bend or angle to it, it might be necessary to fudge the bone positions and orientation just to make the control chain have better IK/FK behavior. In these cases, this type of careful precision is usually superfluous.[/QUOTE]
Thanks Chris, this is hugely useful and supports what I’ve read elsewhere in various papers/books.Since I have to model. rig and animate a few animal forms to demonstrate my findings this gives me a lot to think about when doing my own research.
[QUOTE=Rob Galanakis;14468]Realtime or offline rendering?[/QUOTE]
I’m not researching either specifically, though those sources which are specific to one or the other imply what the best approach for that render method without going into much detail. It’s enough for my proposal to have implied that real time needs to be more minimal in its representation of anatomy in the rig than offline rendering. More detail than that would be nice but I don’t think it’s necessary for my proposal.
I think one of the things you forgot to mention is how will the animators interact with the rig itself. The interface is quite important. I would also look at the hierarchy of nodes and methods. example in Maya:
Direct Connections fastest
set driven keys
Nodes
Constraints slowest
Knowing when to use these will really help keep rigs light. Everything else in your research seems great. I would add an acting class, mammals are very similar, acting like the creatures not only helps animators it helps riggers as well.