thoughts on different methods of animating
I've been having this discussion with a former animation professor of mine about 3 different methods of approaching a shot: layered approach, blocking approach, and extreme blocking approach (for lack of a better name). I thought I'd post that here since I learned a bit in writing it in terms of analyzing my approach. I've put some comments in [brackets] to clarify stuff that was just assumed in the email...
Someday I'll write this out with examples and better detail to make it a much easier to follow tutorial.
[This is the easiest method and best for beginners, though as you get more experienced, you'll find it takes much longer and is less effective than blocking.]
This is what I assume you teach your students, particularly the beginners. Where you animate the base bone throughout the shot first, in I'm guessing a straight-ahead manner. [So you animate a 7 second dialogue shot with just the base body bone bouncing around first.] And then layer in different body parts. [So the head is involved next, and then the arms are layered in, for instance.]
I always use this for walks and runs though I wouldn't be surprised if some people block out walks and runs. But damn, that'd be hard. For most acting and some action shots I stick to the blocking method.
[The problem with a layered approach is that it doesn't allow you to see if your poses are working until you're done. Because the poses aren't there until you're done. The director also can't really critique your work until it's almost done, prompting a lot of possible re-animation and wasted time. It's a handy method for walks and runs though because it simplifies these very complicated actions into something much easier to grasp.]
[This is the method that seems to be used by most professional animators, including myself. The key thing is that this is the fastest method for those working at studios where you don't have as much time as a place like Pixar, to work on your shot]
You have a 7 second dialogue shot. In your acting, the character maybe hits 3-4 major poses to hit the major acting points. And then a bunch of smaller subtler poses like settling out of a major pose or hitting a similar smaller pose like bringing your hands up before bringing them down and out. So you put in all those poses using stepped keys. And you pay attention to every body part in every pose. Nothing is forgotten. This way you're forced to keep everything alive in an intelligent way. If you don't, the blocking won't read as well and won't be as 'alive.' (Now you go get director approval and do some fixing/refining). Next, I'll also tend to put in a small amount of inbetweens and I always put in the holds now too. It can get kind of tricky to put in holds when your motion is stepped because you can only see they're there in the keyframes in the graph editor and not in the motion itself. I recommend putting all the holds in in one sitting, once all the poses are set in stone. Otherwise it can get really confusing. Then you spline the whole thing and get a semi-floaty mess, probably with a lot of spline overshoot that needs fixing. But if you had put in the holds it doesn't look half bad, especially after fixing the overshoot stuff. And you've very quickly done most of the work allready AND gotten director approval on it.
CRAZY OH MY GOD EVERY FRAME BLOCKING (this really needs a title)
[This is the method that seems to be catching on a lot at Pixar, and likely at other studios working on high quality feature animation. It involves taking blocking to the extreme by blocking the inbetweens to the point that you have stepped keys on almost every frame or every other frame.]
Do the same as in blocking, but don't spline it. And include some settle in every hold (on EVERY body part), instead of just making it a flat line hold. Go through each pose and act it out over and over, finding what each body part does subtly, and including it as a piece of inbetween blocking.
So far, it seems like the only way to make this work is to still have each pose occur all on the same frame. Let's say you want to do some subtle inbetween motion of what an arm is doing. So you block that in. Then there's some subtle thing you notice the body doing. But it really occurs a couple frames later when you act it out. But if you block that in a couple frames later, your blocking is suddenly hard to read. Because there's a little arm jerk and then a little body jerk and then suddenly the next major pose. So my conclusion so far is that you need to put those inbetweens on the same frame as eachother and overlap them later in the blocking process or when you spline them.
See, this is where it got complicated and difficult. So I'd love to see how someone else does it...
But if you stick with it and spend a lot of time on making it all work with this method, you will be forced to pay attention to details that are probably left out with just a normal blocking approach. And a lot of potential for great animation opens up there, if you can make it work.