Just a thought: some of the best Poser animation I've seen wasnt necessarily "realistic" because true realism is almost impossible to achieve, even for the big time studios. No matter how hard they try, you can always tell CG work from the real thing... which isnt necessaily bad: I still think BEOWOLF was arguably one of the best pieces of CG work trying to make a human being from a technical POV that I've seen in a very long time. Human beings are so complex that it's nigh unto impossible IMHO to fully capture that, even with the best motion capture equipment and the finest models. That's why IMHO Poser works best with the more stylized models: the toons, the monsters, the robots. We're willing to be far more forgiving -- well, to a point -- because the level of expectation is different. Sure, we want them to look "real", but it's easier somehow to invest in making them real than it is when dealing with a mesh that already aspires to human real-ness.

Does that make sense? Probably not -- I'm still waking up. But just for a couple of examples, these are some pretty simple test animations that I'm doing for a much larger project:

http://www.youtube.com/watch?v=hPfNED6zlco
http://www.youtube.com/watch?v=Y9VdPAvf_VQ

In both cases, you're not talking about much: it's pretty  simple because it's just supposed to be testing out a few things. But it's the little details -- a grin here, an averted look there -- that gives it its own "reality" that, honestly, were you to use M4, would be almost impossible to duplicate. Here you're starting from a stylized origin. But with M4, you'd have to add on so many additional details, lttle quirky things that people just do, that you could get utterly lost in them... and you'd never fully capture them in the first place.

I dont know where this is going, so I'll stop now.

docandraider.com -- the collected cartoons of Doc and Raider