POV-Ray : Newsgroups : irtc.animations : to those that understand : Re: to those that understand Server Time
28 May 2024 21:25:54 EDT (-0400)
  Re: to those that understand  
From: Markus Altendorff
Date: 2 Mar 2007 02:13:27
Message: <45e7ce97$1@news.povray.org>
John VanSickle wrote:

> The posing is done with an app called Marionette, which looks like a 
> spreadsheet of animation variables over time.  Pixar doesn't appear to 
> use motion-capture (although other houses definitely do), which makes 
> their work even more impressive.

I'm not sure how to put this in words, but most motion-captured 
animations in videos appears a bit "odd" to me. Most recent example i 
saw was "Final Fantasy VII: Advent Children" (yes, it's actually a 
movie, not a video game - but it sure feels like just watching someone 
play 1 1/2 hours without ever making a wrong move on the joystick ;) and 
before that, "Final Fantasy: Spirits within". Maybe it's because the 
actor supplying the spatial data doesn't exactly match the proportions 
of the CG character? Or maybe because the data needs to be smoothed or 
oversampled or something, removing what makes it "human" instead of 
"machine-like"? Whatever it is, it flips the "there's something wrong" 
switch in my brain... especially compared to Pixar. Maybe the manual 
tweaking approach, with the animators acting it all out in front of the 
mirror and then playing with the motion curves (i've seen the add-on 
DVD, too ;) works better - like an actor that needs to act 
"larger-than-life" on the stage to make it appear "natural" to the audience?


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.