Gangster Ballet

Anhar_Miah

Veteran
ok, so I've been playing around with skinning a 3d model, (part of my long term goal) which is to use "3d photo scanner" method to capture a real human actors and then animate them inside a virtual set. Trying to establish a efficient pipeline really.

Yea I know its not perfect, I'm learning, used a motion capture data of a ballet dance.

Gangster Ballet
http://www.youtube.com/watch?v=rHlolH2BSyE
 
Last edited:
well, its ok not totally photoreal, of course its only as good as the texture that came with the model, I'll post a frame grab from the video (orginal).

My intention is to use this workflow for "Digital Extras" guys in the background, kinda medium to distant shots, I don't thing I could get away with close ups, thats asking for ILM level of dedication (although not impossible).
 
I wish I had a mocap setup :) sadly thats the preserve of the Big boys, I'm just using mocap data.

Its just a theory I had for a pipeline, aka Digital Extras:

(1) Capture 3d human actors (in custome) 3d photo scanning method (see 3dsom)

(2) Import model and rig using pre-made rigs

(3) Apply mocap (or hand animate :O ! )

(4A) Align camera OR motion camera track

(5) Composite.

Why not just shoot real actors I hear you say?

Well, (1) you can use multiple guys (think crowds), switch to different actors, change angle, change lighting, etc.
(2) digital stuntmen, using ragdoll tech, you can simulate them being hit, thrown about, and die realistically. Something that could be difficult to do in real life,
 
Back
Top