I know this is old news, but I thought I would put my two cents in. There's a technique where you bring a JPG sequence through AI and render out multiple passes which you later composite back in AE. Here's an example I pulled together:
http://smg.photobucket.com/albums/v603/igd/?action=view¤t=test10.flv
Here's some basic info about the clip. I reduced the frame rate by half (12-15fps). The clip I rendered has 4 passes (1 color, 1 dark shader, 1 faded color outline, 1 dark detail outline). The process destroyed face detail, so I took the original footage and overlayed just her face at a low opacity (just enough to make out eyes, lips, etc.) The process itself takes an extraordinary amount of render time (especially through AI) but considering the alternative of drawing each frame by hand..... eh, i'd rather set an overnight render.
http://smg.photobucket.com/albums/v603/igd/?action=view¤t=test10.flv
Here's some basic info about the clip. I reduced the frame rate by half (12-15fps). The clip I rendered has 4 passes (1 color, 1 dark shader, 1 faded color outline, 1 dark detail outline). The process destroyed face detail, so I took the original footage and overlayed just her face at a low opacity (just enough to make out eyes, lips, etc.) The process itself takes an extraordinary amount of render time (especially through AI) but considering the alternative of drawing each frame by hand..... eh, i'd rather set an overnight render.