you are here [x]: Scarlet Star Studios > the Scarlet Letters > principles of pose-to-pose stopmo
<< before pose-to-pose stopmo
after >> the making of "a word from professor ichbonnsen"

February 4, 2008

principles of pose-to-pose stopmo

by sven at 6:07 pm

I thought of a different way to summarize the animation approach I'm exploring. Here are six key aspects:


1) The Sound Leads
When you don't have a voice track, you have a lot more freedom. As soon as you have talking, you're committed to a more structured approach. Lipsync more or less demands using an X-sheet. Yet, there's still room for improvisation with the body.

Personally, I'm finding that I prefer animating when there's a voicetrack. When I know what the character is thinking, I have more fixed points in time to which I can attach interesting gestures. I'm tempted, the next time I animate a non-speaking sequence, to record a dummy voicetrack of the character's thoughts -- just to help guide my performance.

(I know that Nick's animations more often than not don't have a voicetrack -- so this is probably one of the most fundamental points at which our methods begin to diverge.)


2) Use Your X-Sheets
Lots of animation gets done without X-sheets. For me, X-sheets are vital for lipsync... I want to have done my voicetrack analysis and checked it for accuracy before I start the real animation, so I can just focus on the rest of the performance. Further, I figure the more information I'm able to capture on the x-sheets in advance, the more additional details I'll ultimately be able to layer in.


3) Hold-Transition-Hold
The rhythm of the animation that I'm doing is "hold-transition-hold." I make sure a pose "reads" by holding it still -- until it's time to transition to the next pose. Hypothetically, a pose could be held for as little as two or three frames... It needn't necessarily feel like the motion is "stop-and-start." Hold-transition-hold doesn't preclude overlapping action, either; secondary motions can be hold-transition-hold on their own time-line.


4) Non-Linear Pose Development
By using a digital camera locked down in the same position that I'm going to use for the final shot, I can take a whole bunch of shots of different pose options, brainstorming... Instead of treating a pop-through as a sort of dress rehearsal for the final performance, I assemble a pop-through from photographs taken in no particular order. A virtue of this method is that it allows you to select your best options from a large pool. A limitation is that you can only really swap between photos taken when the puppet was standing on the same spot (i.e. walks may be more problematic).


5) Animating To A Beat
How do you decide how much time a transition should take? Rather than acting out each action and timing it with a stopwatch, I'm trying to use a metronome to learn what 4 frames, 8 frames, 12 frames (etc.) feel like in my body. Rather than assign a unique time-period to each transition, I'm trying to learn how to estimate how many frames different actions will take, based upon their musicality -- which is interpreted in terms of "beats per minute."

(Note: I recognize that this is an atypical use of the phrase "animating to a beat," which usually refers to animating a character moving in time with music.)


6) Pop-Thru Rotoscoping
Stopmo animation is inherently a "straight-forward" process; you can't work non-linearly, tweaking different frames into perfection. The closest we can come to a "pose-to-pose" process, I believe, is establishing keyframes -- and then figuring out how to get from one to the next.

In order to simulate keyframes, I am using photographs of the puppet that I am animating, taken from the same camera angle as the final shot. These "keys" are assembled into a pop-through film which can then either be used as a reference clip, or for actual rotoscoping.

posted by sven | February 4, 2008 6:07 PM | categories: stopmo