This week has been all about lip-sync. Let’s talk about lip-sync.
In cutout-style animation like O:P, most of the actual animation is done via motion tweening — limb joints, body bends, walking, etc, are keyframed. The object position/rotation starts on Frame X and ends on Frame Y, and the software calculates the motion in-between (hence “tweening”). For things like lip sync, we fall back to a more traditional-style “drawing substitution” process. Recorded lines are broken down into phonemes, and each phoneme has a mouth/face shape associated. The shape of the mouth making an “ooo” sound vs the shape of a mouth making a “shhhh” sound, etc. As the dialog progresses, different mouth/head drawings are swapped out to correspond with the sound being made.
Lip-sync can be one of the most time-consuming parts of the animation process, and this week I started a scene that (a) is the longest scene in the episode, and (b) has a *lot* of dialog. Whew. It was looking like such a slog that I had to give myself a bit of a pep talk to get the ball rolling, during which I referred to it as “gotta punch through the custard”.
Toon Boom Harmony, the software we use for the animation, has a built-in tool to try and automate the process. It analyzes a sound file and then creates drawing-swaps based on a standard naming scheme. It doesn’t ever really do that good a job, and you always need to go in and fix errors and refine it.
Thanks to the amount of lip-sync I had to do this week, I found I reached a tipping point — going in and doing the lip-sync myself by hand was actually *faster* than letting the software have a crack at it and then fixing it’s errors. It also has the added advantage of encouraging more stylistic choices rather than simply letting The Machine decide, the the characters end up with a bit more life & personality than they would have otherwise.
Really pleased with how this week went. Punched through the custard.