Oh boy, here’s a big one. I would say I have advanced in my understanding of TVPaint, meaning I’ve learned a few things but I’ve still got an eternally long way to go before I know all the ins and outs. This project started out with one goal and then spiralled into a whole host of other experiments and now here we are with whatever this is. At the start of everything, I didn’t care for a story necessarily, I just wanted to have as detailed and clean a sequence as possible. I wanted to avoid line jitter, I wanted the movement to be smooth and believable, I wanted the sequence to look interesting, with lots of moving pieces. I love animation where the entire screen is filled with movement, not going fast or all over the place, but with smooth, seamless movement.
On this, I grade myself a… pat on the back.
I mean, I am proud of myself. This is the most complex project I’ve done on TVPaint so far—there are sequences here (the clothes coming together, the makeup, the burger) that took weeks to do. But when I watch it back, I see all the areas for improvement: The shadows could be more logical, the human figure gets a little wobbly here and there and lacks a sense of solidity. BUT, I also love it. I’m particularly proud of the tomatoes. And the hair.
My first thought was burgers
I wanted an interesting animation with detail and movement, so my mind went to a cheeseburger. Maybe I was just really hungry, but I thought about all the different shapes and colors that it would require and went with it.
I did a little test animation of a pair of hands unwrapping a burger and liked how it turned out. This test one was rotoscoped, technically, which I guess is kind of looked down on in animation the same way tracing would be looked down on in a drawing. I used the main poses from this animation in my final animation, but shortened the movement and added some changes to the drawing to make it feel closer to animation than rotoscope.
Once I had my burger down, I moved on to other things that could interact with it. Hands were one, but I didn’t want to commit to doing a human figure and the complexity that would add, so I did one but covered most of it in a big dress, and had it spend half its time as a blob. You can’t go wrong with blobs. Make your blob a balloon and you have sophistication!
Balloons, burger, repeat
Everything in the final version has at least 1 additional version as a sketch. I worked out the movements at the sketch level, then went over those frames with clean lines. At the beginning, I was naive enough to think I could maybe do just the keyframes as sketches, go over them with lines, and figure out the inbetweens at the clean line level. Anytime there’s an awkward bit in the final sequence it’s because I was lazy and didn’t work out my sketch first! Lesson learned.
Once the lines are down, coloring in TVPaint is a breeze (but sometimes not!). The software also has an automatic shadow/cell shading function, but I’ve never been completely pleased with it, so I did my own manually, which was often mind-numbing. Backgrounds were relatively simple and only required a few tricky experiments to nail down the zooms and pans.
I didn’t mean for any of this to happen, really. I just wanted a fun little project to tryout some techniques and ended up with this behemoth (relative behemoth, OK). The final sequence is just a chain reaction from blob to human, to hair and makeup, and what better way to wrap it all up than literally wrapping it all up in paper in the end.