Facing challenges (final part)

Previously, I said that I'd like to animate small and complex movements, and I hope to use that to make the animation richer and more life-like.
But what makes your work life-like?
That's a question that many artist ask themselves at some point, and there's a simple way to understand the answer.
You look at life.
Artists have done this since the beginning of art, because to be honest, if you don't know how something looks it's quite hard to draw it convincingly.

This is something that helped animators as well. One example for traditional animation is Disney studios, where the artists would go to the zoo to study animals, and hire models for reference shoots at the studios themselves.
I will add that this raises a simple question. Because if you have actors performing the whole scene, why not just trace the reference?
This is true, and has a name: rotoscoping. Historically, this has always looked awful. Every. Single. Time. The artist's hand in the animation is almost completely lost, and the result looks incredibly bland.
I will not look to rotoscoping.

But there's one thing I will look at: motion capturing. You might be familiar with this from Avatar and Peter Jackson's the Lord of the Rings. Here's an example from a videogame called L.A. Noire:

L.A. Noire: The technology behind the Performance by Team Bondi, uploaded by IGN

I really like at how they let a computer look at the actor by making it look at the volumes in the actor's face. I think this is important because these volumes, usually muscles and/or fat, not only define how something looks, but also have an impact on what's around them.
An example: the face.
Eyes are very important in any acting performance. The eyes can move themselves easily enough, but they are also subtly pushed and their shape slightly distorted by the volumes around them. These are the eyebrows and cheeks. Here's a test I made a little while ago, where I marked the eyebrow and cheek volumes.

Now what really fascinates me is the idea that these volumes can be taken separately, and streetched and squashed and moved to fit onto a different face. Maybe something non-human like the face of a snake. And then their movement can be used to make that snake act as the human would, with all the subtle movement intact and the way it would be expected to move...if snakes could talk.
I have not tried this yet, but it will be my other challenge:

Use capturing to include characters' (human and non-human) subtler motions into the animation.


And that was the last of the long technical posts about the function, or reason why I want to make this animation.
The next post will be about the contents, so I get to tell you a story.
And to round up I'll post about the visual design where I get to show you some of the works that inspire me visually. Great stuff :)

Challenging myself (part 2)

Previously I told you about the challenge I set myself. Now I'll go a little deeper and tell you about framerates and keyframes.

But first, here's where I come from.
In previous animations I studied motion. Big motion, small motion, slow motion, fast motion.
Here's one of the animations I made:



What did I learn from my animations?

First off, I looked at the fluency of the motion. When did the animation flow really smoothly, and when did it stutter too much?
Anyone remotely familiar with how movies work can tell you that more frames in each second of animation gives you a smoother image. 24 frames per second (fps, or framerate) is enough for movies to look good, and is a standard amount.
Imagine making 24 drawings for each second of movie. That would be a lot.
Now imagine making 1440 drawings for each minute, and a staggering 43200 for half an hour of animation. That's about one episode of television animation, which needs to be done every week.
So animators cheat a little. Often, they skip a frame, drawing only on each second frame (frames 2, 4, 6 etc.). This is called animating 'on twos'. Sometimes they also animate 'on threes', skipping two frames and drawing only each third one.
This will save an animator a lot of time, but how does it look in an animation?
Short answer: it looks bad, because it stutters and the eye can tell the separate frames apart. The illusion of motion is gone.
In my animations, I also noticed that subtle motions are lost in lower framerates. In lower framerates, a big motion like the moving of an arm would still be visible. But the blinking of an eye would be gone. Someone shaking their head from laughter looks incredibly weird, with the head seemingly instantly moving all over the place because there are no frames in between to show the head moving from one place to another.
So a higher framerate is better.
Digital camera manufacturer RED has written an article on why they started pushing this further, shooting in higher framerates beyond the standard 24 to 48 and even 60 fps, and share some examples so you can see for yourself.
One result of theHigh Framerate (HFR) is that small movements become visible in actor's expressions. This seems very important to me, and I'd like to take this with me for my new animation.

But I'd also like to have some time left in my life to see my friends, family and loved ones instead of drawing all day long.
I learned something else from my animations that might help me.
In animation, the animators first draw a few important frames which will be the key to making the animation work well. Let's use an example of a bouncing ball, with the animation first, and its frames below it:



Images from Wikipedia's Public domain images. Made by Wikipedia user Branko.

Which frames are important?
The first and the last are important, because they tell us two things: when does the motion start and end, and also where it starts and ends. It starts on frame 1 on the floor, and ends on frame 6 on the floor again. So these are the first two keyframes. But what happens if these were the only two frames, and someone else would come along and draw the frames in between (just like your brain fills in the blanks between each frame)? He'd see that the ball was on the floor in the beginning, and would still be on the floor at the end, but how would he have guessed that the ball is supposed to go up? Another keyframe is needed somewhere in the middle to show the path of the ball.
Which frame this will be is just as important as where the ball is drawn in it because that shows the speed of the motion:
If the ball would go up really like a bullet but then land like a feather it will reach it's highest point a lot earlier and our middle keyframe will be put closer to the beginning than to the end. And vice versa, if it would go up slowly then crash down, the middle keyframe would be put nearer to the end of the animation.
If the inbetweener would come back now to draw the frames in between, he would now know that the ball is supposed to go up, and how many frames he needs to draw of the ball going up and back down again.
Now there's one question that I asked myself when animating. How many frames would I need to draw in between each keyframe?
What if the ball was actually supposed to stay on the ground. If we had only frame 1 and 6, we already concluded that those were enough. For a simple motion like the ball above, we'd need six frames total for about one second of animation. Did you read the RED article I linked to above? Then you have seen that a complex motion like the motorcycle jump is best displayed at a really high framerate like 60fps.
It's safe to say that different kinds of motion need a different minimum amount of frames.
One of the things I will be trying to do with my animation is try to capture complex and small movements like that of an eye or a speaking mouth, and to make it look better than what you'd expect from the animation one can see every day, drawn on twos or threes. For that I'll need higher frame rates. But at the same time I will also want to keep the amount of work I need to do to a minimum.
Why not have both?
Here's my first challenge:

Draw at a variable framerate, using what's needed for the motion, up to 48fps.

I hope this will make it possible to add subtle details to my animation, and enabling me to draw more complex expressions on characters and make fast motion look less jagged than what's expected of hand drawn animation.
But there's one more challenge I'd like to set for myself, which has to do with acting. I'll tell you about that in my next post...