Behind The Scenes at AWS: Aldrich on Weight and Movement

+ In this, the first of a series of interviews with the animators and designers of Adam Walker Studio, I talk to 3D animator Aldrich Tangpos about the importance of animating weight convincingly, the methods, and the trials associated with the procedure.
We discuss the evolution of his animation, from his Shadow of the Colossus-inspired second-year art school project, to the professional work he is involved with at Adam Walker Studio today.

How do you go about convincing the viewer than an animated object has a real weight? Well, for example, if a character was carrying a sword – lets use Shadow of the Colossus as an example – when I did that piece, my knowledge of animation wasn’t as developed as it is now, so when my character grabbed a sword there was no sense of weight or anything with the sword itself, he picked it up in the same way as if it were, say, a twig.

Right, now if I was going to animate that today, there would be a more exaggerated movement in him lifting it up and seeing his whole body react to it. There would be anticipation from his lifting up the sword – in particular, I would have his body move first before the hand itself – a synchronisation between his hand and his body.

In what way? The body would go first, followed by the hand. We use our bodies to help control weight and balance. He’d shift his torso first, then the hand. If you do it in that order it looks as though there’s some sort of weight to the sword.

So it looks as though he’s actually having to work to pick up the sword? Action, reaction, and anticipation are all factors. He has to hold the sword in a manner that shows he’s supporting the weight. You can’t have him in some neutral or awkward position, because the viewer will know that it’s awkward. My characters back then, when I animated them, they were very slow. I didn’t have a general sense of speed on screen. Every movement they made was just really slow.

Shadow of the Colossus is based on a small guy fighting this gigantic monster, but both of them were moving at the same speed. Which is not the case in real life because really, the little guy should be running really fast, and all of his movements should be really fast, but in the animation he’s just as slow as the colossus itself. And that goes back to the weight. The reason that the colossus moves so slowly is because he’s gigantic. And I kind of show a sense of weight to the colossus, because he is a bit slow, and I did little camera tricks, like shaking the camera, particle effects every time he takes a step, so you can tell he’s gigantic.

Why is it important that a larger body move more slowly than something much smaller? It’s realistic. Anything larger in life like that tends to… well, visually, it looks slower. Because it’s bigger, I think that you’re got to factor in how much weight it would have if it were real. How would it move. Look at Godzilla. Godzilla’s gigantic, and you can see it doesn’t run as fast as a velociraptor. The bigger it is, the slower it’s going to take for it to make general movements.

From Richard Williams’ “The Animator’s Survival Kit”

So going back to the character picking up the sword, what preparations would you make before animating that now? Do you reference anything? I visualise myself picking it up. In some cases I will actually try to do the movement myself, literally grabbing an object to see how it’d work. But normally for me I’ll just picture it in my head, and I’ll animate it and go through a process of ‘does it look right’, ‘does it not look right’. I’ll make tweaks, I’ll do bits and pieces to make sure everything works right, and just make it look… believable.

If we were to compare the sword animation from Shadow of the Colossus to say, Arcade Trap, which was a training animation that I did here at Adam Walker Studio, there’s just a massive difference because I had more knowledge with animation in general. You could tell his bat had shoulder weight because there was a lot of reaction and anticipation to his movements.

So instead of - say, if I held the sword in my old animation, he’d swing it, but there was no swing-back - no anticipation for the actual swing – he’d do it straight away, following a direct path from point A to point B. What Arcade Trap does is basically add a point in between the two – point A.1 or something. He has to swing back, then forward, and then he has to react to that movement, and then he has to react to the connection of the bat onto something else. I didn’t have that in Shadow of the Colossus.

And what sort of effect did that have on the piece? Well, my knowledge of animation wasn’t so great, so I looked at it and thought ‘yeah, that’s not bad at all’, but I look back now, and I’m just embarrassed that the stuff that I used to know was just… unrealistic. It didn’t look believable. It was interesting, but… I mean, with Arcade Trap, everything’s a lot faster. It’s quicker, everything reacts as it should. General reactions to things like the bat connecting with things go the way they should.

From Richard Williams’ “The Animator’s Survival Kit”

In (AWS-produced game) Piccolo’s Adventure there’s the end segment, so when you complete the game there’s a little animation of (main character) Piccolo. And he has a backpack on his back, and he’s carrying a basket full of tools. That was quite interesting to play around with, because we basically had two objects that were pretty heavy, and I had to work out a way for him to walk. He’s wearing the backpack, which is quite heavy, so he has to lean forward to adjust to that weight, pivot forward. But then at the same time I had to figure out a way for him to carry the basket and to show that, to show that that was pretty heavy also.

And how did you do that? Well, the way the animation goes is that he drops the basket, and takes off the basket and takes off the backpack. So from that I have to animate the scene so that he gets rid of the weight, and so you can see him do so convincingly. He has to react to it. After he drops the backpack and the basket he goes back to his original, default pose, to show the effect those objects had on him. He was moving slower with the objects, and when he drops them his shoulders become much more loose, his arms are down and his head becomes a bit more loose. I had to convey a sense of relief, a reaction.

Was it difficult to animate considering the two sources of influence on the character? Did you have to think more carefully about that process? Did you have someone model the action to show what it should look like? Well, I have a backpack myself, I walk to work with one, so know how it feels to have a heavy backpack. So with that in mind I just played about with it, adding the real-world knowledge I have to the animation. If it doesn’t look right, then obviously it’s wrong. But if it looks as though it has weight, that he’s really carrying weight, you’re heading in the right direction. But there are some situations where if you can’t seem to get it right, you would actually act out the movement yourself to see.

It’s the same case of when I tried to make animation back in college, I used to actually film myself just doing separate things, and I’d try to overlap that, to use that actual footage as a template and animate in front of it and sort of trace the animation, or, the movement of myself. That was another exercise for myself to learn about animation, but it didn’t really get far as I realised how much time it took. It was too tedious.

So now you can just do that on the fly? In your head? Pretty much, in a sense yeah, now that I’ve got the basic ideas to know what exactly the character should look like for the type of animation. It depends on the type of actual animation, but you have to work out certain techniques

Anything you weren’t aware of before you started? There was a lot. The whole anticipation and action and reaction; that was all new. When you animate a character, it doesn’t matter how realistic it looks, as long as it moves believably. Adam taught me that one. My characters back then, when they moved, they didn’t feel like they were believable. Generally their animations were slower. My timing, the keyframes, wasn’t as good as it should have been – some were longer than others, and that’s why some animation, parts of my old animation just looked wrong – there would be fast sections, then all of a sudden it was slow. But when I started working here, and started learning more advanced animation techniques, I had a better sense of the constant speed that a character should always remain at – I just worked with that until I built up more info in my head.

Were there any aspects on human movement that you weren’t aware of before you began animating? Are you hyper-aware of any aspect of these things now? Yeah, it was quite interesting when I came here, I basically had a week to watch Adam animate, and it was just amazing. There’s this one moment I’ll always remember; I was watching Trent animate, and in the very early stages of his animation I was trying to understand how he got his characters to move so convincingly. And there was a certain point when he was maybe 70% done, where it just hit me, like “Oh my god, so that’s how you do it! Wow!”

So what was he doing?The techniques he was using were, say, the little ways in which he’d move a pelvis or the hands or feet and all that – that’s all well and good, but when he played it - the real-time playback - it just looked right. And I saw how he’d done it, and I saw how I should have been approaching it. He did a lot of things with the keyframing and the placement of the body parts – that’s when it sort of hit me: weight needs to be conveyed as a realistic force, otherwise scenes won’t be as believable or convincing, or even taken seriously.