#synchronized animations

So my question is, is it true that more scripted or "canned" interactions between two characters leads to less clipping than say if there are no canned interactions at all? Bc part of their design, is to create more visual accuracy, correct?

So in other words, is a game with a more scripted gameplay style, more likely to have less collision detecting, since one would think that more scripted gameplay leads to more accurate animation clipping. Or is that not accurate?

Here's the thing - animations don't inherently have any sort of [collision detection]. They're a lot like shadow puppets - they have no mass, they have no physics, they just exist as projections onto a screen. Animations just exist in space - they'll pass through anything that occupies that same space because there's no actual matter that can collide.

Since shadow puppets have no mass, they can't physically interact with anything. We can pretend to have them interact by acting as if they were affected by something - a punch, a throw, a kiss, a hug, etc. This is what animators do - they sculpt motion over time and try to line things up if they can. "Canned" interactions are basically paired or group animations that are designed to play simultaneously in order to make the overall motion believable. A shoulder throw won't be believable if the victim being thrown doesn't move in tandem with the attacker doing the throwing. We also often have generic reactions (e.g. getting punched in the face, getting shot, getting knocked down, etc.) but those don't look as good as tailored animations meant to go together.

Paired interactions are more expensive to build than generic animations - you need to build the animation for the skeletons of the performers, which means that it won't work for a different skeleton, and those paired animations can only be used together under specific circumstances (e.g. a shoulder throw shouldn't be done if there isn't enough space for the victim to land on the ground). This is one reason why Call of Duty's executions work - everybody uses the same skeleton, so the same animations can be played on every character. This won't work in a game where different characters have different skeletons - the normal shoulder throw animations won't work if one character is a pixie and the other is a dragon.

Beyond this, there's also the issue of clothing and character model. The animations might not clip, but enormous chunks of material on the character models (like flashy armor) will clip through if the animations don't account for those things. This is one reason why you often see clipping issues during in-game cinematics where the player's flashy wardrobe causes clipping issues because the animators never expected the player's shoulder pads to be as enormous as they are.

[Join us on Discord] and/or [Support us on Patreon]

Got a burning question you want answered?

Short questions: Ask a Game Dev on Twitter

Short questions: Ask a Game Dev on BlueSky

Long questions: Ask a Game Dev on Tumblr

Each participant in a synchronized animation acts using angles and positioning relative to some common point and performs all actions within a specific space around that common point. If one participant is off-angle or off-position, things won't look right. For the animation to look right, all participants must keep the exact proper distance and orientations while the synchronized animations all play at the same time.

This means that, as long as the environment and conditions are conducive (e.g. the ground is flat, there's enough space, there are no objects to clip through, etc.) to choose an in-game location to assign as the synchronized animation's common point, we can play the synchronized animation on the participants and things should line up. Any mismatch and things will look weird. If the ground is not flat, for example, we could see our characters standing in midair. If the positions are not correct, a punch could visibly miss by a mile. If the angles are not correct, it would look like a phantom judo throw. If the environment is not clear of other objects, we would see our participants walking/flying/falling through seemingly-solid objects. Only when all of the conditions are met will the synchronized animation look right.

The TLDR is that every synchronized animation was created with very specific dimensions, angles, and positions in mind. If we can recreate an environment with those dimensions, angles, and positions, we can play the synchronized animations on the participants and they'll line up. In games where you can perform synchronized animations in arbitrary locations (e.g. executions in Call of Duty), the engineering team has built a system to find good places near the position and orientation of the action that meet all of the necessary criteria in order to play the synchronized animation.

[Join us on Discord] and/or [Support us on Patreon]

Got a burning question you want answered?

Short questions: Ask a Game Dev on Twitter

Long questions: Ask a Game Dev on Tumblr