Moho Forum

funksmaname wrote:yeah, just that translating that still means he has to re-do the walk... good practice though

If he used target bones, he can just parent those to the new translation root bone. Then just redo the layer translation on the new root bone.

Oh wait, that would change the animation coordinates wouldn't it?

Edit: if you align the base of the new root bone to the layer's origin, pointing to the right, its coordinates will be the same as the layer, so the parented target bones will work the same.

synthsin75 wrote: If he used target bones, he can just parent those to the new translation root bone. Then just redo the layer translation on the new root bone.
Oh wait, that would change the animation coordinates wouldn't it?
Edit: if you align the base of the new root bone to the layer's origin, pointing to the right, its coordinates will be the same as the layer, so the parented target bones will work the same.

Hum... wait, I don't understand. The target bones are parented to the leg shinbones, as in many Moho walk cycle examples. If I parent them to the root bone, how will they keep the character's feet pinned to the ground?

Reindeer wrote:Hum... wait, I don't understand. The target bones are parented to the leg shinbones, as in many Moho walk cycle examples. If I parent them to the root bone, how will they keep the character's feet pinned to the ground?

Are you sure this is the setup in the walk cycle demo you saw?

In my experience, the Shin bones should point to the Target bones but the Target bones themselves should never be parented to the Shin bones, not normally anyway. Otherwise, the Targets are going to translate with the Shin bones and make the targeting behave oddly.

I usually leave the targets 'freestanding' and not parented to a root...this way I can translate the hip to move the body wholly and the legs still point to the targets on the ground plane. I set the feet to have independent rotation so they stay 'locked' to the plane. Sometimes, I'll mark off where I want the footsteps to fall before I begin animating the scene--this can be done in a separate layer but often I'll just do a quick markup of the screen using the free Epic Pen tool.

There are advantages to parenting the targets to a root bone on the ground. The obvious one is that it can simplify and speed up walk cycle setups by letting you animate 'in place' and then just translate the root bone. This tends to look less natural but if you need to do things quickly, it's usually the way to go.

Obviously, there are many different approaches. The easiest method really depends on what your character needs to do in the scene.

Greenlaw wrote: In my experience, the Shin bones should point to the Target bones but the Target bones themselves should never be parented to the Shin bones, not normally anyway. Otherwise, the Targets are going to translate with the Shin bones and make the targeting behave oddly.

Oops, sorry Greenlaw, yes, of course, the Targets are not parented. I got muddled up there but the concept is clear to me. My character too has his feet at independent angles too.
So yes, now I see what you mean and I'll try the 'parenting to a root bone' thing. I've found that translating using the bones rather that the layer is a nice way to go. One drawback I've found, though, is that the origin of the layer does not move with the bone translation but rather considers the whole translation area (which can grow a lot, say, during a walk cycle or a flight of a bird), and this can be a little confusing.

As far as calculating where the steps fall in the walk cycle, in the case I have at hand now it's luckily not too crucial. When I need a 'real' walk where the feet don't skid, I've found that the additive cycle they added in Pro11 works very nicely.

And.. nice tool, that Epic Pen, thanks

Yeah, I know what you mean with the origin being fixed and not animateable.

What I do is I nest the rig layer inside multiple groups, each with its own origin position for whatever translation I need. For example, if I want some global squash and stretch, I would nest the character in a group with the origin at its feet and translate the layer from there. If the character needs to swing by his head for some reason, nest it in another group with the origin at the head. Obviously, this kind of setup needs to be set up per scene and you typically want to animate these layers last.

All I can add is, thank goodness for the new channel mute switches.

Edit: Naturally, if you don't need scaling and skewing, just translation and rotation, you can use additional 'pivot' bones too. For me, inserting the rig in groups with unique origin is usually quicker but this may be a matter of taste or just what I'm used to.

Reindeer wrote:

synthsin75 wrote: If he used target bones, he can just parent those to the new translation root bone. Then just redo the layer translation on the new root bone.
Oh wait, that would change the animation coordinates wouldn't it?
Edit: if you align the base of the new root bone to the layer's origin, pointing to the right, its coordinates will be the same as the layer, so the parented target bones will work the same.

Hum... wait, I don't understand. The target bones are parented to the leg shinbones, as in many Moho walk cycle examples. If I parent them to the root bone, how will they keep the character's feet pinned to the ground?

This is only if you don't want to redo the entire walk cycle to switch from layer translation to root bone translation. If you parent the target bones to the translating root (placed to match the layer coordinates), then your walk cycle will be intact and you could just translate that root bone to match the feet.

Of course, rebuilding a walk cycle is always good for practice.

ha! I hadn't even considered parenting the targets to the translation bone... good thinking.
This still means you will have to wrestle with potential floating though - I would re-do it personally.

funksmaname wrote:ha! I hadn't even considered parenting the targets to the translation bone... good thinking.
This still means you will have to wrestle with potential floating though - I would re-do it personally.

Very true, and unless the animation is very complex or very specifically timed, I'd be apt to just recreate the cycle too.