Camera reading mouth shapes

Report · Sep 01, 2020

I'm working on my latest puppet and I'm having trouble. The camera is picking up on my mouth shapes and is triggering my "Singles" group of mouths. I want this group to be only triggered by the number keys that are associated with it.

Additionally, my "Sad" or "Alt" mouth group is not working and when I trigger it, it is not locking. I'm sure this is all user error, but I've built several puppets like his and have not had these problems before.

Report · Sep 01, 2020

Could you share the rigging hierarchy? I am not sure what "singles" means - assume a group in the hierarchy.

The swapset shown on the screen looks a little funny - it has mouth in same swapset as surprised (which is normally a child of mouth) and sad (which is not normally a child) - but you might be doing things differently which rigging would help make clear. (eg maybe you are not using lipsync ).

To keep something on that is normally the "Latch" setting on a trigger. See how the letter buttons are different colors (black and gray) for triggers in swap set? i think some triggers have latch turned on and others do not.

Report · Sep 01, 2020

Here is the rigging heirarchy. The layers in the "Singles" group are showing up in response to the camera and creating an overlap with the "Mouth" group.

And ah yes! Latch wasn't checked on the "Sad" group. Thanks!

Report · Sep 01, 2020

Ah, i see. I think what is happening is the face behavior is looking for viseme tags for smile and surprised. It hides the siblings under the same group expecting them to be under mouth with the other visemes. Because it's not in the same group as neutral, it does not know how that it should hide neutral.

My guess is you shoudl drag smile and surprised from singles into mouth so all the standard visemes are together. You may also need to put a copy in the alt group for the same reason.

There may be another way to archive it using swapsets or something (maybe nueutral should be in that swap set from first screenshot instead of mouth), but i am used to the standard visemes being together, and non-standard ones being outside. Smile and surprised (and nueutral) are the ones that are webcam based rather than sound based.