Moho 12

[Greenlaw]

Oh, wait...I see where I goofed up the above. Without thinking, I had copied the 'warp' keyframe in the reference mesh to frame zero--naturally, this invalidated the warp because the base shape no longer matched the original shape. The actual warp keyframe needs to be at frame 1 or later of course. I'm guessing I made the same mistake with yesterday's scene too.

Also, upon reflection, I was doing this 'backwards'. I should be placing the meshes inside the Switch Layer and using Interpolate Sub-Layers to warp a single static image outside of the Switch Layer. This should give you smooth transitions between warps using a Switch Layer. (As yet, untested by me though.)

Oh, wait...no, that won't work. It would only work if the Switch Layer itself could be seen as a 'mesh', which it isn't of course. D'oh! (Although, it would be a cool idea if the vector data could be passed upward and the current state of the Switch Layer was be selectable as a 'mesh'.)

For now, I think this is a dead end. Probably best to stick with the methods described previously.

Moving on...

[Psmith]

Wes:

I revised my Pin Bone setup and the animation as per your suggestions. The result is better - but now it more clearly indicates that there is a compounding of keyframe action as the cycle progresses. Everything that has been Translated or Scaled becomes exaggerated with every cycle.

As the worm crawls, he gets longer and longer - and the portion that has been scaled by means of one Pin Bone - gets fatter and fatter.



Not an optimal result.

Do you see this as I do?


Greg Smith

[synthsin75]

Nevermind...it sort of works but it's not reliable.

Basically, here was my setup:

1. I placed a character face bitmap image in a Switch layer and made a reference copy, and called them ImageA and ImageB.
2. I created a vector layer, traced out the face and placed guides where I wanted the warping to occur. Then I converted this into a mesh.
3. I made a reference of the mesh layer, and renamed the layers MeshA and MeshB.
4. For each of the Switch Layer images, I selected the corresponding meshes for Smart Warp.
5. I set the Switch Layer for Image B and and then edited Mesh B for my second pose.

This seemed to work at first but after a while, the Smart Warp stopped deforming the second image. It would still mask the image, just no deformation.

I'm not sure why you'd need references here, Greenlaw. Wouldn't duplicates of the image layer all controlled by a single smart mesh work?

[synthsin75]

Greg, the last frame in your cycle needs to be a copy of the cycle's start frame, only with all bones (ctrl+A) translated to the end position. Since your last frame includes the offset and scaled bones, Moho thinks that you intend that cycle to add those changes (from you first to last frame) to every cycle. If the only change between the first and last frame of the cycle of a straigh-line translation, Moho will only add that to each subsequent cycle.

I hope that's clearer than my previous answer. Some of this takes translating what's somewhat intuitive for me into followable steps.

[Greenlaw]

I guess using duplicates or references in this case really doesn't matter does it? They probably use the same amount of RAM in Moho. In 3D animation programs I use, a duplicate object will add to RAM usage but instanced object typically does, not but I'm not sure that idea translates to Moho and referenced image layers. But anyway...

The idea was to have differently deformed images inside the Switch Layer, with differently 'posed' meshes for each image in the Switch Layer. But this setup is pretty silly...in practice, I can't see why anybody would to do this. It's a lot of extra stuff going on with no real benefit. I think you'd be better generating static mouth shape images using a straight forward Smart Warp setup, and then importing those static mouth shapes into a Switch Layer for your rig scene. That would be less complicated and a lot more efficient cpu-wise.

Now, if the idea is to use a single mesh layer to deform various static images in the Switch Layer then, yes, that should work. However, I think the results could be dubious because a mesh warp 'pose' is probably going to be specific to a given image in the Switch Layer, and it will probably look wrong for a different image. I guess this depends on how broadly different the Switch Layer images are and how far you're going to distort these images using the Smart Warp mesh.

This was fun to explore and I did learn a few new tricks but, IMO, I really don't see any of this reducing the tedium of animating lip-sync and facial aniamtion.

Personally, I don't think animating lip-sync is really all that tedious once you've done it a few times. Just import the dialog track, drops some markers on the timeline to notate keywords and certain syllables, and then work through the dialog using the pose presets method of choice. With practice, you should be able to get a decent first pass done lip sync done in a few minutes, and refine the mouth-shapes in a second pass using transforms, Smart Warps or bones maybe still within the hour or so. All the extra stuff mentioned above isn't going to make this process go faster.

[Psmith]

Learn to use Pin Bones to create general animated Deformation. Utilize the Pin Bone's abilities to Rotate and Scale portions of a "Soft Body" - thus correcting any deformation problems.

Create a "Worm Cycle" with only 2 Key Poses:



Learn Fast!


Greg Smith

[Greenlaw]

Pretty nice! Thanks for posting that.

BTW, remember that in Moho 12 we can now separate the channels and remove any unwanted keys on x that might be adding vertical distance with each passing loop.

[Welsh Jester]

Now, if you do want the warping effect to be animated, you''re probably better off using a static image with a Smart Warp applied, again without the Switch Layer, and creating a set of mouth poses as Morph Actions (phonemes or expressions). If you label the Actions clearly, you can just click through the different poses for your lip-sync, and Moho will interpolate between the different Smart Warp positions.

Hey, when you say to use Morph Actions do you mean just Actions or Blend Morphs? Also i appreciate all the information you posted, thank you.