Page 1 of 1
Convert image layer into vector shape with smart bone dial?
Posted: Mon Mar 20, 2023 4:12 am
by Mattyj
Just wondering if there is a way to morph an image layer into a vector shape with a smart bone dial? I've had some fun transforming my vector characters into other forms e.g. old man into demon into Santa Claus, but is there a process where you can do the same with image layers?
Would it require having the vector layer you'd like to transform your image layer(s) into connected to the same mesh(es)?
Re: Convert image layer into vector shape with smart bone dial?
Posted: Mon Mar 20, 2023 5:35 am
by Greenlaw
I think the way to treat this is the way image and video morphing is done in compositing programs like After Effects and Fusion, which is to use a mesh warp to deform both images and then cross dissolve between the two. (A true morphing effect has more going on but that's the main nuts and bolts of it.) In most fx programs you would use spline guides to define the features you're blending between; Moho's mesh warping tools should let you do something similar.
To make the transition a little easier, you might pre-render the vector art to a bitmap so you're actually blending between a bitmap and a bitmap. Hayasidist wrote a neat script a while back that can convert Moho vector layers to bitmap. The script is called
Bake To PNG.
Naturally this depends on the type of artwork you're morphing between (i.e., images with matching target features work best,) but if you're clever about it you can pretty much morph anything into another thing.
Once you have that figured out, anything that can be keyframed in Moho can be copied into a Smart Bone Action.
Hope this helps.
Re: Convert image layer into vector shape with smart bone dial?
Posted: Mon Mar 20, 2023 5:49 am
by Mattyj
Awesome thanks Greenlaw I'll try and work it out.
Re: Convert image layer into vector shape with smart bone dial?
Posted: Mon Mar 20, 2023 10:59 am
by hayasidist
This was bitmap to bitmap "morph" using animated masks in Moho:
https://youtu.be/rz0TZzQhDgU
exactly as Greenlaw said:
Greenlaw wrote: ↑Mon Mar 20, 2023 5:35 am
images with matching target features work best
the hardest part was aligning the images!