Say source A is an input movie, and source B is an image still.
I understand the what the first morphed frame would result from:
The first morphed image result will be the result of two frames, the first designated frame of A and we'll call that frame A[n] and image B.
What images are morphed for the second morphed image:
- Frame A[n+1] and image B, progressed two iterations in a morph of A[n+1] and imageB
- Framet A[n+1] and image B, progressed one step in a morph
- Frame A[n] and image B, but progressed two iterations in a morph of these frames, perhaps superimposed upon frame A[n+1]
- Frame A[n+1] is morphed with a the nth morph of B towards A. (poorly worded - hard to describe)
Also, are the points specified in frame A[n] used for all subsequent frames ... A[n+1], A[n+2] and etc.? Perhaps the user could manually adjust points in subsequent frames if there is movement of the source face?
The input movie function sounds like it might be deep feature, but I don't want to wrongly assume if or how a morph is being applied to subsequent frames.
Is there a sample morph on the site which shows the morph of an "input movie" and an image, or two "input movies" together?