Erika & Nate:
Thanks for all these examples and thank you for the news of upcoming features for Spine.
In all of the linked examples, there are demonstrated only subtle changes in position of all (non-limb) character elements concerned - which, for games, when everyone's attention is focused on violent limb action - and quick, almost jerky motion of the entire character in general - is a fairly adequate solution.
Because games are not my current focus - and having been obsessed with the unsurpassed animation quality of cartoons which helped raise me as a person, (30's to mid-50's Disney and Warner Bros animations) - I seek the likely impossible, but hopefully probable solution for smoothly animating a character's movements through the entire range of foreseeable angles and poses - without having to individually draw each frame of an animation, manually.
Smooth interpolation between poses is the practical key to any such method.
Using the Head Turn (270 degrees - right to left and behind) as an example - and performing this movement on a typical cartoon character with a prominent snout (think Wiley Coyote as he prepares a trap for the Road Runner - imagining all of his necessary angles and poses for this task) - producing the movements of his head from right 3/4 to left 3/4 in Spine is completely possible using only one rig.
It is when the snout, mouth, ears and one eye pass the 3/4 mark and continue toward the back that computer mesh interpolation will become inadequate. Snapping of facial parts (due to image substitution) will inevitably occur and be blatantly visible - especially during a casual turn of the head.
The solution, typically, would be to draw each transition frame individually (and laboriously) - swapping these images in, frame by frame. And, maybe this is the best we can hope for - (but I hope not).
In the olden days of computer animation, we used simple techniques for transitions between images that were pixel based and crude. "Morphing" was the term for this type of image transitioning. I believe that the quality of pixel based transformation has greatly improved and become more intelligent as of today. Maybe this technique could be the answer to drawing every transition frame by hand.
When did you say the ETA for Spine 3.7 is expected to be?
What do you think?
Greg Smith