Can an original tween's animation curves be preserved?

Can this be done? When placing a key along a motion tween, I’d like the key to be placed with the original tween’s easing to be preserved rather than recalculated to the new key. (Also known as freezing a curve in 3D terminology) Freezing the keys on a curve would be acceptable too so I can keep the section that moves the way I want it to and reanimate the small bits that don’t behave.

Also, I’m using harmony 10.3.


A stop-motion keyframe might be your solution to this. Are you using 3D motion paths or separate axis paths?

Alternately you can go into the Position: Velocity and Position: Path function graphs to edit your keyframes’ bezier handles to modify/retain the original motion path.

Motion paths are described more fully here:

Thanks for your response, though none of these work for my purposes.

I need to know how to do this using separate paths as the animators don’t often use 3D paths to animate. Applying 3D path removes all existing tweens on the peg and none of the examples in those links show how to place a key on an existing tween with applied eases. Lilly is working with pre-keyed or non-eased curves.

My last ditch effort is to try to place stop motion keyframes entirely on 1s but even that doesn’t work as the curve of the tween is recalculated any time there’s a key placed along the stretch.

You can do this by having the original keyframe applying the ease then you add a 2nd keyframe at the end of the ease sections that is non-ease. You then move the two keyframes in tandem. Any keyframe placed after will not affect the eases of the original two-keyframe unit. You can do the same for the ending keyframes in reverse order (i.e. last keyframe being the ease keyframes and 2nd to last one starting at the beginning of the ease sequence but not applying any eases itself.)

As long as you don’t place a new ease keyframe between the original ease keyframe and its “capping” keyframe everything should be good. Of course I may be misunderstanding your question.