Or: When is a Time Loop Layer Like a Movie Clip?
Okay, so I’ve figured out at least one way to use the Time Loop layer to get actions to repeat when I want them to; now I’d like to find out if it’s the best way and, if not, what is.
In this instance, I have a character who is nodding his head in response to a series of questions. So, he needs to nod, then stop, then nod again, stop again, and so on. I put the nod - plus a blink and a resting expression - in an exported canvas and put a Time Loop layer over it in the main canvas. So then I animated the parameters of the Time Loop layer to either hold on the resting expression, cycle once through the blink, or cycle twice through the nod action (for a nice, vigorous, double nod). Once I’ve set keyframes for each action in the Time Loop layer, I can then just duplicate those to repeat the same action later in the animation. (And what could really speed this process up would be the ability to right-click over the new keyframe widget marker, copy, then right-click and paste that keyframe elsewhere in the time line.)
While doing all of this, it occurred to me that you could do a lot of character animation this way and maybe save time. (I’m sure more seasoned synfig folk have already figured this out. 
) You could basically replicate Flash’s mouth comp system in an exported canvas, then use a Time Loop layer to control what fires off when. The advantage over just copying and pasting frames in the animation canvas itself is that, with the Time Loop approach, you’ve only got one copy of each frame or action, so edits instantly take effect for all “instances” of that frame or action (just like with movie clips and flash).
Anyway, does that seem about right or am I still doing stuff the hard way?
Matt