Firstly, I’m sorry if I you can’t understand all what I mean. My English may not be good enough to express all my ideas. 
The scenario is, I’m interested in making music visualization videos (my first work is here). The biggest problem I face is syncing the visual animation with the music. Probably for a long & complex music track this method can’t work well. So I propose a new feature which should solve it more easily.
[size=150]Here’s how the feature works:[/size]
A user (like me) has already created all the waypoints to animate an object (or layer), but the its movements hasn’t corresponded with the sound layer yet. So he uses this feature in the following steps:
- First, choose all the needed waypoints and put them in a queue in an order.
- Click the “Play” button. The animation (including the sound layer) plays from a chosen start time but the queued waypoints have no effect yet since their time value is temporarily blank.
- At the right time (e.g 2s 20f), the user press a key (this action is record) and the first queued waypoint’s time is set (2s 20f). When the user press the second time, the second queued waypoint’s time is set. Same goes for the rest waypoints.
Advance:
- This feature can apply for waypoints of different layers
- User can record with multiple keys on the keyboard
It’s easier to understand when you compare this with playing Synthesia on the keyboard.
Now it looks like I’m the only one to need it right now. I totally understand if it’s not the priority in the development of Synfig. That means it’s best for me to learn to code and then add this feature myself 
.