View Issue Details
|ID||Project||Category||View Status||Date Submitted||Last Update|
|0000269||Cinelerra-GG||[All Projects] Bug||public||2019-07-17 08:34||2019-07-17 08:34|
|Target Version||Fixed in Version|
|Summary||0000269: Keyframe fader automation|
|Description||From the forum, dankinzelman made these requests (https://www.cinelerra-gg.org/forum/help-video/ideas-for-ui-modifications-improvements/):|
Keyframe fader automation: Ardour/mixbus have a very nice system going here, with the option of switching automation states from 'manual' (which ignores automation) to 'write' (which writes new values according to fader position as the track rolls, overwriting existing automation until the transport is stopped and it switches automatically to 'touch') to 'touch' (which plays existing automation and switches to 'write' as soon as you click on the control, then switching back to existing automation as soon as you release it) to 'play' (which ignores manual input and forces the control to follow existing automation). The current cinelerra implementation seems to be a sort of version of 'touch', but it does some strange things if I try to write automation while the video is playing back, and it would be nice if it gave up control for the duration of manual input. Currently, if I try to draw automation as the track is playing, I end up with a sawtooth in which Cinelerra alternates between writing my inputs into automation and jumping to the preceding state for the parameter I am automating (if this isn't clear I can make a video to illustrate). Sample rate for automation capture seems very low as well, which isn't ideal for more delicate work.
I have done some more testing, and I made a video to illustrate the problem I'm talking about (check the attachment). To describe in words, if the fader automation has already been modified, further modifying that automation creates zigzags since there seems to be some indecision about whether or not to overwrite the already existing automation once I assume manual control, and it ends up alternating between existing automation and new automation points (at its low automation sampling rate). I think in the video it should be clear what I mean, check out the zigzags.
|Steps To Reproduce||See:|