Can Openshot use Audacity for special audio effects like ducking, denoising, etc.?

[raindropsfromsky]

Openshot uses Blender and InkScape for implementing different features. Can it use Audacity to achieve ducking and other effects, such as denoising? Probably the idea is to invoke Audacity in commandline mode with parameters.

Note that ducking was discussed 3 years ago (issue#2831), but IMHO there was no discussion about the real ducking algorithm, where one track (typically voiceover/speech) is sensed and another track (typically music) is reduced (link). Instead, the entire discussion was centered around reducing the voiceover+music tracks and then mixing them to avoid clipping. Finally the stale bot closed the issue.

Note that in ducking, the emphasis is NOT on clipping (which distorts the combined signal), but of intelligibility of the speech track in the presence of loud music track. So in ducking, the music track is supposed to be suppressed as soon as a speech token starts, and there is a soft (gentle) recovery after the speech token ends.

That said, if Openshot is simply able to invoke Audacity by passing both audio tracks and other quality parameters as variables, then we need not even discuss the nitty-gritty of how the function is achieved inside Audacity: Let it remain a black-box approach!

[Colorjet3]

This is a bit beyond my expertiese. I am going to assign it to the lead developer for his review before we make an entry in Trello.

[raindropsfromsky]

I searched for Audacity's CLI capability, and found indications that Audacity is not very good in that aspect. (But I am not a programmer, so I could be wrong!)

Tenacity is a fork of Audacity that may have this capability. (Fingers crossed!)

[leodf1]

Maybe the existing audio Effects need to be fixed first. Like all the milisecond Property fields are actually in seconds!