Make "Silence Audio" non-destructive

Already — at any rate, in Beta 2.2 — Silence Audio only suppresses sound, leaving volume envelopes intact (making albeit rather irrelevant adjustments to absolute silence).

Obviously, if I had virtually no alterations to the volume envelopes within a selection, then the simple thing to do would be to create an envelope at either end, and lower the inner section to zero. That might easily be undone at any time. But this is not so easy when there are fifteen or thirty envelopes to adjust/remove.

Where there have been a lot of adjustments, it would be a huge timesaver instead to use Silence Audio — so long as I might then “Un-Silence” the same section, should I decide to return to it at a later date.

How easy would it be to implement this?

A variation on the same theme would be the ability to apply a sort of variable "Master Volume Control” to a selected section: one might be able to drag, say, a horizontal yellow line up or down, raising or lowering the overall volume of everything within that selected section?

Any ideas whether this would be an easy thing to implement in 2.2?

I see that someone has already suggested this in the Wiki — could you add my vote.

Silencing audio is intended to be destructive, what are you talking about?

I use Audacity for one main reason: it’s a destructive editor, not a DAW. Sometimes, you need to be destructive.

If you’re building really complex sound assemblies, use a DAW. These are non-destructive by design.

Can you imagine a use case that would need 30 envelopes?

It’s unlikely to be implemented soon, though I am keen to see a number of improvements to the “Envelope” feature, and “non-destructive silencing” could possibly be implemented as an enhancement of Audacity’s Envelope feature.

Vote added.

I think this is a good idea. It would be useful for more than just the “make silent” effect, but all volume-related effects - like amplify and fades, or in my case auto-duck.

I’ve been using Audacity for a while, but for the first time I tried to do some speech with an audio track playing in the background. I discovered the “auto-duck” tool and the “envelope” tool.

Envelope was great as I could fiddle with the level it and it was quick, and each “fiddle” didn’t result in progressively distorting the original. However, it meant I had to manually go through and do each “duck”.

Auto-duck was great as it automatically did the ducking for me, but 1. it was a lot slower as it had to go through and modify each of the samples, and 2. if I got the level wrong and wanted to adjust it, I had to “undo” and then reapply each time (so even slower).

If the auto-duck analyser was re-written to apply an envelope transformation rather than editing the original clip, it would combine the best of both worlds.

I agree.
I’d like to see the idea of envelopes expanded to become more general “controllers” that can control other things in real time, such as panning, and (for certain types of effect that could support it), as effect parameter controllers.