[Feature Request] Mix second (unfiltered) source

[m1ga]

NoiseTorch does an awesome job in filtering out noise and make the microphone completely silent when I'm not talking :+1: But in some cases it is too much and it would be good to mix in another unfiltered source (e.g. desktop audio) to have a base level of background music or notifications going.

Would it be possible to mix in another audio source inside NoiseTorch that will be added to the filtered mic? I know it is possible inside OBS or through different sound sinks but it would be very hand to have this in NoiseTorch so it can be used in tools where I can just select one recording source.

[lawl]

can you elaborate on your usecase a bit more? Elaborating what you want to achieve would help me decide how/if to implement something.

[m1ga]

I want to combine background music from a browser/web-radio with the mic output of NoiseTorch to avoid the 100% silence so the users know that there is sound and I have a base level of music in the background going on the whole time (since there is no "noise" many people thing audio is not working). And in an online conference I can only select the noisetorch mic as my input.

In OBS I would create two audio sources: Desktop audio + NoiseTorch so I can stream both sounds combined.

[lawl]

avoid the 100% silence so the users know that there is sound

I think something that could realistically be done is to have a checkbox to send comfort noise instead of complete silence when the voice activity detection is muting the mic. Would that work for you?

Letting the user chose their own "noise" would be harder to configure (i'm trying to keep things as simple as possible), and harder to implement.

[m1ga]

That would be good, too.

I though it would be easier to select "default audio" (or even select "Firefox" from the application list) as a second source like microphone since you are looping over the sinks already. Didn't think about creating a noise generate or adding a file player.

My idea would be another "Select background audio" section below the "Select Mic" with the output sources (ctx.outputList?) where I can select to tool/source and it will be mixed into the "NoiceTorch Mic" sink. But I didn't debug the source yet to see if everything is already in there :smile: And I don't know how to connect the sinks afterwards

And thank you for the fast replies!!

[dmaahs2017]

I don't know if this is what you mean, but this feature request reminds me of sidechain compression/ducking within OBS https://obsproject.com/wiki/Filters-Guide#sidechain-compressionducking