Gain Groups

[leonerd]

Often I have multiple different kinds of sounds in the same show (atmosphere/music, background sounds, interruption sounds, pre-recorded dialog/etc...), that would want to be played at different volumes. When moving between venues or sound setups it can be useful to adjust levels again for different environments.

I'd find it very useful if I could group up all the sounds into distinct sets, and have one common volume/gain fader to apply to the entire collection. The UI could then display these as a set of virtual faders in a little mixer box somewhere, possibly in its own window.

I'd then use that by creating a few named ones for "ATMOS", "SFX", "DIALOG", and assign each sound to one of these faders, so I could adjust the entire set of them in one place.

[leonerd]

Oh I forgot to add - most of the point of this is to allow live editing of volume levels, via some interactive faders on the screen. Rather than just editing numbers in the individual cues ahead of time.

[fnetX]

To be honest, I think that implementing a fader GUI is out of scope for LiSP. This is better done in virtual mixers where this feature set is already more complete.

I think the cleanest option might be use Jack and a mixer to send the outputs to separate channels, and allow for easy selecting an output channel for each cue. But I think the user experience for this approach is probably a little lacking - it is probably complicated to get right.

A compromise for this work like this:

There is a new cue type to group cues for permanent settings. Like collection cues. But if a volume action is run on this new "group" key, it does not only affect cues currently running, but reduces the volume for all cues in the group to the maximum volume (including fade actions if selected).
The feature could be extended to allow other settings to be defined for multiple cues at once, and selecting the cues they should apply for later.

[clemradio]

+1 for the feature of having more than 1 audio output. Would love to see decks, maybe like 4 of them and each one would send to a different output. You could then assign cue presets to each of these decks and every cue in the list with a preset would then be played on the preset’s deck.

[FrancescoCeruti]

Something similar has been requested, at least in the form of a "global volume control".

Now, the way the backend is designed doesn't allow to mix cues together, and, as far as my experiments has gone, it's quite tricky to build a mixer directly in GStreamer while keeping each cue state independent. That said I would be very beneficial to have such design.

I think it came up in other circumstances, and similar to what @fnetX wrote, we could try to keep cue settings synchronized, but it's a very fragile solution especially outside of a "live" scenario, also an independent volume control for each cue should be retrained anyway.

[moonshaadow]

I think like @fnetX that a robust solution is the use of JACK, with a simple mixer like jack_mixer. You could even manage your sessions with NSM tools like RaySession. This is working well for the asking task.

Another approach is possible if you don't want to use JACK, but it is depending of your hardware : different outputs from your soundcard, and enough free inputs on your mixdesk, or a mixdesk as soundcard directly.

I am not aware of how such a feature could be hard to achieve, but we could imagine a cue acting like a VCA, not like a group . This way the volume of the controlled cues will be relative of the volume of the VCA cue.