Create marker in recording based on user action

[havremunken]

Hi there,

I have been using Mumble (Murmur server on Ubuntu with DigitalOcean, various clients on Windows, Mac and phone/tablet) for years for recording a podcast. The primary reason is that Mumble is the only app I've found that gives me files which properly sync up; This makes post processing and editing much easier.

However, I would love to be able to do one thing - to have a script, command etc. to have Mumble store a "marker" in the wav files we end up with. I am not sure exactly how these markers are stored (I am not familiar with the wav file format in intimate detail), but when I use for instance Adobe Audition to edit, I can save a .wav file with markers in it, and they will be there even if I open the wav file on another computer. So I am guessing there is a somewhat "standard" way of doing this, accepting that I might be entirely wrong about this.

So what I would love to see is some sort of command, or even GUI element, that would allow me to set a marker, that is stored in the recorded wav file.

Bonus 1: I could be able to enter a "label", marker name. This could be "Segment start", "Edit point", "Participant burped", "Episode end", that sort of thing.

Bonus 2: I could choose if the marker was only created in my local ("Recorder") file, or also in the local recording of the other participants.

Why does #2 matter? Well, in order to have the best possible audio quality, we have all participants record multichannel locally. Then they upload their local ("Recorder") audio files to me. Now I have a sync challenge - since audio boards and computers have slightly different timing, my local recording of the remote users will be slightly longer or shorter than their local recording. I can "stretch" them in Audition and sync this up manually based on the waveforms, but if I could tell Mumble to place a marker ("Episode start", and "Episode end") in each of the participants local "Recorder" file, that job would be MUCH easier. I realize network latency means this wouldn't be nanosecond precise, but it would be good enough for a talking head kind of podcast.

I don't know if this is at all feasible - could it be done? Or would it require adding functionality to the Mumble client itself?

Thanks for any input on this! Really loving Mumble for the audio quality and excellent features!

[kissaki]

Hello,

good to hear it’s of great use to you.

WAV files can hold metadata. The question then is what kind of metadata Audition uses. Possibly a propiertary format (= not defined as a standard).

https://en.wikipedia.org/wiki/WAV#Metadata

Any change like that requires changes to the client, because (a part of) the client is the one recording and storing the recordings.

For Bonus 2 it would also mean extending the messaging protocol and network code handling it.

In principle there is nothing standing in the way of this. It is technically possible, and not too difficult. But we’re short-staffed and have little time and a lot of open tasks.

Have you considered creating a feature request in our issue tracker? I think it would be better there because it will be tracked, can be followed (for updates and discussion) and can be voted on.

Your description is very good as it provides context and is elaborative enough. It would serve well for a feature request ticket as well.

https://github.com/mumble-voip/mumble/issues?q=is%3Aissue+is%3Aopen+recording+label%3Afeature-request

[havremunken]

Hi kissaki,

Thanks, I'll be happy to create it as a feature request. I understand I can't expect this any time soon, and of course I wish I could contribute some code instead of just increasing the burden, but hopefully enough people will see the value that this will be given some consideration. I have no doubt this would make Mumble even more complete as a podcast recording solution!