Mumble Communication Recording

[tsi2112]

I am a graduate student that is involved in a research project looking into communication, and I am looking into various communication products for an experiment that we are attempting to develop. The experiment will involve 4 individuals communicating over a computer network, and this will require us to record the audio between all of the individuals.

After looking at various products and software, we are interested in Mumble due to its capability in multichannel recording and the whisper function. However, we are looking at if there can be possible modifications on the recording end. Testing out the recording done by Mumble, it appears that multiple audio files are created for what the recording client hears from the other clients, whether this is whispered to them or it is communication to everyone. What we want to do is attempt to separate the recording based on whether what is communicated by whispers or what is said to everyone in the channel. I am wondering since Mumble is open source, is it possible to have this capability? Could there be multiple channel recordings that records audio to a given file only when a specific hotkey is pressed (whisper hotkey for each person)?

[kissaki]

> is it possible to have this capability

Yes, of course.

Are you looking into implementing it yourselves then?

[tsi2112]

We are assuming that we will need to host our own servers, but we would greatly appreciate help with setting up the network for this type of recording.

[kissaki]

With the current code-base, AFAIK it is not possible.

So you will have to create a feature request and/or implement it yourselves, potentially submitting the code.

[tsi2112]

Thank you for letting us know.

We are looking into recruiting someone with experience to help us in modifying the code for what we need. Could anyone tell us what software and coding languages are needed to take on such a task, so we know what type of experience the programmer needs to help us?

[kissaki]

He definitely needs good C++ knowledge.

The toolchain for building should not be a problem, all free/open source, and our wiki has a guide on how to build. As long as you don’t want to distribute installers, no problem.

I would also consider this a relatively easy task (compared to general complexity). You will only need to modify Murmur, which at the moment simply transmits the audio packets. Recording-code is there for the client, so that may come in handy in how to write out the audio streams.

[GrantWilson]

tsi2112study writemyessay wrote:

I'm a graudate student that is doing a research paper project on communication products and the ideas how to imlement communication methods attampting to develop an experiment involving 4 other students.

>>>>

View full quote

 

We made multi-track recordings with SoX. I can send you the materials: + communication products > Ventrilo vs Matrix and the other project was: Riot vs Overtone vs Growtopia. We analyzed the results based on the students' commentary and syrveys: 230 experiment participants.