Mute myself and mute own channel leads to sound coming back to headphones

[ann0see]

@mulyaj discovered a bug:

Steps to reproduce:

1. Click the "mute myself" button
2. Click the "Mute" button on your own channel in the mix

Actual behaviour:

You hear sound coming back from the server (it's local monitoring)

Expected behaviour

No sound coming back from the server

[DominikSchaller]

The sound is actually not coming back from the server, it's direct monitoring and I believe it's a feature not a bug.

[ann0see]

Ah. Interesting. But that shouldn't be the case if you're muted in your own mix.

[DominikSchaller]

I believe the mixer is always referring to the sound coming from the server and as you are not sending anything the mute button has nothing to mute. I wonder what the use case would be in which you are playing but nobody (not even you) should here it.

I'm using Janulus 3.6.2 on linux which works correctly (not like you mention), and as expected. Please give more details.

[ann0see]

Yes. Can verify that it's not coming from the server.

I believe the mixer is always referring to the sound coming from the server and as you are not sending anything the mute button has nothing to mute.

I think Me and mulyaj just fell through an issue @gilgongo mentioned (mixing board analogy).

I'm using 3.6.2git (on Debian 10. Not sure which commit I used to compile it).

[chrisrimple]

I've just tested this on 3.6.2, on a server for which I have 200ms Overall Delay (making it easy to differentiate between local and server-returned sound)...

1. MUTE MYSELF = false and MUTE = false. I DO hear myself returned from the server.
2. MUTE MYSELF = false and MUTE = true. I DO NOT hear myself returned from the server.
3. MUTE MYSELF = true and MUTE = false. I DO NOT hear myself returned from the server. I DO hear myself locally.
4. MUTE MYSELF = true and MUTE = true. I DO NOT hear myself returned from the server. I DO NOT hear myself locally. In other words, I hear nothing.

I believe that all above is "as expected".

[ann0see]

Ok. Just trying to compile the latest version from master. Will reverify if this "solves" the issue.

[ann0see]

Ok. Compiled the latest code from master and if I connect to a server and click on "mute myself" and mute, local monitoring starts.

@chrisrimple:

MUTE MYSELF = true and MUTE = true. I DO NOT hear myself returned from the server. I DO NOT hear myself locally. In other words, I hear nothing.

Is not the case for me on Debian 10

[ann0see]

I could now also reproduce the other behaviour: Click on mute and then on mute myself.

Strange: Now this behaviour seems to be ok for me.

[pljones]

Does the order of doing it matter? i.e. If you had "Mute myself" on when you clicked mixer mute, is it the same effect as if you had mixer mute on and clicked "Mute myself"?

[mulyaj]

What happened exactly was the following:

• @ann0see and I were in a Jitsi call together, and were both using screen share with audio muted
• We were both in the same Jamulus server
• My personal mix was muted first and after when I enabled Mute Myself, the personal mix mute was overrided and MUTE = false

[pljones]

1. Connect to a distant server (London to Los Angeles, for interest)
   
2. Neither mixer mute nor mute myself - can hear audio from the server
   
3. Mixer mute only - cannot hear audio, although the server can
   
4. Mixer mute still on, add Mute myself - cannot hear audio: nor can the server
   

[pljones]

More tests...

Does it make a difference if only one user has connected to the server?

Does it make a difference if the user joins an existing session?

Does it make a difference if another user then leaves the session (leaving more than one user)?

Does it make a difference if another user then leaves the session (leaving only one user)?

I compiled the latest jamulus-master.zip (3.6.2dev-) and compared it to my installed jamulus 3.6.2 and found that Both work 'as expected' (as chrisrimple shows above).
Thus, there is no regression (and no problem either) from my point of view.

The testing both versions used 'the same personal configuration'. Maybe this problem depends on how you set up Jamulus or your computer.

Note: that I came across this new build message with jamulus-master.zip 3.6.2dev-
(did not affect building) :

qmake-qt5 Jamulus.pro
fatal: Not a git repository (or any parent up to mount point /home)
Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set).
Project MESSAGE: building an intermediate version: 3.6.2dev-
Project MESSAGE: Jack Audio Interface Enabled.