Voice Activation Detection (VAD) Support

[Poco]

Operating System: Linux
SinusBot Version: SinusBot (0.9.21-87a90ce)
TS3 Version: 3.0.18.2

Problem Description
I have exported the settings.db from my TS3 Sinusbot into a Linux TS3 client and enabled VAD with a threshold at which I want my bot's voice blip to become enabled. I then exported the file back to the Sinusbot's settings.db location.
I expected the Sinusbot to respect my edited settings, however it forced continuous transmission back into the settings.db file..

I believe it might be useful for some people to have an option to enable VAD via the instance settings menu; if enabled you can select at which threshold level you want your bot to pick up audio.

Kind regards.

 

[irgendwr]

but... why..?

 

[Poco]

Well why not?
Some audio files have line-level white noise which can be resolved via a VAD option.

Is this a needless feature? No, I do not think it is.

 

[Diesmon]

but... why..?

Well why not?
Some audio files have line-level white noise which can be resolved via a VAD option.

Is this a needless feature? No, I do not think it is.

But the problem is not every song is mixed to the same DB level. So if you want only to start transmitting at X db then another quite song could be massively cutted and distorted due to cutting out of it. Aka another source of QQ threads.

 

[Poco]

Sinusbot has already has an option to offset the audio levels per individual audio file.
If this is an option, why is there no option for VAD?

I feel I should elaborate more on why I think this is an important feature..
I am curious if this could be implemented for use with podcast streams that are uploaded to the Sinusbot.
Typically podcasts recorded by the same author tend to have the same line-level white noise which can be eliminated via VAD.

Is this really something that is not useful for that?
If so, is it possible to have a feature implemented so that Sinusbot respects the VAD that I implemented into the settings.db file manually? This is another solution which will work for our server's purposes.

Regards.

 

[irgendwr]

Okay I kind of get it now but this could easily be fixed by editing the file in audacity.
Maybe this could also be implemented in the future but as this is quite a rare case it doesn't have high priority.

If so, is it possible to have a feature implemented so that Sinusbot respects the VAD that I implemented into the settings.db file manually? This is another solution which will work for our server's purposes.

This won't make sense/won't be possible in the next version.

 

[Poco]

I understand.

I've tried editing the files in Audacity.. however the results have turned out tinny and lacking quality in the end unfortunately.
Hopefully in the future there will be a better way than my current solution (I've been using Audacity..)