Hello,
Is there a way to to remove inbad DTMF from the recorded wav file (conditionaly)?
If not, how could be that done the easyiest way and a way it could get into FS git?
As FS has a great DTMF detector and it has probably all the stuff doing such a feature, I think it would be much better than using an external tool for removing DTMF from finished wav file - and it would consume more I/O probably.
Removing it would not be that hard (ie, inserting silence instead). Removing it without leaving a trace any DTMF was ever there sounds a tad more difficult. If you're ok with inserting a silence instead of the tone, extending the record_session app to detect and remove DTMF is probably what I'd do.
It is also possible to create a new app that monitors the audio stream and silence the stream when DTMF is detected, but since the detection takes a few milliseconds (~50ms), there would be a bit of "bleeding" of DTMF in the signal, unless the new app also delays the signal by 50ms and then has a way to strip the DTMF completely.