MP3’s are better for me for a number of reasons: size, portable integration into web content, etc. If possible I would like to avoid going the route of simply creating a scripting solution to do en masse conversions from WAV to MP3.
Is this possible to have Asterisk emit call recordings as MP3 from the beginning?
There are patents on MP3 encoding that make this a problem for GPLed software, particularly if distributed in the USA.
Normal rate MP3 recordings are larger than toll quality voice, recorded in an appropriate format.
In Asterisk .WAV means GSM in a MS WAV wrapper, which should be about 13.6kbps. That’s smaller than all but the most drastic MP3 compression, and likely to encode speech better than the latter.
I seem to remember that you can specify a script to run when a recording is closed.
If I am going to convert anyway is there a preferred or best way? RIght now I send the Asterisk generated .wav through sox to generate a pcm, which I then send through lame. I’m tweaking settings, but the result is not great.
Thanks again David.
If you really must use MP3, which I suggest you shouldn’t, you want the initial recording to be either 16 bit linear (including Asterisk .wav), or, for a smaller file, the actual codec that was used over the phone. sox can handle G.711.
Yeah, this is looking more like not a great plan. Input->sox->lame->output is not good, and the alternative to switch over to 16 bit linear or G.711…I don’t know how I would convert all the old files. Out of curiosity which .conf would I want to alter to start playing with doing recording as 16 linear or G.711?
Thanks again.
The one in which you enable recording. In our case it is queues.conf, but that may well not be your case.