>>>>>We have a zillion WAV files that have both voice converstation and DTMF tones in it. Obviously, WAV files are not the smallest sound files - so I'd like to convert the WAV files to some other format. I haven't messed with it much, the problem is the gizmo I use to pull the DTMF tones out of the audio files seems to fail when I convert (compress) the WAV files.
>>>>>Anyone have any sugestions on a CODEC I should try that will keep my DTMF tones?
>>>>
>>>>Have you tried converting the WAV files to MP3? I think you might have a hard time compressing the WAV files and keep the DTMF tones. Compression usually means some type of loss of information. If I recall correctly DTMF tones are sine waves and your "gizmo" is probably pretty sensitive i.e. high sampling rate.
>>>>
>>>>Do you have any specs for your "gizmo"?
>>>
>>>Well the gizmo is just an activex control that reads the DTMF tones...I didn't write that part of the app and I forget now which activex control was used (I can find out though) - but we can use whatever we want to read these DTMF tones. I tried various bitrates and such with MP3's, but it just doesn't seem to work very well.
>>>I would think there would be an audio CODEC specificly designed for what I'm trying to do here...one that does compression (or at least gives me smaller files) that doesnt kill my DTMF tones....
>>
>>How big are these WAV files? Unless you are calling to a foreign country with an insanely long extension number the files should be relatively small right?
>>The Windows Media Format (WMF) might work for you but I don't think the compression is very good).
>
>The WAV files are recording both voice converstation AND DTMF tones in the same file. Some of these phone coverstations are quite long so the files can get pretty large...hense our problem.
>There are things like "press 1 if you want to ...blah blah blah"....and I have to keep the whole file with the DTMF tones together for legal reasons.

I think Einar's on the right track... IIRC analog phone audio has a frequency range of 200Hz to 4KHz rather than the full 20Hz to 20KHz of music/full range human hearing. So if you could tell whatever codec you use to filter out anything below 200Hz and above 4KHz you might get substantial file size reduction without noticeable, or any, loss of quality.

Then, you might be able to apply a smart compression codec to the result for even smaller files.