MOH (Music-on-Hold) quality and formats


I have recently been hired to write MOH for a number of large companies. I’m a pro composer/producer and have through the years worked with variety of formats for different platforms. This, however, is new to me and I was hoping someone in this forum could help clear up some issues for me.

I’ve done some research about the different formats for the asterisk platform and basically what I need to know is, what is the highest possible audible quality I should mix for. And which applications would you recommend for doing so,- I have found that Audacity can do most formats. If it has got anything to say the version of asterisk that is being used is

So far all I can find is
-wave (waw)
-ulaw (alaw or ulaw whats the difference?)

I hope some of you out there, can help bring some light on these issues.



If you are going to play the music on hold over the PSTN, there is no point in going any better than µ-Law or A-Law, µ-Law for the Americas and Japan and A-Law for Europe, and probably the rest of the world. These are companded, 8 bit, mono, 8kHz sampling rate, formats. The cost of converting from Windows .wav 8kHz, 16 bit, signed, mono is quite small, so this is probably the best format to store in, but you should be aware that what the caller hears will be lower quality. (International calls may get transcoded between µ- and A-Law, further degrading them.) You should not have significant audio power above 3.4kHz, as it not get through. Nor should you rely on anything below 300Hz.

If you expect any significant use of mobile phones, the audio will be degraded to GSM quality (Asterisk .WAV or .wav49, with Windows .wav wrappers, or .gsm, raw). This is based on vocoders and is only intended for human speech, so you will have to experiment, or research specialist literature on composing for such formats. If Asterisk is directly using high compression formats, like GSM, the cost of converting on the fly is high, so you should provide versions in the alternative formats. Asterisk will use the format that is cheapest to convert.

Some people use higher audio qualities on internal networks, particularly I suspect for senior executives’ phones. You would need to check specific details.