When building Asterisk, there is two different entities related to audio formats - formats and codecs. What's difference?
For example, I need a support only for alaw+ulaw+g.729 on voice traffic itself, and only plain wav + mp3 for announces and other sounds played to subscribers. What modules should I enable?
Thanks.
format_ are modules for working with mediafiles, and codec_* are modules for transcoding human speech.
In general, if we're talking modules, you should enable all codec_ and format_ modules that you have (and then you will enforce the needed codecs via peer configs), because when not used - they basically do nothing. The modules that you shouldn't enable in a bunch are various app_, cel_, cdr_, pbx_ and res_ that you don't use (because these are doing much of unneeded stuff when loaded but not enabled); but you want to have the widest compatibility with incoming/outgoing speech codecs along with on-disk mediafiles formants that you can provide.
