I am planning to analyze audio via linux command line. There are a lot of analyzers out there which have a graphical interface, but since I want to make an automated input/output of this information I can't use a GUI.
The optimal case would be to send an audiostream from an embedded device to powerful machine, which analyzes the audio and then generates a csv or db entries with the data and builds some graphs out of it for a website.
If there so no such thing for audiostreams, an analyzer for audiofiles will also be a great step forward.