First off, you're alluding to Layer Markers and not Composition Markers. What you want is very achievable but you'll need Expressins or even Soundkeys from Trapcode. It helps if you can prep your audio file.
But honestly, since layer marker creation is done in realtime, I don't see how you will want to speed this up. Further, you will still need to check on the results once you're done. An automated process that is lined up against an audio file that isn't suitably setup for the job may result in errors further down the production line.
FWIW, Trapcode's Soundkeys is able to discern audio frequencies as well as amplitude.
HTH
RoRK
Sell your AEPs with broadcastGEMs' DVD series of templates. Click
here for more
