Ive been experimenting with gain automation for my mix, its sort of a necessity for the music Im mixing - not because of major recording problems, but simply to get as much dynamic range out of the music and have it as natural as possible. Ive read a couple times andy wallace does this, along with a few other engineers, but think he does it after compression.. My question is this - how much is too much? Im finding the dynamic range Im creating by the end could be too much, after all - If your automatimating a volume swell of roughly 6 dbs per instrument for a particularly big section - is this going to be too much? (prioritising the vocals obviously) Whats the minimum output level for an average song thats recommended? Or even a guideline db range. I know, generally for mixing before mastering, youre best bet is to leave -3dbs of headroom. Im really influenced by buckleys grace in this area - the massive amount of dynamic in mojo pin is really my aim. I do want the result to be that theres a lot more dynamic in the music than most music currently out there. Any help or advice or tips would be really appreciated!