gently hard limiting or using compression can change the sound entirely.
Over-doing it can, sure... but I've never had an entire file be adversely changed in its relative dynamics, or have the sound ruined, by light limiting and/or compression to simply tame a few erratic peaks. I *could* make that happen on purpose if I wanted to, but you don't squash the good -- only the bad. If the entire file is noticeably changed aurally I'd say the person doing the processing doesn't know what they're doing. Same as running an outboard compressor -- if you don't know how to manage the parameters then there will certainly be an unpleasant outcome.
Moreover, if you were to manually locate the peaks and simply reduce their volume, what do think is going to happen?
All the highlighted part will be reduced in volume, from the topmost sample of that bad cymbal crash or string pluck all the way to the bottom -- including all the instruments and vocals mixed in that you
wouldn't want brought down.
Regularly I have to process files of live performances where all I get is a final 2 channel live stereo mix. Often, there are some loud trumpet parts that overshadow the rest of the music and they have to be brought under control. I can't remix anything because there was no multitrack session to begin with, so I split the file into bands. The upper band is from 2K on up. In that file is the bulk of the trumpet sound, easily seen because in the waveform its peaks are way out of line with everything else. Correctly compressing and/or limiting those peaks brings the trumpet volume down to a more acceptable level. Once the bands are remixed, the trumpets are at a more tolerable level, better balanced with everything else -- without adversely affecting the vocals and instruments.
Author
