View Single Post
Old 02-26-2015, 08:57 AM   #4
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 7,272
Default

I've been using a technique lately for a very transparent kind of volume leveling. It works well for vocals (a lot like that VocalRider plugin, I think), but I also do it right on the Master to get a bit more "loudness" without losing so much transient information. The trick is to set the RMS window pretty big, but then set the pre-comp to somewhere around half the RMS window. This way, the compressor is reacting to the average level around the current sample - both past and future - so that if one bear is significantly louder than the next, it sort of gently dips the volume leading into the louder hit, and then comes back up afterwards in a pretty natural way. Attack and release are basically built into the process, so I'll set those parameters to 0, and I usually use a very low ratio - 1.1 or 1.2. Then, for even greater transparency, I'll usually set the threshold so that it just barely starts squishing on some of the louder parts, and then crank the knee where up so that it's almost always doing just a tiny bit, but it's almost always in the knee, and almost never hits the full ratio. I sometimes try to calculate the time constants to cover a beat or two of the song's tempo, but I've found that it's not super critical, as long as it's pretty long. Pre-comp only goes to 250ms, I think, so that limits things a little. All of the really fast transients still get through with the same proportional relationship compared to what's going on around them, but the overall dynamic range is reduced very transparently.


Edit - it's worth at least mentioning that the RMS level of anything that's not a square wave is always less than its peak level. At least to a certain point, the longer you set the RMS window, the lower the level that hits the detector and shows on the meter, and the lower the threshold needed to get the same amount of gain reduction. It can be kind of tough to A/B different RMS sizes, or even say for sure where any perceived differences are coming from.

Last edited by ashcat_lt; 02-26-2015 at 09:16 AM.
ashcat_lt is offline   Reply With Quote