Everyone's favorite topic - compression for vocals

[andrushkiwt]

I am usually editing/copying/pasting from 6-7 vocal takes and, occasionally, my position changes or I sing slightly louder/quieter in one phrase than the previous/next line of a separate take. When I select the "best takes" of each lines (yes, all the way down to lines for me, not whole takes) the volume changes slightly from one line to the next. (I am not touching the gain). I have received a ton of positive comments on the vocal quality of my tracks, but I am really struggling to understand what i am actually doing sometimes.

I use multiple compressors on the vocals but I have a tough time with their settings. My vocal tracks will typically run through this buss chain, in order: EQ, de-ess, comp, verb, delay, comp. Then, I have a "send" compressor that I link some vocal tracks to. That's 3 compressors not including the de-ess. The problem is that I am STILL having to do a ton of manual volume editing and some words pop out too much. I am guessing that my compressor settings are not optimal.

I usually have the first comp at 4:1 with a very low threshold (-30) and keep the gain knob on the compressor at "auto". Should I turn any and all gain adjustments to 0? The second compressor is usually 6:1 with a slightly higher threshold (-25) and the "send" compressor is about 2:1 with a higher threshold.

Help me out here. I love the flavor and sound I get from these compressors, but help me get the balance right in settings so that I don't have to rely on editing volumes by hand so much.

let me know if i left anything out that's relevant. thnx

 

[andrushkiwt]

crap - wrong sub-forum. moving it now, sorry

 

[mixsit]

Use gain or clip automation before going into the compressor to do most of the bigger leveling.
The compress has a more consistent track to respond to, has less 'fixing to do with it's static setting', and it can sound more natural.

 

[Jake_JW]

Same as mixsit said I normally sort levels using clip gain, trying to get the sections of the track mostly in line before hitting compression, then automating track/buss volume later to sort the levels in the final mix.

I normally stay away from 'auto' settings for makeup gain - not sure why or if it's just me but I always find it works out better (for my workflow, at least) to level parts in matching the pre/post compression volume by ear (percieved volume, not absolute level). Sometimes an 'auto' setting makes the output seem louder (adding more gain that I would manually to compensate for the GR of the compressor) so it can skew my judgement on wether the compressor is doing what I want it to by upping the percieved volume...

 

[mixsit]

... I normally stay away from 'auto' settings for makeup gain - not sure why or if it's just me but I always find it works out better (for my workflow, at least) to level parts in matching the pre/post compression volume by ear (percieved volume, not absolute level). Sometimes an 'auto' setting makes the output seem louder (adding more gain that I would manually to compensate for the GR of the compressor) so it can skew my judgement on wether the compressor is doing what I want it to by upping the percieved volume...

This exactly. And I'd offer rather than sometimes' -more often than not. FWIW I'd prefer that plugs drop in transparently. These 'global setting guesses add another layer of complications to undo.

 

[Farview]

It sounds to me that you are doing too much processing.

Try putting a compressor on each vocal track. Send all the vocal tracks to a buss and compress that. Put the de-esser first in the chain, before the EQ and compressor.

Use the clip gain to get all the phrases about the same level.

Know that you are also fighting proximity effect. When you are closer to the mic, the sound will be thicker and more in-your-face. When you are farther from the mic the sound will be thinner. The takes where you are closer to the mic might stand out even if they are the same volume or quieter than the ones sung from farther away.