balancing the vocals: compression / waves vocal rider / manual automation on particularly loud words or phrases

MartinJohn123

New member
I've been mastering a song for a client for most of today. And one thing has continually taken me back to the mixing session: the vocals being too loud on certain words which causes pumping when put in the mastering chain.

I've messed around with different compressor settings to tame it, I've tried waves vocal rider at the end of the chain (as well as at the beginning) and nothing was really doing the job. And then I just went through each word (probably about 10 or so throughout the whole song) and manually reduced the level. Voila. Was it really that simple?

I wanted to share this on here and see how others do it. It's one stumbling block I keep coming up against when mixing.

The next song I work on I will do the following:

- manual automation on particularly loud peaks
- waves vocal rider
- compressor 1 - slow attack, gentle gain reduction
- compressor 2 - fast attack, harder gain reduction on peaks only

I'd love to know what others do.

Martin
 
I use a handful of techniques:
Mix the raw vocal track with a heavily compressed copy of same. The compressed one will pump, but you turn it down so that you only consciously hear the raw one
Drenched reverb channel (turned down pretty quietly). The verb channel smooths off a lot of the volume edges.
Double and triple up the vocal parts. The more takes you have, the more consistent the gestalt will be volume and pitch wise.
 
Oh wow, that's completely different to anything I've tried. Do you have any examples online of songs you've worked on with that chain?
 
[1] A better vocalist with decent mic technique.
[2] Parallel compression (which Steve is talking about) can be a life-saver for out-of-control vocalists (can be a copy but with many DAWs that are sample-accurate, can be done using an aux send also).
[3] Ladies and gentlemen - The LA2A compressor. Savior of countless vocal tracks for decades.
 
Certain words/phrases are naturally going to be lower volume than others - these are ones that require the lips close down (partially or fully) during the pronunciation. I have become pretty deft in drawing in 'points' on the automation curve ahead of the playing line, then I can grab the level and raise/lower as needed when the playback hits that point. If I'm not sure I got it right, I'll rewind to before the section and adjust as necessary. I only use compression to smooth it all out.
 
I usually go through and use automation to pull down the volume on any vocal sections / words that are causing such a problem. I also pull up some that were not well placed at the mic. Sometimes it's as much the tone of the word as it is the volume so you can do some automated EQ adjustment as well. The proximity effect might be too much and pulling down lower frequencies is as good as lowering the volume but retaining clarity sometimes. Spectral editing is also helpful on some words or phrases. I use Reaper BTW.

My 2 cents worth of dribble.

Mick
 
I use a handful of techniques:
Mix the raw vocal track with a heavily compressed copy of same. The compressed one will pump, but you turn it down so that you only consciously hear the raw one
Drenched reverb channel (turned down pretty quietly). The verb channel smooths off a lot of the volume edges.
Double and triple up the vocal parts. The more takes you have, the more consistent the gestalt will be volume and pitch wise.
Yeah. This one uses all those techniques in various combinations


I also tend to stack up a lot of harmonies. As long as your timing is tight, they'll reinforce the lyrics.
 
"Manual automation on particularly loud peaks." There you go. Determine where they are, and use automation. It can be tedious, sure, but well worth it. You have to be careful, but you'll get used to it. As you yourself pointed out, you did it and it worked. Every vocalist is different, and so every vocal is gonna be differently handled. A good vocalist who understands how to work the mic will give you less problems obviously. I don't know how you check your mixes, but I'll always listen at night whilst in bed to the the day's mix (es) on an IPod Nano. It works great--a lot of times its' instant realization like, "aw shit, she's too loud at the end of the second chorus." How did I miss it? I missed it because of ear fatigue, which will happen after long stretches of doing stuff like automation tweaking. Have you thought about sending your mixes out to a mastering engineer after you have everything mixed, including taming the vocal? One less (really important) thing to do. Sure it costs money, but to paraphrase my old employer, that final bit of quality doesn't cost, it pays.
 
The order that things happen is usually clip gain -> Fx -> fader.

It might be easy to overlook, but now that you’ve had this experience, keep in mind clip gain is generally the frontline for these things.
 
"Manual automation on particularly loud peaks." There you go. Determine where they are, and use automation. It can be tedious, sure, but well worth it. You have to be careful, but you'll get used to it. As you yourself pointed out, you did it and it worked. Every vocalist is different, and so every vocal is gonna be differently handled. A good vocalist who understands how to work the mic will give you less problems obviously. I don't know how you check your mixes, but I'll always listen at night whilst in bed to the the day's mix (es) on an IPod Nano. It works great--a lot of times its' instant realization like, "aw shit, she's too loud at the end of the second chorus." How did I miss it? I missed it because of ear fatigue, which will happen after long stretches of doing stuff like automation tweaking. Have you thought about sending your mixes out to a mastering engineer after you have everything mixed, including taming the vocal? One less (really important) thing to do. Sure it costs money, but to paraphrase my old employer, that final bit of quality doesn't cost, it pays.
Thanks for this TimOD, really helpful :)
 
Back
Top