Processing Vocals Part 5 – Compression

Ah, compression. It can be a great tool, and it can be easily overdone. However, I can’t imagine mixing a song without using compression on the lead vocal. It both tightens up the vocal and helps it fit into the mix.

If you’re not all that clear on what compression does and how it works, I’d recommend watching my Intro to Compression video first. It’s a pretty succinct overview of compression in general.

Alright, assuming you have a basic understanding of compression, let’s look at how it applies to vocals.

Determine what you’re goal is with compression.

Take some time to listen to the dry vocal in the mix. What is it lacking? What does it need? You know what compression can do, how can you use that to your benefit? Be patient. You need to have a plan before you start turning compressor knobs. Otherwise, you’ll end up knee-deep in compression that doesn’t make sense or even sound right.

Control the Peaks

Does the vocalist sing at a consistent level or is he/she all over the place? For doing strictly level control, I would do the following:

  • Set the threshold just below where the louder portions happen.
  • Use a relatively high ratio, maybe 4:1.
  • Keep the attack under 10 ms

With these settings, the compressor will turn down these louder parts and provide a more consistent level across the vocal performance. Important Note: Compression shouldn’t take the place of automation. Even with a perfect compression setting, you should still “ride the fader” on the lead vocal to dial in the perfect level throughout the song.

Lightly Compress Everything

Perhaps the vocal performance is pretty consistent. If your goal is to use compression to alter the tone a bit, maybe make it sound “tighter,” then try the following:

  • Set the threshold well below the average level of the vocal. The compressor should be compressing whenever the vocalists sings.
  • Obviously this will sound overly compressed at first. To combat this, turn the ratio down considerably, perhaps 1.5:1.
  • Use a longer attack.

With these settings, the vocal is always being compressed. But since the ratio is so low, it’s a softer compression. On the quiet parts, the vocal will still have a little bit of a “compressed” sound. As the vocal gets louder, the amount of compression will increase.

This is my favorite way to compress a vocal. To my ears the compression sounds smoother. As the vocal gets louder, you don’t hear the compression “turn on.” It’s always on, it just compresses more and more the louder the vocal gets.

No matter which method you use (I’ve actually seen engineers use both on the same track), you want to keep an eye on the GR (gain reduction) meter on your compressor. If you’re knocking off over 10 dB with compression, you’re probably overdoing it.

Parallel Processing

I’ve posted a couple videos on parallel processing. The idea is that you duplicate the track, then process each copy differently. This can work wonders on a lead vocal. Perhaps you overly compress the second track and blend it in with the first. Or maybe you need to add a little distortion to the second one and blend it in.

Doubling the number of vocal tracks certainly doubles the complexity, but this is something to keep in mind. Check out the video on parallel processing bass to get an idea for what I’m talking about.

De-Esser

Since compression turns down loud parts and turns up quiet parts, you may have the PERFECT compression setting, only to find that all the vocalists S’s are just too loud.

That’s why God created the de-esser. I’ve not done a video on de-essing (hmm…that’s not a bad idea), so here’s the basic run-down. De-essing allows you to compress a very specific range of frequencies without affecting the rest of the audio.

For my voice, the sibilance sounds occur around 7 kHz. When needed, I set a de-esser to that frequency, then have it knock down the S’s when they come through. It takes some practice, but it can really help keep the S’s from cutting your head off…a nasty bi-product of compression.

How do YOU like to compress vocals? Let’s hear it! Share your tips by leaving a comment.

Other Articles in This Series

  • Chris

    Sorry to be a grammar nazi but i think you meant your not “youre”

    haha besides good tips. Ive been trying to figure out compression and this makes it a bit easier

  • Sean

    Hello Joe,

    Your site is truly amazing. Thank you so much. I am only just beginning to get au fait with compression. I have produced my first five songs, and am gradually improving. My studio is basically a Boss BR800 – an 8 track digital recorder. It comes with a compression “plug-in” but I have absolutely no idea how loud the input of my track is, and where the threshold is kicking in(there are no units on the threshold setting – simply a setting of 0-100). As such I feel like I’m working blind. The louder I make the threshold the louder it makes the whole track, implying that the whole thing is above the compression threshold? But again a digital multitracker has no meters. Is the only solution to buy an external compressor with meters so I can see what is going on?
    Thank you again!
    Cheers, Sean (Cape Town, South Africa)

    • http://www.homestudiocorner.com Joe Gilder

      Hi Sean! Glad you like the site.

      It’s hard to say what’s going on on the BR. Does it have all of the normal compression settings? Attach, release, threshold, ratio, output/makeup gain? If not, it may be doing some of the things “for you,” like makeup gain. I would just spend a lot of time with it, and do what sounds best. If it doesn’t give you much to tweak other than the threshold, then really spend a lot of time dialing that in. Meters are nice, but you can still get great sounds without them. It just takes a lot of time and practice.

      Keep working at it!

  • http://www.homestudiocorner.com Joe Gilder

    I think you should get away from measuring things in dB. Meters aren’t musical. Your ears are. :)

  • http://www.homestudiocorner.com Joe Gilder

    That’s really hard to answer. I wouldn’t base it on what the meter is doing. Compress it until it sounds right to you. And don’t forget that fader automation can be a much more effective way to get a balanced vocal than applying too much compression.

  • http://www.dirtysouthkingpenrecords.com Drew Murret

    What up man, just letting you know I stumbled here and figured I’d give the compression part especially a thumbs up.. It’s been years since I’ve taken Audio Engineering but I record a lot, and it’s WAY too easy to get in your own habits over the years and just keep kickin it like you always do..
    Sure, it’s good to have a style, more importantly to know what works for YOU.. but it’s not good to get stuck in a single mode… n jus forget that life should be a progression and learning process… always.

    I totally agree with your preferences of less compression ratio, lower threshold, and higher attack.. good way to go on most recordings in general, in my opinion.

    Personally should probably get in a better habit of using the de-esser more often.. As for EQ.. I do mostly underground trip-hop / hip-hop and am a bass “FREQ” so I don’t do as much cutting on the low end usually but that’s a given for the style. Things in general are a bit more compressed to tighten, as opposed to EQed to keep the dynamics.

    Anyway, just wanted to shout at ya- gunna go post this on my facebook for a few others.. if you’re into any Trip-hop type Electronic/hip-hop type shit, check my site out @ http://www.dirtysouthkingpenrecords.com (it’s alternative/underground style electronica hip hop… not gangster rap or anythin)

    More abstract than the site probably sounds but ay, that’s just the type of dude I am.
    Be easy fella

    • http://www.homestudiocorner.com Joe Gilder

      Welcome Drew! Always nice to meet another engineer. :-)

  • Anthony campagna

    Hey Joe,

    I haven’t heard a mention of this, but I like to put the desser in front of my vocal plugs or sends, it tends to warm things up going into the verbs/delays where the ss can get out of control quick:0 This allows me to keep more of the sibs in the main track with out becoming a thorn in the earball:)

    Out
    T

    • http://www.homestudiocorner.com Joe Gilder

      :-) Great point, Anthony. There’s nothing quite as annoying as a super-sibilant reverb.

  • Maria Ayerbe

    Hi Joe, great great posts! I was wondering if you could explain how sidechain plugins work now that Jon mentions them.

    Thank you so much!

    • http://www.homestudiocorner.com Joe Gilder

      Hi Maria! Sean over at KeyofGrey.com has a great post on side-chain compression actually. Here’s the link: http://www.keyofgrey.com/?p=2789

      • Maria Ayerbe

        Wow this was fast and really useful!! Thanks again!

  • http://www.fruittreemusic.com jon arnold

    Hey man,

    Another great post.

    I tend to use your “light compression” model above, but I find that the way I do it takes off about the same amount as a standard approach.

    If I set it for -20 threshold, 1.2-1.4 ratio, it normally takes 4-6 dB off. (Which for me is perfect on a normal pop/rock vocal)

    To your point, it’s always “on”, which not only saves the hassle of a breathing compressor, but you don’t have to worry about fluctuating tone. The compressor’s coloration is always effecting the vocal, meaning you don’t have weird changes during softer passages.

    Regarding de-essers, have you done much with sidechain compression? I find that putting a cheap EQ (or simple plugin) in the sidechain, set to the sibilant frequency, helps a lot as well.

    Hope you’re well!

    Jon

  • Mac88s

    So Joe, do you put the de-esser after the compresor. I’ve seen other demonstrations where they put EQ-De-esser-compressor. What do you think?

    • http://www.homestudiocorner.com Joe Gilder

      I tend to put it after the compressor, since the compressor tends to exaggerate the sibilance. It seems to make it easier to de-ess the signal.

      It depends on the track though. Sometimes it needs to be de-essed before the compressor.

  • Martin

    Hi Joe!

    I was just recording and mixing my vocals for a new song of my band as I visited your great website once again and just when I needed some good advice, you posted this thing on vocal processing! So first of all: thank you!
    I always had problems with getting an “even” sound and volume into my vocals because of the multiple takes I made and I was pretty helpless with EQing my vocals the right way. In fact, I find processing and mixing my own vocals is always much more difficult (and sometimes annoying) than processing anybody elses. What do you think?

    I probably just recognized a little mistake you made at part 5, “lightly compress everything”. At the second point I think you meant “ratio” instead of “threshold”.

    Kind Regards,
    Martin

    • http://www.homestudiocorner.com Joe Gilder

      First of all, thanks for catching the mistake, Martin! Just fixed it.

      As far as evening things out between multiple takes, compression can help, but you’ll really want to actually change the volume of the region itself. If one line is much louder than the next line from a different take, then highlight the louder take and use a Gain plug-in to turn it down. This will make your life much easier.

      • Martin

        Okay! I’m going to try that. Thank you!