Skip to main content
forums, blogs, song critique, support & more...

compression on vocals

Hello all,
I am tracking some vocals and have been trying to figure out how to set my compression. The singer's volume has lots of fluctuation in it due to his style, so I want to be careful not to over-compress the signal before putting it to disk. This is the setup I am using: (now don't laugh please, this is the best I can afford at the moment...)


  • Shure KSM-27
  • dbx 376 Tube Channel Strip Has a tube preamp, compressor, eq, de-esser,
    digital out (SPDIF)
  • Echo MIA
  • Steinberg Nuendo
    So the question is, how do you guys that know what you are doing set your compressors for vocals?

    Thanks in advance,
    Vlad Khavin

  • Comments

    Profile picture for user anonymous

    anonymous Tue, 11/27/2001 - 08:05

    Since you are always able to add more compression in the mix stage, and since you aren't working with a box that is going to add a lot of desireable color (like an LA2A), I'd go as light as possible on the tracking compression - just enough to keep you from getting digital overs. I don't know the dbx box specifically, but if there is gain reduction metering, my guess would be you'd want to keep it at or under 7db. Start with maybe a 2.5 to 1 ratio and gradually lower the threshold until you see that amount of gain reduction. (Hopefully that will still be a reasonably transparent sound.) Otherwise you may be stuck with a squashed sounding vocal that you may not like later. If there is an "auto' mode for attack and release times, you may want to try that first. If you haven't had a lot of experience tweaking compressors, sometimes that gives you the best results. Of course, as always, your ears are the ultimate judge, not the knobs or meters.

    etheriagtr Tue, 11/27/2001 - 10:42

    Thank you much, littledog. I'll deffinetly try that out. I don't think the compressor on the box has anything but an auto mode, and the only time I've ever played with a comressor before was in my guitar rig so, any sort of input is much appreciated.

    Thanks again,

    Ang1970 Tue, 11/27/2001 - 11:29

    If the dynamic range of the singer is that drastic, I would avoid compression altogether. Maybe a limiter so you don't clip, as littledog suggested. Once you get the track into Nuendo, automate it to whatever degree is neccesary. (a.k.a. "Vocal ride") Maybe then send the automated signal to a bus for compression, if it needs it. (Or can fx be inserted post-fader in Nuendo?)

    etheriagtr Tue, 11/27/2001 - 11:41


    To my knowledge, all inserts in Nuendo are Pre-fader. You can set up sends to be post or pre but not inserts, so I would have to do as you suggested and send the output of the track to a bus with the compressor as the insert of the bus. The limiting idea sounds great too. The complete project is not going to be mixed in Nuendo, I am just tracking it and then going to dump the tracks to a real studio's DA88s. (We recorded all the music at TraxEast, but wanted to cut the vocals ourselves because there was way too much experimentation involved for our budget) The whole reason I am asking is that I wanted to get the best signal I could into those tracks, because I think the less doctoring we have to do in the studio to compensate for something I missed while tracking, the better the final mix will sound. Thank you very much for the suggestion, it all makes a whole lot of sence now that you mention it.

    Thank you,

    Ang1970 Tue, 11/27/2001 - 12:06

    Cool... One more suggestion... If you are cutting takes in sections as opposed to full passes, keep readjusting the gain to get the hottest level on each take. That will make the vocal rides that much easier, and get the hottest level to "tape" for the whole performance, not just for the loudest part. Try to do it quickly so as not to disrupt the flow of the session. With most r&b type songs I'll give a rather hot level until the second chorus, back off one notch, back off another notch at the bridge, and another 2 notches before the last ("bringing it home") line of the bridge and leave it that way thru the tag. You know, match the level to the dynamic of the song.

    Be sure to do your vocal rides at home and dump them to tape that way, your mix engineer will thank you. Also dump it to another track without automation, just in case.

    Cheers :)

    etheriagtr Tue, 11/27/2001 - 19:52

    Wow, I NEVER thought of that! (Changing the gain mid-session) I will most deffinetly try it the next time I cut tracks. Thanks again Ang1970!

    So what you are saying is that you lower the over-all level of the vocals to match the dynamics of the song, or are you just lowering the gain on your preamp and then compensating for the lowered gain by doing the volume ride in the mix to keep the level consistent?


    Ang1970 Wed, 11/28/2001 - 06:48

    A little of both. If the singer is louder, cutting the gain back will make it level with the previous section. If you get the right gain on each part, the vocal will be more level thru the whole track, and then the ride will be much less drastic and easier to do.

    RobinH Wed, 11/28/2001 - 13:59

    Biulding on what Ang said you can ride the fader to tape as well. If you have a visual with the vocalist you can often tell when they are going to give it some. After a few practice runs you can judge your fader riding.

    planet red Wed, 11/28/2001 - 17:38

    Why not just mess around at home and figure out exactly what you want to do vocally and go back into trax east and record it. If you're already spending all that money in trax east i wouldnt take the chance of messing anything up and they have way better stuff than your shure and dbx pre. By the way is steve evetts recording you guys? I was up there one time... pretty cool place.

    etheriagtr Thu, 11/29/2001 - 12:02

    planet red,
    Steve doesn't work at TraxEast any more, unfortunately. Steve has a hell of a reputation around here. He went out on his own, so we are recording with Eric Rachel. Six months ago, going back to Trax East would have been the only option worth considering, but at this point, we pretty much have to get the thing done. The gear I have seems to be giving me a nice full sound and we should have no trouble working with it later on in the studio. I considered posting the reasoning behind this move of ours, but it's a long story. In a nutshell, we started with a different engineer that screwed us over, and now that Eric Rachel is working with us, we have to get this album completed as soon as possible. Not to mention, there is no better incentive to learn how to do something right than when it's your own project, so it's also a great learning experience for me. If we record at home and it won't work, then we get to rerun the tracks in the studio, but I have faith...


    PS: If you actually want to hear the whole story, let me know and I'll post it.

    MadMoose Thu, 11/29/2001 - 21:08

    Originally posted by etheriagtr:

    PS: If you actually want to hear the whole story, let me know and I'll post it.

    I wouldn't mind hearing it. I've got a whole bunch of horror stories to trade.

    etheriagtr Fri, 11/30/2001 - 09:21

    Well, ok. This is not so much a horror story as it is just the story of an album gone wrong. Here goes...

    We (Etheria, my band) started recording our debut album "Genesis" at the end of October, 2000. We arrived at Trax East, expecting to be working with either Steve Evetts or Eric Rachel. Instead we found out that our engineer was going to be Eric Kvortec, Steve's replacement. Eric seemed like a nice enough guy, was easy to work with for us, and we tracked all of the music with him, and were reasonably happy. And then it was time to do vocals...

    Our singer and Eric did not see eye-to-eye on alot of things... Like his singing style, and his perfectionism as far as a part being delivered with the correct feel, not only correct pitch. The disaggreements were not the real problem tho. Oppinions are like assholes, everyone's got one, and I completely respect that. The problem started when Eric started acting completely unprofessionally. Several sessions went by where half the day was spent arguing with our singer and telling him that everything he was doing was all wrong, and the rest of the time was spent either having his feet on the console, (which is technically his right if he wants to treat the console that way) or bitching about the argument he just had with our singer to our manager. Either way, the guy was not paying attention to what he was doing. Now I can make allowances for someone having a bad day, but this happend several times. I understand that our singer can be an extremely difficult person to deal with, but that was uncalled for. One of these times, he actually neglected to stop the tape after the song was done because he was too busy bitching or daydreaming, and erased part of the next song on the reel. It took alot of restraint on the part of our keyboard player not to kill him for that (it was keybrd player's baby that erased on him...)

    So after a few VERY unhappy and unproductive sessions, we decided to talk to Eric Rachel, who owns the studio. We told him what's been going on, and Mr. Rachel (from here on to be called Eric) said that he would now be working with us and he would make it up to us. Eric Rachel is an incredible engineer, and a hell of a guy. We had a deadline by which we needed 3 songs to be complete so that we may use them for a demo. Eric did not have time in his own studio, so he got us in touch with Mike Romeo of Symphony X. Mike had some awesome equipment and was kind enough to record vocals for three of the seven songs on the album for us. Symphony X went on tour, and we had a few options as to how we would finish the album.

    We spoke to Eric Rachel for some time and he said that he could give us mono mixes of the music with SMPTE code on the other channel and we could record the vocals on my setup at home. The schedule at Trax East has always been insanely busy because it's one of the best studios in the area and is very reasonably priced compared to what I've seen. We would wind up stretching it out over several more months of tracking if we did it at Trax East and we just don't have the luxury of time anymore, nor do we have much left of the budget between having to retrack a few things in the studio. He told us that as long as I can give him a decent signal, he can match it to what came out of Romeo's house and we will be alright.

    At this point, it's been over a year and this album is way overdue. I have always wanted to learn to record, and this seems like a golden oppartunity for me. Not to mention, since it is my band's album, it's extra incentive to do a good job.

    A while ago, my singer bought a mic and a preamp (CAD M37, and JoeMeek VC3Q), and we didn't like the sound we got, so I did some homework, and bought the best that I could afford. the setup seems to give a good sound, and I went from Cakewalk to Cubase to Nuendo and finally got what I wanted out of the software and the hardware. Now it's just getting the vocals down, and making sure that what I think is good, is actually good.

    So this is the way the story goes, and in the end, I get to learn to do something that i've always wanted to learn, but until recently has been a complete mystery to me. Perhaps I can post some of the stuff that I tracked and you guys could give me some feedback and pointers if I'm doing something wrong.