Thursday, October 21, 2010

What's in Your Vocal Chain?

I thought today it might be fun to compare our favorite effects chains for using on vocals. Every track obviously differs according to what you need, but most of us probably have some basics we use fairly regularly. I'll start, and you can feel free to join the discussion in the comments.

I record my vocals on an AKG C-3000 which I've had forever. It's nothing fancy, but I think it sounds pretty good with vocals, as it seems to have a little bit of a mid-range bias, and this can be helpful in getting your vocals to sit right in the mix. The down side is, if you've got a singer with a lisp, or just a sort of "spitty" way of singing, the mic will exaggerate it.

The mic is plugged directly into the XLR input on my MOTU 828 mk II. I'd love to have a fancy, high-end recording strip for stuff like this some day, but for now, this does the job just fine. Very neutral sounding, so it gives you a good starting point.

Once I've recorded several takes, the first thing I do is to build a comp of the best bits from each take. This way, I'm starting with the best results possible. Now's also a good time to listen for mic bumps, mouse farts, or other unwanted sounds that need to be edited out.

The first item in my typical vocal channel is an EQ. The sole purpose of this EQ is to filter out the unneeded low frequencies. The reason it's first in my chain is that I want to get rid of them right away, so they don't effect the way the compressor works. Having an EQ cutting out the lows is really a good habit to get into for almost any channel (aside from bass and kicks, although even they can use a cut around 40Hz and below). It will clean up the low end in your mixes and free up all sorts of headroom in your mixes.

Next up, I have my compressor. Even though I have a LiquidMix, which has models of classic analog compressors, I actually prefer Logic's built-in compressor on my vocals. It's a very clean and neutral sounding compressor and really lets vocals come through clearly. If I need "character", I can always add it later. I tend to compress my vocals pretty heavily... a ratio between 6:1 to 10:1 with an aggressive amount of reduction and relatively fast attack. There really is no magical setting that will work for every voice. You just need to spend some time experimenting to find what works the best with your vocals.

I follow the compressor with another instance of EQ. Unlike the previous EQ which consisted of a heavy low-end cut, this one is aimed at boosting a couple of key areas. I'm putting it after the compressor because I want to exaggerate these frequencies in the already-compressed signal. I find this allows for more subtle results and a greater ability to fine-tune easily. My typical boosts are around 1500Hz - this is an area often associated with intelligibility of vocals, and can help fit a vocal "in" the mix instead of on top of it - and at 10k, which adds a little "air" and presence. Neither of these boosts are more than a couple db's a piece.

The second EQ is followed by Logic's Sample Delay plug in, which is designed to allow you to offset the left and right channels of a track at the sample level. When you delay one side versus the other, the result is a stereo "widening" effect. I think this makes the vocals sound a lot slicker and more professional, but it also helps the vocal fit in the mix better. It might not be right for every case, but give it a try and see if you like it.

Finally, I have to aux sends, 1 which goes to a reverb, and another that goes into a delay. I'll use different ones for different purposes, but I always keep my send levels pretty modest... about 20% or so. This helps give the vocals space, but without overwhelming them. Vocals are generally mixed pretty dry these days, so mess around and find what sounds right in the context of your song.

That's it. Pretty simple. How about you? What's your go-to effects chain for mixing vocals?


Paul Seegers said...

I use a Rode NT-1 into a N.I Kontrol right into my duo core PC running Cubase SX3 or SX5.
The Rode has a definite sound with a bit of sparkle that either you like, or you don't, if you don't want the bit of high end character, well as our friends in the old German war movies say "Ve have our veys."
One way is Antares Mic Modeler, which is a fancy eq that can give you good e.q characteristics of other mics. Another route is to have the mic mod-ed by a guy on the Net who will take your NT-1, put a Chinese U67 and adjust the wire basket and give you a very close Nuemann sound, for about $300.
I record the vox flat and dry with no compression (May ride input faders with a wild singer or rapper) then I make all the decisions about compression and e.q. I am comp a few vocal passes togeather for the best bits, or just make the vocalist "do it right". Some people really want to "do it right" because they feel they have to pull it off live.
After the vox is recored, I often use the excellent UAD-LA2A line leveler. (I use it on all kinds of material) or the UAD-1176 compressor. The fist is somewhat transparent, the second
has "a sound", it just all depends on what I am going for. I am also a big fan of the Waves SSL channel strip (I like the color it adds to everything). I e.q to taste and have Reverbs and Delays on aux sends.
One more bit of kit I like is the Vintage Warmer plug. Great to warm up vox with, and can be used and abused on a lot of different material(guitar, vox, drums, bass,synths)

Anonymous said...

Hello, I usually use a Shure SM7 for the vocals as it has an updated broadcast vibe to it that is dense but organized like the RE-20 of olde. Lately we've been using a Nice Pair for the preamp. It's fast like an API but FET and has a larger tonal range between the input/output gain. If that pre is too fast sounding i use an old NEVE 22415 that we racked up a long time ago. Our main engineer usually prefers a Distressor next for compression ( i'm not so much a fan but he gets it to sound very good ). I would tend towards the GYRAF SSL Clone, you can use just one side with the stereo controls and it can do a nice up front but under control vibe that blends well with layered vocals. The coolness of that design is in the side chain mode mods. thanks for blogging! I read regularly! z

Anonymous said...

This is a great post.

I use the Rode NT-1 as well and think it does add a character that's quite good for it's price range. When I worked at commercial studios I used to go nuts with cascading different compressors together for vocals. LA 3A into LA 2A into a Distressor? Sure, why not?

Nowadays I just use a little bit of FMR Audio's RNC at a modest setting with the Really Nice button engaged just to tame the peaks so it sounds as transparent as possible.

Then yeah, high pass filter judiciously, some compression from The Glue VST at a pretty drastic setting, post comp EQ, and a little bit of drive from Wave Arts Tube Saturator.

I use that last plug in varying amounts on every track. Normally you can't hear the difference with it on or off if you solo the track but I've found that a little bit of drive from that component modeled plug in on everything increases the perceived volume of the whole song. Pretty awesome.


krell said...

Rode NT1-A going direct into the Apogee Duet. Fairly flat EQ from the standard Logic Channel EQ mainly to kill the bottom end. Logic's sample delay is next with only a very subtle setting. I have two sends, a touch of grit from the D-16 Devastor and some reverb from Eos.
Finally a bit of compression from Logic's compressor. Simple but does the job.

nulldevice said...

I mix it up a little bit, depending on the track. Used to use a Rode NT-2, then swapped it out for a Oktavamod 219PE which had a much less "hyped"upper midrange (my unfortunately nasal voice did not benefit from the midrange bump of a lot of modern condensers). I've recently started using a GT Velo8 ribbon mic which has been pretty awesome.

Run that into an MetricHalo ULN/2 (I upgraded from the 828 Mk1 to the ULN/2 a few years back and couldn't be happier), and then usually one of the PSP sQuad EQ's (they're awesome. So awesome) then logic's comp. Sometimes I put a little delay or verb on it. If I'm feeling like my vibrato was a little too much I'll add pitchcorrect up at the head of the chain.

Seems kind of like overkill for my occasionally amateurish attempts at music, but it works. And it's fun!

-sihiL said...

I record my screams, shouts and grunts with a JTS NX8, a copy of the SM58. It's plugged into a Presonus FP-10. The first thing on my channel strip is Logic's EQ, I cut out just about everything below 300Hz and boost the 3kHz range at a wide Q-setting for a couple of dB to get more air, presence and clarity to the sound. After the EQ I have Logic's Pedalboard with just the Treble Boost, works awesomely with such a dark-sounding mic. I bash in about 20-23dB of gain. After that, a chorus at a moderate setting. On the sends I have Logic's pitch shifter at 50cents set on "invert", to widen up the sound even more, I tend to keep it pretty dry. I also use a reverb and a stereo-delay on the sends. The delay and the pitch shifter send into the reverb as well as the main vocal track. I tend keep the reverb and the delay pretty wet for a more commanding, massive effect. If I want to add delay for emphasis, I slap Logic's Echo on the last insert of the main track and automate it to activate on certain parts.

For clean vocals (and screams for records), I go to my friends studio, but that's an entirely different story.

fractured said...

Lately, I've been using a BLUE Baby Bottle into an API 512c for preamping. That goes into my RME Multiface and Logic. I usually set up the Channel EQ with a low end rolloff and then go to Logic's compressor, set to the optical or the Class A U with a 4:1 or so. I'll then insert a small or mid Space Designer with just enough reverb to feel, but not really hear. Next the vocal hits whatever effects I'm after to tailor the sound and a Stereo Delay.

Rayzr said...

Studio Projects C1 via XLR input into MOTU 828 Mk3. Cubase 5 audio channel, has several presets that load up strips and EQ the channel automatically. I usually go with one of the Male Vocal ones, and then from there change the built-in native EQ up to match the vocal itself, usually cut out the lows @ 140khz and boost the highs around 2k or so.

From here it goes through the Voxengo Voxformer (last in the chain), Cubase 5's vocal preset will have loaded a compressor or other native plug's, however I usually disable/remove them to taste.

There will be two sends usually @
-1.25db, one going to Cubase's DoubleDelay with a hard 100% left/right and a 1/4 - 1/8 timing.

And I've been trying to use two Reverbs (native, either Reverb A or Reverence) one hard right and one hard left, both set as sends.

I used to use a lot of the Waves stuff instead of the native plug's. But they are such a pain in the ass now that I'm on a x64 architecture.