Mixing Rap Vocals – Part 3: Compression

➥ Take your hip-hop beats and rap vocal mixes to the next level

Time for the third installment of the Mixing Rap Vocals series: Compression.

I highly recommend you check out part 1 & part 2 before reading this article.

Compression is a difficult subject because there is so much you can do with it. So let’s look at the main reasons to grab a compressor before getting into some of the more intricate uses.

Quick Macro-Dynamic Control

Macro-dynamics refers to words and phrases. These are the clear dynamics you can hear as “this part is louder, this part is softer.” The most transparent way to get things sounding even is to actually automate the vocals manually. But sometimes time doesn’t allow for this approach. So if you aren’t automating, a light ratio, slow attack, slow release, and just catching the louder moments with the threshold is a good way to even things out.

Micro-Dynamic Control

What volume automation might not catch is the very quick dynamic changes: loose spikes at the fronts of words. These spikes aren’t heard so much as “volume” but more as an overall quality to the vocal.

The issues with these spikes are two fold: first, they eat away at your headroom pretty quickly. Second, they will trigger any compressors you are trying to use for purposes besides micro-dynamic control.

It can be useful to dedicate a compression stage toward pulling back these vocal spikes. Generally a fast attack and release, and a light ratio does the job. The light ratio is to retain the articulation of the word and minimize frequency skewing. The key is to set the threshold low enough to catch as much of the peak as possible while effecting the body of the signal as little as possible. I try to avoid using limiters for this purpose. I like the Empirical Labs Distressor for this (especially for controlling peaks while tracking), as well as digital style compressors such as the Logic or Pro Tools stock compressors or the Waves C1. The attack setting is very important — usually between a number of nano-seconds and two or three milliseconds in the digital world, and on the faster side of things for the analog world (totally varies unit to unit).

Getting a Vocal to Stay Audible Through a Mix

The power of compression is that you can make something louder while not actually raising the peak volume of the signal. This becomes extremely useful for making something cut through a dense mix or come forward. This is probably where the majority of compression work for rap vocals come in.

Rap is generally an in-your-face, visceral style of music. The kick is physical, the snare is physical, and subtlety isn’t really the overall goal. The vocals are paramount. I’ve mixed a number of rap records where the vocals are lower in the mix, but never have I thought it was a good idea. Generally I want the vocals to be equally as strong as the drums or stronger, and I want them as “forward” as possible. Compression is usually a part of that equation.

Optical Compressors

The smoothest way to get those vocals forward is through optical compression. The rounding quality of the attack and the unique shape of the knee in an optical compressor makes them ideal for vocal work. Examples of optical compressors would be the CL1B, the LA2A, LA3A, your stock Logic compressor has an optical mode, RComp has an optical mode, and don’t quote me on this, but RVox has an “optical” sound to it, as does the “smooth” setting on the UBK-1. One of the advantages to opticals is they tend to have easy access. Many have just one knob to control the degree of compression.

Attack and Release Time

Of course you’re not limited to simply optical compressors or fixed-time settings. Many other compressors work very well for rap vocals. In fact, any decent compressor can yield great results if set properly.

The key is setting the attack and release times right. People will suggest milliseconds or time ratings for the best attack and release for vocals but the issue is that 300ms on one compressor might give you the same results at 75ms on another. So, I’d rather advise your compression technique based on the expected results. Your goal is to pull up as much of the “sustain” of the voice — the weight of it — while minimally affecting the articulation. Taking notes? It’s about to get heavy.

When dealing with the articulation of the words, you’re primarily gauging your attack time. There’s generally a substantial range of attack speeds that work for vocals. What you don’t want to do is set the attack too short, or the shaping of consonants will be blurred. Nor do you want to set the attack too long, because you’ll allow the consonants to poke through too hard. So you want to find a middle ground. A good way to experiment is to temporarily pull the threshold down a little farther than you normally would and find your attack setting that way, as the effect of the attack time will be more exposed with the lower threshold.

With the release time, my goal is to pull up as much of the body of the voice as possible. So I’m going to set the release on the faster side. I don’t want the voice to distort or become unnatural sounding, but I want as much body as possible before I get to that point.

In terms of both the compressor ratio and the release time, I tend to be a little more aggressive with rap vocals than “softer” music. The “integrity” of the vocal sound is not really as important as the prominence of it. For a more relaxed, natural sound, I might do a medium release and 3:1 or 4:1 ratio. For a rap vocal I’m going for a pretty quick release, and I’m doing 4:1 up to 8:1. Rap isn’t really supposed to be “pretty,” so I don’t worry if the compression becomes a bit audible.

Thicker Vocals

Another great use for compression on vocals is to make the vocal sound thicker — particularly in rap.

Rap is frequently recorded in home studios, even by big name artists. And home studios rarely produce the thick, full vocal sound that one can get at a professional facility. So being able to thicken and give weight to a vocal is an extremely important skill. In order to do it right though, you need a little more than compression. You need an EQ to make sure the vocal is as even and smooth as possible. Then you need some “saturation”. Saturation is just a nice name for friendly distortion. Saturation moves and enhances the harmonics of the vocal. Over-saturating will sound like crud, but just the right amount gives the impression of a richer sound.

So the formula is: get the vocals sounding clean, saturate to get the vocal sounding richer, and compress that signal. You have to tread carefully though — over-EQ’ing, over-saturating, or over-compressing will make your vocals sound horrible. Unless you do all of that in parallel!

Parallel Compression

Parallel compression basically means making a copy of the signal, compressing the snot out of it, and then blending that parallel signal back in with the original signal.

The advantage here is that you can get really liberal with the effects, and just blend it in to where it doesn’t sound unnatural. This is great is you are trying to fill out frequencies that weren’t really there in the original recording, because you can really saturate and compress the parallel signal and generate some very consistent dense harmonics. Then just blend that in until just before it starts to sound too effected. One of the reasons I really like the UBK-1 for vocals is because it gives you a saturation stage followed by a compression stage and the ability to blend both in parallel.

Distortion Free Equalization

Lastly, compression can be used to tame frequencies without the artifacts from EQ. Often times with vocals, you’ll have moments where the vocalist changes their tone. A common example is vocalists will often become more mid-rangy as they project, because they tighten their neck and push more air through their nose. If you simply notch out some midrange, that might work, but it might also might take away some of the energy of the overall performance, or some of the frequency information you need to make the vocal stand out.

A good alternative is to use a compressor with either an adjustable sidechain, or an external sidechain input. The sidechain signal is what the compressor reacts to, so if you can EQ the sidechain to target problematic frequency areas, the compressor will “intelligently” react to those tones and pull them down.

Conclusion

Compression is a powerful tool that many people struggle to fully understand, so try to get your hands on one and start experimenting.

As always, I’ll keep an eye on the comments in case there is anything that needs clearing up. I also encourage you to share your own compression tips!

We just released Mixing Rap Vocals 2: Advanced Techniques featuring Snoop, Murs, Juicy J, Arrested Development and more. Watch the video to learn more.

Matthew Weiss

Matthew Weiss

Matthew Weiss engineers from his private facility in Philadelphia, PA. Credits include Snoop Dogg, Gorilla Zoe, Arrested Development, Dizzee Rascal, Gift of Gab, J-Son and many others. Get in touch at Weiss-Sound.com.
  • Pingback: Compressing rap vocals (and other thick and aggressive vocals) | Kim Lajoie's blog()

  • http://www.audioczar.net czar

    75% of the music I mix is rap vocals. I normally use an opto compressor (softube cl1b or mcsp 6030 opto-c)to start with. If I use another compressor I’ve been backing off using super fast attacks to try to preserve transients lately. I’ve recently discovered how great the saturation on the softube focusing eq can be to.

    • http://www.weiss-sound.com Matt Weiss

      I usually start with an opto or the “Glue” setting on the UBK-1 (which models the attack and release of an 1176 at 10&2). Then a touch of RVox at the end.

      I find I’m doing a lot of layers of compression – a little UBK-1 at the beginning, a little RVox at the end and a little parallel compression overall.

      A good saturation just to give a little spark is also a great touch – the softube stuff is pretty damn good. Only reason I don’t have it is because I have other stuff I’m happy with, but the CL1B and softube passive and active are both GREAT.

  • Uouin

    Thanks Matthew, for this detailed explanation. I hope I can apply it. A few questions: A) To clarify, you recommend the vocal chain in order to be De-Esser, EQ, then Compressor ?

    B) You mention a lot of fairly expensive equipment here. I’ve been trying to get vocals right using strictly Pro Tools stock plug-ins. I have also played with the BF1176 plug-in on vocals, but nothing has really worked for me yet. Pro Tools EQ has not worked well for me without creating ‘artifacts’ as you say. I’m admittedly novice at audio engineering. What plug-ins would you recommend (whether a compressor or EQ, etc.) for a novice with only a few hundred dollars to spend? What makes the most difference for the buck?

    • http://www.weiss-sound.com Matt Weiss

      Hi Uouin.

      A) Definitely changes. I usually address what stands out to me the most. I’m mixing a project now where the first item in my chain for the vocals has been a compressor most of the time. I try to de-ess after compression, but before high-end boosting. Because a compressor will just bring esses out again, and it’s counterproductive to boost high end into a de-esser. But EQ or Compressor the order changes around. Convention dictates you “correct before you enhance”.

      B) Honestly, a lot of times the very first processor in my chain is the stock digidesign compressor. I really like the versatility I can get from it, particularly with the knee function. I’ve come pretty full circle on the equipment I use – I find the better I get with my sensibility, and the better I understand a particular piece of gear/software the better results I get. But I don’t feel I would be terribly limited in my capabilities with only stock plugins. Best things you can do are to step up your monitoring and if you are the artist as well, then step up the tracking chain. The rest is experience and experimentation.

  • Uouin

    Thank you very much Matthew for the quick and thoughtful responses. I am honestly pretty happy with my tracking, and my girlfriend (i.e. “other ear”) tends to like my vocals best with just the compression. Once I start trying to use the BF76 and EQ it starts getting wonky. I have a nasaly voice (somewhere between Kanye and Biz Markee) that doesn’t have a ton of range to begin with, and I find that cutting reduces the range even further, while boosting tends to make me mid-rangy and/or trebly.

    The problem with just using the compressor and saying F it to the EQ is that the vocals sound thin. And, thus, it is a tightrope act in my vocal processing. I am seriously considering the UBK-1 to see if that can add some thickness without sacrificing the clarity and range I want in my vocals.

    I am sure you are right that I simply need to refine technique, and I appreciate the good tips you’ve given to get me in the right direction..

    • http://www.weiss-sound.com Matt Weiss

      I’d have to hear the vocal to really make a fair assessment. But from your description I’d say the UBK-1 is a good bet. The saturation stage is perfect for filling in the “thickness” of an otherwise thin vocal, and the density stage is very good for pushing the top end forward on a midrangy vocal.

      But it’s really smart to know that if your EQ’ing isn’t improving the sound that “no EQ” still counts as a choice and a setting. It’s good to be honest about what you are hearing because sometimes vocals really don’t need EQ. A fair amount of time when I’m EQ’ing I’m really only doing it to make one element sound different relative to another element. Like an acoustic guitar monitored against a vocal, I might take something out of the acoustic guitar to let the vocal pop through – something that I wouldn’t necessarily remove from the guitar if it were on it’s own.

  • Uouin

    Thank you again! Have purchased the UBK-1, and am starting to put your tips into motion. Do not expect any more feedback, and am very appreciative of what you’ve given; however, consider this the “BEFORE” picture that I am working with: http://soundcloud.com/lqoau/texas-relays Hope to give you a good after snapshot as well. Many thanks!

    • Uouin

      My song is sounding much better. Not sure it’s “there” yet but at least decent. Thanks again for all of your help and guidance. I ended up re-tracking the vocals (a few thousand times), and really get what you’re saying there now– you can’t expect to fix poor vocals through processing, and tracking vocals properly is far more important than EQ and compression, etc.

      On the UBK-1, used the glue setting you recommend, with minimal saturation and the density off. I still can’t get EQ to sound right by adding anything, so I used a HPF between 80-100 Hz, with a notch or two out between 100-250 Hz depending on the verse. On the beat track, I took out a notch at ~1.5, 5, and 10 KHz to try and let the vocals pop out a bit. I’m sure the EQing could be better, but I decided to take the safe route, and this works okay.

      Still long for the day when I can pay someone to do all the production for me, but this has been an important learning experience, and your 3 articles on here are the best I’ve found on the internet in really teaching concepts. I’m sure I’ll be reading and re-reading them again here soon..

    • http://www.weiss-sound.com Matt Weiss

      Well – it’ll help your EQ to remember that you’re not boosting or cutting frequencies. You are boosting or cutting sounds and tones that occur at frequencies. It helps to listen to the voice and make decisions about the sound. Is it thin? Is it thick and gunky? Does it feel hollow? Nasal? Start prescribing adjectives to the sound and see what makes sense. Sometimes the best EQ setting is BYPASS. Sometimes the EQ looks like someone handed an etch-a-sketch to psych patient.

      Also, the density setting on the UBK can be cool for vocals. I tend to use the “Top” density setting fairly often. And occasionally the “mid.”

  • Jay

    Big thanks to the author of this post!!

    Ive been strugging with compression for years and no matter how hard i tried to understand, it still confussed the hell outta me..

    Luckly im currently working on a track that features hip-hop / rap vocals that needed so much work, re-recording was not an option as im not the rapper lol.. this info had helped me understand compression a little better, still not sure about how the threshold works and what its for lol, but for this track i found i had to reduce the threshold by loads but increased the overhall level with a “makeup” setting on the vst..

    I used a ratio of 4:1, attack 2.5 (ms) and a release of 0.001 (ms)
    gave the vocals a nice punch :)

    thanks again

  • http://www.audioczar.net czar

    @jay

    the threshold determines at what singal level the compressor will start to compress. So for your example with your settings if you have a threshold of -12db, then 2.5ms after your signal gets louder than -12db it will start to compress. When it goes below the threshold no compression will occur. Your basically telling the compressor when get this loud (-12db) I want to start to start compressing. Then your attack determines how fast it starts to compress.

    • Jay

      Hi czar thanks for the info, i so needto get to grips with compression but its tricky to understand lol…
      Need to find a good tutorial on the net that explains all the settings and what they do in simple terms.

      Ive had a few hardware compressors in my time but even with the manual the explinations can be very complicated and they explain things in a way as if you already know lol..

      back 2 basics i think, i really needto understand compression before i use it, the vst has a few presets but it would be nice to be able to understand what im using so i can take control of stuff ..

      Thanks again :)

    • http://dancomerchero.com Dan Comerchero
  • Pingback: Mixing Rap Vocals – Part 2: EQ()

  • http://tonedango.com J.P. Leach

    Cool compression guidelines and a nice complete post. I am going to reference this post the next time I record someone’s vocals.

  • Pachakuti

    Hey! thanks for this article!
    When i put a compressor on my vocals some of the plosives start popping to much. No matter what i tweak this effect stays. The only to get rid if that effect is to tweak so much that nearly all of the desired compression-effect is gone.

    Can anyone help me out?

  • Pingback: Anonymous()

  • Flex

    Thankz for the tips……I spend 2 weeks right now to get the best out of my voice ….it tooks a lot of experience ….the best is now I read your tips and i saw I did a lot of things, like you wrote in here, on my own…..seems to be I´m on the right way !

    Big up from Europe

  • BB.KOHL

    what i usually do is start off with automation killing any white noise then work on the vocals so my compressor wont have to work to hard before compressing i eq to taste and throw a noise gate desser and some distortion or saturation to add a little more giddy for rap vocals.

  • kai

    I know that you said a lot of stuff here that was extremely helpfull, but it would definitely benefit me if you could tell me what you’re ideal vocal chain order would be. I get it for the most part, but there were some places were you discussed using two or more compressors/equalizers, and that would be nice for you to clear up. Thanks for all the help!

  • Pingback: How to Make Your Hip-Hop Beats Stand Out()

  • Pingback: Ideale compressie instellingen rap - homerecording.be forum()

  • Pingback: Tips for Mixing Sample-Based Hip-Hop()

  • Johnny Loose

    GREAT article!! Thanks for giving a thorough writeup!!!

Recommended