Using compressors and limiters

by Keith Hatschek on February 25, 2011 · 38 comments

in Fast Forward,Recording & Mastering

Dynamic changes give music shape and emotion – using compressors and limiters allows you to smooth out passages where the volume is too loud

For many recording musicians, the compressor presents an enigma. Most of us know that using compressors and their more specialized cousins, limiters, controls the dynamic range (aka volume) of an audio track or mix. But exactly how do they work, and from a practical point of view, how can you use them to the greatest advantage? We’ll provide a brief overview, and then head into the studio and see how the use of a compressor can help the recording process.

While dynamic changes in volume help give music its shape and emotion, too great a change in volume, and sudden changes, may create challenges when recording and mixing your music. Using compressors and limiters allows you to smooth out passages where the volume may be too loud.

What’s the difference between a compressor and a limiter?
A limiter is a compressor, but a compressor is not necessarily a limiter. Limiters are designed to prevent the sound level of an audio program from going beyond a certain, pre set point. They generally have very high compression ratios of 20:1 variable up to Infinity: 1. (A compression ratio of 20:1 means that for every 20dB that the input signal increases beyond a preset point, the limiter will only allow output gain of 1 dB.) Limiters are critical for in-ear monitoring systems, radio and TV broadcasting as well as vinyl disc cutting systems. Limiters are seldom needed in a typical home recording session. Compressors will normally use lower compression ratios, often in the 2:1, 4:1, 8:1 range, meaning that the amount of gain reduction applied is not as drastic as with a limiter.

Compressors, on the other hand, are frequently used tools that reduce or control the dynamic range of a recorded track. A good way to envision the dynamic range of a track is the difference in volume between the loudest notes to the quietest notes. Compressors are handy for controlling many of the typical tracks you may be recording such as bass guitar, vocal, acoustic guitars, keyboard, etc.

Not every track will benefit from or needs to be compressed. If the original performance has a fairly limited range from the loudest to softest note, then using compressors will make little or no difference. However, if we consider a lead vocal track where the singer starts off near a whisper for the introduction and by the song’s end is wailing over a pounding band, putting a compressor on that lead vocal track may help reduce the dynamic range of such a vocal track so that it can easily stay on top of the band and allow for an easier overall mix.

Compressor Settings
The controls on most compressors are fairly similar. The first and most crucial setting is the Threshold. This is the point at which the device will begin to reduce the level of the output signal.

How much the signal is reduced is determined by the ratio. At a 2:1 setting, for every 2 dB of increase to the input signal beyond the threshold, the output signal level will increase by 1 dB. At a 4:1 setting, for every 4 dB of increase of input level beyond the threshold, the output level will increase only 1 dB. Higher ratios result in more dramatic gain reduction. If you have a home studio, put a compressor in your audio chain, and click between the different compression ratios available while you compress a track and you’ll quickly be able to discern which ratio works best on various instruments.

Two other controls affect the sound of the compressor. These are the Attack and Release controls. The Attack control allows you to fine tune just how fast the compressor reacts to the input signal with gain reduction when the signal overshoots the Threshold point. Conversely, the Release control allows you to set how fast the compressor lets the original signal return to normal once it is no longer above the Threshold. If the Release is set too fast, it may result in a sound often referred to as “pumping” or “breathing,” which draws attention to the compressed track. It’s up to you to decide whether or not such a sound may be useful in your overall concept for a song. Once again, spending a few minutes in the studio playing with these controls is the fastest way to hear how they work.

One other important benefit of using compressors is that when set properly, the net effect is to increase the apparent overall level of a track, making it seem louder, by compressing the loudest parts without creating any overload.

Into the Studio
I stopped by our campus studio where some musicians kindly offered to make a quick demo recording in Pro Tools and then process it with a plug-in compressor. We started by recording an electric guitar through a small amp with no EQ or dynamic processing. We played back the track and inserted the Pro Tools DIGI Rack Compressor/Limiter Dyn 3, a common plug-in with most Pro Tools systems. (Although we recorded without effects, some engineers prefer to record with a compressor engaged from the start, while others record without any compressors, preferring to make decisions about dynamics after the performance has been captured.)

At my request, our guitarist Nic played a soft picking pattern first and then shifted abruptly to a more aggressive (and louder) chordal texture. Without compression, the chordal texture dominated the demo guitar track.

before using compressors on guitar

Guitar before compression

To demonstrate the before and after effect of the compressor, we actually decided to over-compress the track, so that when readers listen it to by clicking on the link below, you’ll be able to clearly hear the difference. For an actual session, this level of compression would likely be judged excessive, unless you were going for an effect. (Over-compressing a track for an effect is also a very creative tool, but outside the scope of this article. Listen to some of the Beatles recordings from Revolver or Sgt. Peppers to hear the heavily compressed sound of Ringo’s drum kit for an idea of how effective such a use can be.)

“Guitar

After tweaking a bit, the student engineers settled on a threshold of -25 dB, so even the softer part of the guitar track was being compressed slightly. We use a ratio of 6.6:1, which gave us a fairly consistent overall final track reducing the level of the chordal texture to a point where it was only slightly louder than the softer first section.

This particular DIGI plug-in offers a real time analyzer (RTA) screen, similar in look to an old school oscilloscope, allowing the user to see the point at which the signal is being compressed and by how much. The vertical orange line on the RTA screen represents the threshold point and the intersecting white line represents the compression ratio beyond the threshold point. So with this particular ratio of 6.6:1, the slope of the compressor looks similar to a human knee, hence, this type of setting is often referred to as knee compression.

After listening back to the extreme compression we used, we realized that although we now had a much smaller dynamic range, if we actually wanted to use this track, we would have to boost the level of the guitar a bit, but nonetheless, the wide dynamic range on the original recording was fixed.

How much compression you use is also determined in part by the complexity and number of tracks in your overall mix. The denser the mix, the less apparent the individual compressed track will be, as opposed to say, a solo guitar and vocal, where the compressor must be very carefully set to avoid distracting the listener.

Compressing Acoustic Piano
Next, we made a stereo recording of an acoustic grand piano, and asked Nic to play both a quiet and a loud passage. For this compressor demo, we used a lower ratio of 3:1, but to show the compressor at work, kept a threshold point much lower than one might normally select for a piano, setting it at -24 dB. This way the compressor would kick in a bit on the quieter passage, but really be working hard on the louder section.

The overall dynamic range of the piano track was roughly 25 db, from a soft intro around -30 db up to some peaks in the louder section that hit -6 db. With the compressor settings chosen, we should be able to reduce the overall dynamic range to one that is more usable.

before using compressors on piano

“Piano

On playback, we found that the overall dynamic range had been reduced by about a third, from 25 dB uncompressed to about 16 db, when compressed. As a result the overall track sounded more consistent and the softer intro was more apparent. The net result was a track that would fit more consistently and will neither disappear in the quietest passage, nor overpower the mix when the volume kicked up in the louder section, when added to a full band mix.

after using compressors on piano

Piano After Compression

This particular compressor has another nice feature, which shows both the input and output levels as well as a third meter, which displays the amount of compression or gain reduction that is being applied. In the adjacent screen grab, the two green meters represent stereo input and output to the compressor, while the orange meter just to the right of the green meters tracks the overall amount of gain reduction being applied by the compressor in dB.

using compressors on stereo piano

Stereo Piano

As a general rule of thumb, it’s best to avoid heavy gain reduction on vocals, acoustic instruments, and percussion unless you are going for a specific effect. Conversely, electronic instruments such as bass, keyboard, synth, and electric guitar can handle a good bit of gain reduction and still sound like the actual instrument. As with so much in recording, it’s up to the ears of the recording engineer or musician to determine how much is the appropriate amount to use.

Start recording and get your compressor in the circuit to see how you can help smooth out your tracks to best effect.

Audio files
Guitar without compression
Guitar with compression

Piano without compression
Piano with compression

Keith Hatschek is a regular contributor to Disc Makers Echoes blog and directs the Music Management Program at University of the Pacific. He’s also written two music industry books, How to Get a Job in the Music Industry and The Golden Moment: Recording Secrets from the Pros.

Special thanks to musician Nic Chang, and student engineers, Joey Frantz, Noely Gonzalez and Josh Walkover, as well as studio manager Jeff Crawford for recording the sound files for this article.

Story Links
Understanding Audio Compressors and Audio Compression
More on compressors in this extensive article by Barry Rudolph – he also gives a few of his personal settings near the end of the article.

Classic Compressors
Here’s another article about using compressors on vocal tracks as well as a mention about the vintage Fairchild compressors used by Geoff Emerick recording the Beatles.

Build 
your own home recording studio

Read More
Recording Electric Guitar
Home Studio Recording Tips From a Pro Studio
Home Studio Posts – Recording Tips For Producers, Engineers, and Musicians
Signal Processing For The Home Studio Owner: Part 1, Compressors, Limiters, and EQs
Signal Processing For The Home Studio Owner: Part 2, Gates, Delay, and Reverb

Share this articleShare on Facebook939Tweet about this on Twitter78Share on Google+6Share on StumbleUpon1Email this to someoneShare on LinkedIn17Share on Reddit0Digg this

{ 2 trackbacks }

JSonar Project
March 7, 2011 at 2:02 am
Fly Tucker Productions
March 17, 2011 at 3:21 pm

{ 36 comments… read them below or add one }

juan negron March 1, 2011 at 4:57 pm

thanx for this info was very helpful. in other words, does this same method of using compressors and limiters applies to hiphop music and the rappers’ vocals?

Reply

DAS March 1, 2011 at 9:13 pm

Hey Juan, yeah the use for compressors definitely applies to hiphop, as well as every other style of music. The hard part is kind of what you’re asking, knowing how to apply the basic techniques of compression for the style of music you’re working on.

In other words, there are certain techniques that are used specifically in electronic music, like in house/dance or trance when you hear that “pumping” effect of the beat. Or even in a lot of the classic Michael Jackson tracks from Dangerous , History albums, etc that have a real big “crunch” to the snare drums and claps. But as well how Bruce Swedien and other audio engineers used compression to make Micahel Jackson’s vocals sound like they’re 5 inches in front of your face. Or the crunch of big rock guitars using compression.

All of them use compression, but it’s used “artistically” for your own purposes. It took me a few years to finally “hear” the differences in using compression and limiting, both on individual tracks, and even on the 2 track master. But the more you practice, the easier it gets.

Reply

Jim Hougen March 1, 2011 at 6:58 pm

The compressor demos were very effective taken as a set. Would like to see more on this subject as concerning classical style symphony, with choir, piano, solo

Reply

Omar March 1, 2011 at 8:28 pm

Props to Keith Hatschek! This is by far one of the best explanations of the basics of compressors and compression application I have ever read, and I’ve read my share. Keep these articles coming.

Reply

Brad March 1, 2011 at 11:36 pm

Agreed. Great article!

Reply

Adam B Harris March 2, 2011 at 7:33 am

“Not every track will benefit from or needs to be compressed” – good advice, too many people I’ve worked with want to use accross the board compression settings on everything they record regardless of the instrumentation or genre of the song or tune. A little compression can go a long way, particularly when recording acoustic instruments. Trust your ears first always.

Nice article, clearly explains a concept that many struggle with, well done.

Reply

Dave Lister March 2, 2011 at 9:33 am

Thanks for th refresher course. Great timing for me as I’ve been bouncing commpressor settings off of other studio owners. It is all in the ears and what you want to do a track, mix or Master.for us Studio Junkies, the more info the better ! Thanks and nice job

Reply

Barry Higgins March 2, 2011 at 11:00 am

In looking it and listening to the samples given this article it shows the good and the bad. With the intent to illustrate over compression to make it obvious to the eyes and ears, it also illustrates the war on loudness or loss of dynamics in much of todays pop and rock music. The guitar track loses out by the loss of the dynamics which most likely once make-up gain is applied becomes an intense ear fatiguing squashed performance for the listener. The piano track was well applied and enhances the listeners experience by bringing out parts that the player dynamics were too soft or hard for the recording chain makeup ie; mics, placement, pre amp, gain, convertors, and bit rates. End result is an expressive play was retained and a better listening experience. That all said creative use by over compressing may make the difference in a mix or intent of a song. Just make sure that if your intent is for radio or even YouTube that these people will re-compress your audio for play so leave them some room so that it does not get totally squashed.

Reply

THE GAYLE March 2, 2011 at 12:35 pm

Compression across your whole track is the best way to ruin your music’s “tone”.
The biggest mistake I hear being made on CDs is…over compression.
It KILLS the tone (requiring excessive re-EQing). And sucks the life out of it.
As soon as you can hear the effects of compression, you can hear the loss of tone in both the high frequencies (over-tones) and in the bass articulation.

My view is that compression on individual tracks can be a good thing (and is NEEDED almost always on vocals) but should be avoided or used VERY minimally on the master mix.

Just my opinion!

What do you guys think?

Gayle

Reply

Moose March 18, 2011 at 2:08 pm

What “kills sound” on CD’s isn’t overcompression… its brickwall digital limiting. Very big difference. Overcompression can have a cool musical effect, use the example of RIngos drums above or any number of modern rock records… STP, Foo Fighters, Mastodon or even Elton John etc. Brickwall limiting is what makes things really loud and destroys musicality. Very large differences exist between an 1176 and Waves L2/L3 even though both are tagged as “limiters” rather then “compressors”. Cheers!

Reply

Mb February 6, 2014 at 11:18 pm

The discussion just quickly gets comical. The same producers who preach against the loudness wars often seem to often be the ones doing it anyway, just to compete.

Since you asked for an opinion, mine is this. If mastering techniques are used to create the loudest possible track, everyone loses. It has absolutely ruined popular music for me.

Every time the subject comes up, there is an agreement about whats wrong, then everyone goes back to work trying to master the loudest song of all time.

So Gayle, I encourage you to hold the line. Do it your way and forget what the rest do. Make music that sounds good to you, and be happy.

Reply

Da Chedda Monsta March 2, 2011 at 1:07 pm

Loved the article – thanks. I’ve always been confused by the limiter, myself. I’m going to experiment tonight!!

Reply

Vulcho Bonev March 6, 2011 at 5:10 am

Yes compression/Limiting can be very tricky, as well as, being very simple, i tend to end up going with the ear. I’ve been using Universal Audio, focusrite, and avalon products. But I have to admit I am constantly experimenting with it , and am still learning every time .

Reply

giomakyo March 8, 2011 at 1:20 am

>nonetheless, the wide dynamic range on the original recording was fixed.

This assumes you see wide dynamic range as a problem. Instruments naturally have this range, and in many cases, it’s worth preserving. Generally the fewer instruments you use –or esp. in solo situations — then wide range is not a bad thing. However, in more crowded mixes with a lot going on, you will need to decrease the range somewhat to get an overall more consistent level. However, it’s all too common these days for people to assume EVERYTHING needs to be compressed heavily, which is definitely not the case.

Reply

Jerry Jean March 9, 2011 at 1:54 pm

Excellent article on compression. This explains the concepts very clearly. Thanks for an informative read! ~ J

Reply

Peter D'Amore March 21, 2011 at 1:39 pm

Thank you for this article.
I have a nice Tascam digital 8 track, so there is no computer visual going on for me to “see” what the compression is doing.
When I ask professional people about using compression, which I haven’t yet, I get so many different answers. This article helped me out a lot.

My GNX1 Digitech Guitar Effects pedal has a very detailed Compression setting in it, with Threshold, Attack, and Release settings. I was never able to make sense of what it was doing until now.

So with that, maybe I could use the compressor in this guitar pedal (I know, sounds strange) to do an external effects loop for the final vocal track, to bounce to another available track with the desired amount of compression??

I worry about how much hiss the GNX1 would give the new track…
Sorry for my ‘caveman’ question, but the album sound quality sounds great so far, but I’m hoping that compressing the vocals would improve things.

Reply

Peter D'Amore March 21, 2011 at 1:41 pm

Sorry – I forgot to mention, for my question above- the vocal take is done already, without compression, at this point.
So, I would be playing the vocal track back, sending it through the GNX1, and bouncing it another track with compression now.

Reply

brian lipski March 21, 2011 at 8:06 pm

Peter, personally I would be pretty concerned about the amount of hiss that pedal is going to add to the vocal track. It can’t hurt to try, but a multi-effects unit designed for guitar is likely not going to cut it for a vocal track, especially if the arrangement is sparse and there is a lot of space for the hiss to show in the mix.

Some compression will definitely help smooth out your vocal track, so it may be worth it to look into renting or possible borrowing a dedicated compressor from someone.

Reply

Mung March 27, 2011 at 8:04 am

Another good comp chain I’ve found for vocs is using the SSL comp and Waves RenVox together in a chain. 4:1 ratio on the SSL with a slow/medium attack and release set to auto. On the RenVox I just play with the settings till I get the vocal to sit where I want it.

Reply

Rocker Bryan Eddy May 11, 2011 at 2:55 pm

You’re not kidding! its been the bane of my recording existance.I never got a good one,never learned to use it and my stuff (1000s of songs)sounds awful without it .Thanks for all the knowledge. everyone.

Reply

compression is so 1980s lol May 11, 2011 at 5:10 pm

Great article however Nic is terrible at attempting to play guitar. Find a musician or guitarist for your next example. Other than the painful sounds Nic bestowed on us, your work was quite informative, thanks! =k

Reply

Flaming is so 1990's May 11, 2011 at 8:00 pm

trolololol

Reply

Dennyps July 27, 2011 at 1:09 am

Thanks so much what a very good article and very clear and concise description of the compressor/limiter and it’s settings.  In fact as I write this I was trying to eliminate some problems in a great bass line I had just layed down on a track and this article helped me in “real time”.  Thanks again, Denny

Reply

Charles Ray Batchelor July 27, 2011 at 1:54 pm

This is Really Cool that you placed the sound on here, for that is the best way to explain such things.  

Charles R. Batchelor
c/o A “Higher Call” To Artists

Reply

Studio_mike September 22, 2011 at 1:27 am

you guys are awesome for sharing your craft with all us home recording wannabees!

Reply

John Paul November 25, 2011 at 6:10 pm

Very informative and concise. Exactly what I’ve been needing to know.  Thanks!

Reply

WACHO December 29, 2011 at 2:20 pm

Thanks for the great knowledge !!

Reply

Ricky Barnes December 29, 2011 at 3:44 pm

Thank you for your help.  It is truly appreciated. http://www.barnesbarbq.com

Reply

Poppareno January 29, 2012 at 6:44 pm

I’m using a compressor/limiter on my bass drum mic in live performance. Any suggestions on how to keep it loud & round without booming?

Reply

Jack February 4, 2014 at 7:57 pm

“compressors and their more specialized cousins, limiters, controls the dynamic range (aka volume)”

Not EVEN close.

Volume DOES NOT EQUAL DYNAMIC RANGE.

NOR does a “compressor” provide “make up gain” That would be an amplifier, usually at the end of the compression circuitry. IT is NOT a mere “resistive” pot, it is really an amplifier.

Please present facts and have a qualified fact checker check your facts. This type of “mis” infomation is only ONE of the reasons I left and will not return to “pro” audio. Same reason for Video (The “Difference? LOTS more money in video therefore more smoke, mirrors, half truths and general no goodnik “mis” information. i.e. How many video producers does it take to screw in a light bulb? I don’t know – what do you think?)

Barry Rudolph REALLY knows his stuff. Read his article on the $300 mic kit and what he produced, then check the track.

Thanks Barry.

Reply

Steve Lushbaugh February 5, 2014 at 4:14 am

Ditto on the brick wall limiting comment. Most of today’s music is produced to make everything as loud as possible. Look at the waveform of most any Metallica song (I know, not exactly “today’s music” but it seems to me that’s where it really became a widely used formula). Typically there is a soft guitar intro with some dynamic range, then everything comes in, turned up to 11 and limited into oblivion. After the intro, the waveform looks like a solid block, almost like a tone, with the very occasional dip or gap lasting a few milliseconds. It’s hell on the ears, causes listening fatigue, eliminates all dynamics and rapidly accelerates hearing loss. If people want to be deaf, it’s quicker to shove long needles through your eardrums, and as far as I’m concerned, far less painful than slowly going deaf listing to lifeless music. Everything is loud all the time. It is bereft of nuance. It saddens me that this is how music is made today. Crescendo, diminuendo, forte, fortissimo, piano, pianissimo, mezzo, sforzando and every other dynamic change is absent from most popular music thanks to the ham handed use of gain reduction. The notes are only half of the music. The dynamics are the other half and they are squashed out of most music, if they were ever there at all. Sure makes music easier to play, and harder to listen to.
Radio has long over compressed and limited programming as part of “loudness wars,” a belief that sounding louder than other stations on the dial made a station sound bigger and better to the listener. Now the program material comes pre-compressed so the chance of the needle moving is pretty much nil. Everything is in your face all the time. The only upside I can find is, if you are listening to music in a noisy environment, like under the engines of a fighter jet, you can still hear everything that comes out of the speaker over the roar of the jets. In most environments it all sounds like a flat wall of dull noise that becomes oppressive in short order and makes you want to turn it down. Remember when music made you want to turn it up? Guess I’m showing my age but I fail to get the appeal.
I’m not anti-dynamics processing. It is one of the most basic, original and important processing effects in production. But a monkey could strip all the dynamics out of a recording. It takes a skillful producer to use the tools judiciously to make music smooth and balanced without stripping it of every bit of dynamic range. All that gain reduction has other side effects, like distortion, especially in the low end, and a grittiness that becomes grating, especially when listening to stuff like that at loud volume. There is a middle ground but it is no longer occupied by anyone in rock or pop music these days. It’s a sad state of affairs and a great loss to the art of much music.
Don’t get carried away. The meters can move during a song, not just hover at the red line from start to finish.

Reply

Mb February 6, 2014 at 11:29 pm

The discussion just quickly gets comical. The same producers who preach against the loudness wars often seem to often be the ones doing it anyway, just to compete. Every time the subject comes up, there is an agreement about whats wrong, then everyone goes back to work trying to master the loudest song of all time.

If mastering techniques are used to create the loudest possible track, everyone loses. It has absolutely ruined popular music for me.

While this old article may not dig into details with complete geekdom accuracy, it gives a fairly good overview. I appreciate the attempt.

Reply

Ken McCray March 28, 2014 at 11:34 am

I was under the impression that a limiter worked exactly opposite the way a compressor does meaning that if a sound goes below a set db level then the limiter will raise the volume on it. Is that correct. If not please explain

Reply

Adrian Redden April 2, 2014 at 3:49 pm

Rap

Reply

Adrian Redden April 2, 2014 at 3:53 pm

Rap!

Reply

clash of clans hack free download no survey 2013 April 16, 2014 at 4:13 am

On a methodological echelon HTML5 is of course just
a arrangement for a mark-up language. megalith megalops
megaplex megapode megapods megasses megastar megatons.
coassume coasters coasting coatings coatless
coatrack coatroom coattail.

Reply

Leave a Comment

Previous post:

Next post: