COMPRESSORS AND LIMITERS CAN MAKE YOUR MUSIC SOUND LOUDER; THEY CAN ALSO MAKE IT SOUND AWFUL!
Compressors and Limiters are a way of life in audio. Without them, listeners would be constantly reaching for the volume knob. But they can also really mess things up. If you understand how they work, chances are they won't be messing you up. So here we go. Sorry for the math, but it's necessary.
You put a compressor in the master buss. You set the "threshold" to -6db and the "ratio" to 2:1; anything above -6db will be only half as loud. That means that if the signal has a 0db peak (6db above your threshold of -6db), the compressor will only let that peak reach -3db (only 3db above your threshold of -6db). So your song won't have any peaks louder than -3db in it. In other words, any peaks above -6db will only be half as loud. Your song is no longer reaching the 0db digital ceiling.
(NOTE: READ THE ABOVE PARAGRAPH A FEW TIMES UNTIL YOU GET THE MATH THAT'S BEHIND IT.)
But hey, you're doing this to make it louder, and all the compressor did was make it quieter. But - and here's the magic part - you can now turn the whole song up by 3db. Those loud peaks, which previously determined how loud the song could be, are now 3db quieter. Turning the song up by 3db will raise the level of all the other information, bringing the peaks back up to 0db, and you can now hear the entire song a lot louder and clearer.
This is a waveform of raw vocal. Note that the third peak reaches the ceiling of 0db:
Here's the same waveform after compression. All three peaks were brought down in volume, and then the overall volume was increased:
Note that all three peaks now reach 0db. More importantly, note that the rest of the waveform is considerably louder. So the quieter parts are much easier to hear now.
That's how compression works. (Limiters do the same thing, but their ratios are set to infinity:1. That means that NOTHING gets louder than the threshold.)
Compressors are used for all types of things, like taming a vocal that has wildly varying levels or an acoustic guitar that's sometimes slammed and other times strummed lightly. It helps keep instruments in balance during mixing. But if you don't know what you're doing, compression can either ruin a song or not make any difference at all. Here's some examples:
1) THE COMPRESSOR'S NOT DOING ANYTHING. Your threshold is set to -6db, but the waveform never goes higher than -9db. Since the signal never goes above the threshold, the compressor's not doing anything!
2) YOU'RE OVERDOING IT, AND THE WAVEFORM IS DISTORTING. You're compressing so much that the waveform is squished. Here's the same waveform from above, but over-compressed:
This waveform sounds terrible. It may have enough distortion on it that is sounds like you're putting it through a fuzzed-out guitar amp.
3) I CAN HEAR THE MUSIC CHANGING VOLUME. The release time is too long. The peak triggers the compression, and the volume stays low for too long. You can hear the overall volume slowly rise back up (they call this "pumping", "breathing", or "ducking"). You've probably heard this happen on the radio, because most broadcasters compress the heck out of the signal.
Here's the raw vocal again:
And here it is really squished with a long release time:
Notice that on the raw vocal the signal dies down smoothly after the third peak, but on the squished vocal, it ducks down and then slowly rises to a loud level after that third peak. If you were listening, you'd think someone was messing with the volume knob. This takes all the life out of music and should be avoided at all costs.
You can fix this situation by adjusting the "release time". Make it shorter until you no longer hear the music "pumping" in volume.
Compressors and Limiters are necessary tools in recording. If an album doesn't use them at all, you'll be changing the volume knob constantly, because one song will be too quiet and the next song will be too loud. But using them too much (like during the "loudness wars" of the 1990s) can cause ear fatigue. A "squished" song is very hard to listen to at a decent volume level.
I may hold a little "workshop" at my studio in the near future to show you how compressors work, because there's nothing like turning a control and hearing (and seeing) what it's doing to the signal. Let me know if you're interested in attending.
A WORD ABOUT USING VINTAGE COMPRESSORS (EITHER OUTBOARD UNITS OR SOFTWARE EMULATORS):
Old-school rackmount compressors from the fifties and sixties were finicky beasts; many of them lacked all but the most basic controls. (Some of them were called "one-knob squeezers".) These compressors were needed in order to overcome the inherent noise floor of analog tape or AM radio. If the signal was too quiet, you couldn't hear it because it would be buried in the hiss and hum. With 24 bit digital recording, this is no longer a problem.
These compressors were made with tubes, which completely colored the signal and gave it that old-school vibe. You should consider these as special effects units and treat them as such. Some people put their microphones through them while recording because they know how they work and they can adjust the controls to their liking. Other people put their microphones through them because they've heard that they sound cool, even though they have no idea how to control them or what they're doing to the signal. Bad idea.
If you're going to use old-school compressors when tracking or mixing, don't just plug them in blindly. Turn their knobs and see what they do. Make sure you've mastered how they work before committing your audio to them. One of the reasons that the first two Beatles LPs sound bad in stereo is that Abbey Road had just bought a stereo compressor. It took the engineers a year to notice that there was a "fast / slow" selector switch on the compressor. It had been on "slow" all through 1963 and into 1964. He set it to "fast" when mixing the "A Hard Day's Night" album, and he was floored by the difference. So there you go.
DAWs (such as Digital Performer or Pro Tools) have basic compressors that are more surgical in nature and have great graphical interfaces. They simply change the volume without "coloring" the signal. These are the compressors I prefer, because I use them to solve problems, not to change the overall sound. When I master a recording, I usually add 2 db of gain with a 1.5:1 compressor, followed by 4db of gain using a limiter which is set to limit the peaks very quickly. This gives me a 6db overall boost in volume, which makes for a hefty master that can hold its own against professional recordings in a playlist. It still sounds like what I was shooting for; it's just louder. (I adjust the attack and decay times as needed, based on the music.)
Digital Performer has a compressor called a "Masterworks Leveler". It looks and feels old-school. It's also extremely difficult to control and doesn't really add color, so although I was excited to see it when it was introduced, I don't even consider using it anymore.
The best thing about the compressors in today's DAWs is the graphical interface. You can SEE where the waveform is getting compressed to your advantage numerically and then use your ears to judge if the signal sounds better or worse.
A PARTING WORD TO EXPERIENCED RECORDING ENGINEERS:
If you've been recording for awhile, you probably already know how compressors and limiters work, and you may find fault with my attitude towards vintage compressors. This page wasn't aimed at you - you already know what you're doing. This page is for the uninitiated recordist who may not be using compressors all that effectively. I have no problem sitting down and arguing the merits of opto-compressors over a few beers....
THE TAKEAWAY: COMPRESSORS AND LIMITERS CAN MAKE YOUR MUSIC SOUND LOUDER; THEY CAN ALSO MAKE IT SOUND AWFUL!
back to index