What is Compression...

 

Perhaps you heard of albums or tracks being "compressed" to make the sound better.  You may have also heard about audiophile albums boasting that "no compression" was used as a positive thing.  Huh? What gives here? We'll get to that and many other issues in this article, designed to make you fully conversant about the compression process and where to use and how much to use and when not to use.  We'll end up with a discussion of software vs. hardware compressors and when software is appropriate.

Compression, ideally, is an "invisible" sort of effect that can bring your audio material up to spec with professional recordings.  Most audio professionals do use compressors in every piece and sometimes on nearly every track in every piece. 

And sometimes compressors are overused.  Ever listen to a radio broadcast talk show and notice that when no one is talking you hear noise and hiss coming through until someone talks?  That's a compressor doing that.  Radio stations, especially those with weak transmitters, pump the gain so they can get every ounce of volume out of FM radio's limited bandwidth. They know that the loudest channel will attract and keep more listeners than the ones at lower levels.

And the same is somewhat true of the music we buy and listen to.  Top 40 music is always compressed, polished and buffed so when it comes across the radio or TV, even on tiny speakers, it's fully listenable and accessible.

3 Ways to Use a Compressor in your Studio

There are 3 places in the audio chain where compression can be used to enhance your work of art.  They are the recording chain, the tracking chain and finally the mixdown chain. We'll spend a little time on each one.
 

The Recording Chain

Here the compressor is put on a direct out or insert of the mixer which takes the microphone signal after it is boosted by the preamp. Other methods are to place the compressor "in between" a mic preamp and an audio interface, or on the inserts of an audio interface or preamp.  The purpose here is to optimize the material for the recorder.  You want to make sure all low volume passages actually do have a strong enough level where they won't bring in noise later, and you also want to stop and loud "peaks" from overloading the recorder's input, which will ruin the track.  That is compressor theory 101.

The Tracking Chain

Once you have your audio tracks recorded on your computer or multi-track, you will be in the process of tweaking each track to make it sound the best it can, in reference to all the other tracks.  Here the compressor is added as an insert on a mixer.  That is, the signal goes out of the fader, goes through the compressor, then goes back to the fader's channel.  If you recorded your vocals and acoustic instruments without compression, and you are mixing on an analog board, you almost certainly have to use one here to get the track up to spec.  This can be done in the computer sequencer's mixer with a plugin, in the multi track if it has onboard compressors, or you do at at an analog board on inserts or busses.  No matter how you mix, the idea is to get the tracks uniform, so you don't have instruments or vocals suddenly dropping out because they went soft on you.   You may also need to clamp down on those pesky peaks. Compression helps.  If you have a single guitar note that peaks 15 db higher than the rest of the material, for example, your whole track will have to be mixed 15db down which will definitely put it in the background. The compressor, by clamping down on that peak, allows the whole guitar track to be boosted higher in the mix, where it can, at least, be heard.

The Mixdown Chain

In the mix, a variety of compression techniques may be used.  Compressors can be put on busses or even on sends and returns to affect (and effect!) certain parts of the mix. 

An advanced mix technique is often called Parallel Compression, where the uncompressed source tracks are mixed in with the compressed signal coming back on a return or on a bus.  The advantage here is that the compressor fattens the overall sound yet the peaks (which come from the source signal) remain clear and "on top" of the compressed signal.  Parallel compression can work for drums and vocals, or anything really.  It can also be done with groups of tracks.

But sometimes there is a temptation to put the compressor on the master bus (the main outs) a compressor may be added here too, and can have a dramatic affect, for better or worse.  Some professionals advise against using compression here.  Particularly if you are sending the mix to a mastering house for cd replication, let them use their gear.  However, if this is a home cd production, you will have to master it yourself.  But again, more cautions.  See if your mastering software has any software tools for the finalizing task.  Mix to wave without compression and use a mastering processor there.  But if you are mixing down direct to a cd recorder or DAT and this is the last stop, then go ahead, compress the mix. If done properly, the whole thing will come out louder and stronger.

Dynamic Range at Mixdown

This is a good thing, right?  Not if you want to be loud.  Dynamic range is the difference between the softest passage and the loudest passage in a song.  Compression shrinks dynamic range.  it makes the soft part louder and the loud part the loudest it can be.  So.   Got to be Loud?  It's at this point where you would consider multi-band compressors, like the TC electronics Finalizer and brick wall limiters, like the Waves L2, or the UAD Precision limiter. These will let you use every bit of space in the audio bandwidth and you will be able to maintain consistent loudness.  Because those writing top 40 hits all seem to do this, you may need to go this route if that's your bag.  Want to be soft and loud? You might consider not using compression or just extremely light limiting at all at this stage and preserve the dynamics of the material.  Orchestral and ambient works benefit from this approach as it makes for great dramatic passages when the orchestra does get loud. This is where some producers boast, "no compression was used".  Of course, they are not aiming to get played on car radios around the globe.  

Hardware vs. Software Compressors

Just a few notes here. Most modern sequencers have software compressors these days. These come in many different styles and types and many of them sound quite good. However, these are mainly for post-recording. You apply them to an audio track or soft synth as a plugin, after the recording has been made. Software compressors do not help as you record, so they cannot limit the peaks coming off the microphone through the preamp and into the converter. Hardware compressors, on the other hand, when setup correctly, modify the signal before it is recorded, thus preventing the overloads that can ruin a take. If you don't want to use a hardware compressor here you simply have to be careful about overloads. With 24 bit recording you can record at a lower level to avoid overloads, however, it is a great idea to have the protection of a hardware compressor all the same.