When you have a compressor every problem looks like a nail. Or something like that.

True story: two days ago I was driving to the office of a label to drop off the final masters of an album I’ve been working on for most of this year. At the start of the drive I popped the CD into the car’s stereo for one last listen. Halfway through the drive I pulled the car over. I replayed a track that sounded just plain wrong. Well it didn’t sound any better just because I stopped the car. Damn. I got out of the car, took a walk around the block, got back in the car for a third listen. Then I turned the car around and went home.

At first I thought perhaps I had burned the wrong master take to the CD. But in fact, being distracted by other projects and life in general I almost let slip a bum mix.

I opened up the project for the song in question (I hadn’t done so in months so there was some scare that some upgrade or plugin changes might cause some problems but I was in luck, everything was there just as I left it.) I then proceeded to completely remove all the effects and pushed all the faders all the way down and started over with the original tracks. It was tough medicine.

I’m not proud of this story but the technical lesson was a necessary one and hopefully others can learn from it:

In the old days I used to painstakingly edit the volume envelopes in Sonar (then Cakewalk) and later ACID to get tracks to “even out.” All the way though it I kept thinking “there has got to be an easier way to do this!” Eventually friends, collaborators and the Web taught me about compressors.

According to the experts I was doing a disservice, imprecise and needlessly laborious, to the tracks by manually building these long complicated volume envelopes. When you have variation in a track’s dynamics (i.e. some parts are softer than others) you’re supposed to get out the dynamic compressor and it will auto-magically raise the levels of the low parts and lower the levels of the high parts. Brilliant!

So when I got the vocals tracks for over the Internet for this song and noticed a pretty drastic variance in levels (some words were whispered right next to some that were shouted) I pulled out the best compressor I know of and started smashing away at the knobs. In the end the loud parts were being squished so much to meet up with soft parts the results made the vocals sound, er, less than natural so I did (what else?) added a ton of delay and reverb. It still didn’t sound natural but at least now it sounded like the proverbial “I meant to do that.”

What bullshit.

A few months later, listening in the car it hit me harder than a side curtain air bag smacking me upside the head: the vocals sounded like someone was sitting on the singer’s chest while he lay on the floor of a marble bathroom. This was a beautiful song, sung beautifully in an arrangement I was otherwise proud of and I had really made a mess of the whole thing because I read somewhere on the Web that real mean use compressors.

So it turns out there is no shame is actually moving a volume knob here and there. It turns out the best way to avoid a sterile over-compressed sound is to apply the Bush Sex Doctrine: abstinence.

In reality I found a wonderful balance between the volume envelope and the compressor. The actual lesson learned here goes something like this:

The difference between good compressors and great ones is how effective they are manipulating the dynamic range of the input while maintaining the authentic nature of the original sounds. But no compressor is even designed to handle huge swings in volume. So unless you really (as in: honestly) meant to sound squished and unnatural use the not-so-macho volume knob before whipping out the compressor.

Or something axiomatically along those lines.