The most I knew about distortion was just making things sound dirty and gritty, I really know little about how it's actually done within a piece of hardware. I do know that turning up the input heaps of a piece of hardware will create distortion depending on what kind of processor the hardware is. Then turning down the output can give you a reasonable level but with distortion. So I wanted to put this theory into action and try creating distortion through processors that aren't typically used for distortion, in this case compressors.
Distorting a Sine Wave
Just so I could visualise and understand what distortion actually looks like I tried running a sine wave through some hardware compressors. The first piece of hardware I used was a dbx-166xs compressor without really adjusting any parameters that create compression and just pumping the output gain. Here's an audio example of a sine wave being put through the various tests I'm about to go through. In this example I add the dbx compressor, then the M350, then I remove the dbx and leave the M350. For the example I've used a sine wave at 100Hz as it's a little easier on the ears than the 256Hz I've used during all the tests.
I was curious what kind of sound the compressor actually makes so without running the sine wave through it I just turned up the output gain to max and the picture below is a frequency spectrum graph representing the subtle noise that the compressor generates. It's pretty insignificant but when it's added to other sounds I'm sure it makes a hell of a difference.
Now what does a sine wave look like on this frequency graph, without anything distorting it. Well as expected it looks to be mostly hitting 256Hz. It does sort of cover a bit more of the lower end which could be due to the compressor adding some harmonic content, however we should be expecting to see most of the harmonic content in the higher end as harmonic frequencies are created a multiples of the original frequency.
Okay so this is more like it, in the picture below we can see tones of harmonic content being generated above the original frequency. This was achieved by cranking the output gain on the dbx all the way up to it's max. It sounds pretty cool, but not in a nice way, it isn't exactly pleasant. This is a pretty common attribute of odd distortion, as opposed to even distortion. Basically if the distortion is even and we have a fundamental frequency of 256Hz like our sine wave, then the 2nd harmonic would be double, 512Hz, and then the 3rd harmonic would be double the second, 1,024Hz, etc. If we take a look at the frequency spectrum graph we have here, we can see that the 2nd frequency is a bit higher than 512Hz meaning this is odd distortion. I can't see this changing with a different fundamental frequency, I think if anything it's the kind of distortion method I've used. I wouldn't expect to get even distortion out of a clipping signal through a compressor. Let's remember that I haven't activated anything on the compressor except for the output gain, just so we can get the sound of this compressor coming through. The input from Pro Tools for this test is set to 0, so there is no attenuation or accentuation to the sound. Let's see what happens when we do accentuate the amplitude of the input from Pro Tools.
So with the input amplitude and the compressor output volume cranked all the way up, the waveform develops all this extra harmonic content and we end up with a really dissonant and abrasive sound.
Now just to take things even further lets add another compressor to this chain and see what happens to the harmonic series.
Again, tones of harmonic content, but nothing really different. It sounds different, definitely a lot harsher but if anything we're just adding more to the harmonic content. In the frequency analyser we can see that the peaks in the higher end of the spectrum are a lot more dense and closer together.
This is all very interesting but how can I actually use this practically? Luckily a class mate that I was working on all this with suggested we run a synth lead that she had recorded through it and see what that sounds like. I would have absolutely loved to have put it through the chain I was just working on but unfortunately we had to leave the studio we were in at this point, so we went to a different studio and decided to use a EL8 Distressor instead. As you can see in the picture we've set the ratio to 2:1, turned the attack to around 2 and a half, and pretty much turned the release all the way up. Meaning that it's pretty much always compressing the signal we're running into it. Then, as per our usual techniques we boosted the input and output gain up all the way. Here's some examples of what the synth sounds like with and without this distortion.
Personally, I think it adds great texture to the sound and really brings it to life. I think this kind of distortion would be fantastic on something like a clean guitar signal, or even a vocal.
Over all, distortion isn't very complicated, pushing a signal hot through a processor of some kind creates distortion, I think that's kind of the end of it. If anything it's the different kinds of distortion that you get from different equipment that intrigues me. I can't wait to find all different types of gear and see what it sounds like when it's pushed to it's limit. I've definitely discovered some production techniques that I'm excited to try in my further endeavours.