Understanding intermodulation

[EuphAural]

As I understand it so far, intermodulation distortion (IMD) is a product of multiple frequencies passing through a non-linear system. Apparently what happens is that the lower frequency modulates the amplitude of the higher frequency, which produces the sum and difference frequencies.

Now, when I see charts showing IMD levels, it usually involves rinning two sine waves through at different frequencies. It shows the IMD products pretty well, but I don't think it tells the whole story.

What about guitar (or bass or any other instrument really). It doesn't produce a sine wave. I'm sure you've heard the IMD products when playing chords; for example playing a fifth through a distorted amp produces a tone an octave below the lower note in the chord. A single note from a guitar consists of multiple frequencies though, which can be analyzed as the sum of multiple sine waves. When you play a note (say an open A-string) through that distorted amp, wouldn't that result in IMD also? There are multiple frequencies present, which is the requirement for IMD.

Simply put, if you play a single note on your guitar through an overdriven amp (or other distortion maker) will there be IMD products as well? Even though it's one note, there are still multiple frequencies present, which should produce some IMD. Is it just at a low level relative to the regular harmonic distortion? Does it have something to do with the harmonic series present? I know I can hear IMD with chords, but I'm not sure if it's there in a single note.

Thanks for any advice and explanation. For some reason I can't quite get the concept. I'd assume it would be there but I'm not totally sure.

[Chuck H]

I can't speak from an engineering standpoint on the issue, but your spot on. Of course there are sum and difference frequencies present in a clipped signal with multiple notes of greatly variable amplitude and frequency. You can hear it all the time in most highly saturated vintage tones. You'll hear terms like "beating", "ghosting", "swirl", etc. All possible artifacts of sum/difference harmonies that occur in highly clipped and compressed conditions. Now, as to IMD as it's recognized by most people discussing it!?! Since it's actually a commonality in clipped tube guitar amps I usually hear it as a reference to a particular tubes character. As in, "el84's have a lot of intermodulation distortion". Which is true. Too much and you end up with an A-tonal mess. But a little (or some) is just the spice of life WRT clipped signals. Sorry I have NO proofs or studies to offer. I deal with IMD all the time. But I've never troubled to do any testing and documentation. If an amp sounds bad, you tweak it until it sounds good. Research on the matter would surely be helpful. As in, what sorts of IMD sound good/bad and what can be done to manipulate it? But I'm not aware of any such research (not to say it hasn't been done).

Just my low tech $0.02

[Steve Conner]

In a single note, the frequencies are all harmonically related. This means that their IM products all end up at harmonically related frequencies too, so they just look like regular harmonic distortion. To see (or hear) IMD, you need to use two non-harmonically related tones.

That's why the chord voicings used by guitarists get simpler and more consonant as the gain is increased from jazz to metal. When a complicated chord is played with a lot of distortion, it disappears in a throbbing mush of IM products. In music theory terms, a single note is the simplest and most consonant interval of all, it will survive any amount of distortion.

Executive summary: IMD is just another way of looking at non-linearity by exploring what it does to dissonant intervals. It's not a special kind of distortion.

[Jazz P Bass]

That's why the chord voicings used by guitarists get simpler and more consonant as the gain is increased from jazz to metal. When a complicated chord is played with a lot of distortion, it disappears in a throbbing mush of IM products.

Here is an interesting Wiki article on Power Chords, to add to what Steve said.
Link: Power chord - Wikipedia, the free encyclopedia

[Chuck H]

There was a reason I mentioned "saturated vintage tones". This is where things really turn to mush WRT difference intervals. Modern uber gainers are designed to articulate better while clipping hard. Usually by graduating gain over many stages rather than just a couple and by carefully tuning which frequencies are most distorted. That's why grunge metal, with it's lower 9's and 7's added to five and six string chords, sounds more articulate on a modern amp than a vintage amp. And this is why I mentioned that a better understanding of IMD as it relates to musicality could be beneficial. As subjective as that seems. But I love an old Marshall cranked to uber gain levels as much or more than most modern amps designed for such gain.

[EuphAural]

Thanks for explaining things a bit better. The fact that all the extra frequencies in a single note explains why you don't get beating pretty well. I think the 7th harmonic and up are pretty dissonant, but I doubt they show up in any strength.

I agree that knowing which chords to use can be very important. Especially at higher gains. A lot of those notes sound less than pleasant together. That does explain why metal guitarists like power chords more.

I also remember reading somewhere that in metal (or other high gain genres) bassists should drop tune if the guitarists do. Apparently it is because of the octave down tone that a heavily distorted power chord makes.