typical IM distortion in a digital compressor
The general and most obvious effect of intermodulation components in audio signals is distortion of course – hence the concept of “intermodulation distortion” (aka “IM distortion” or simply “IMD”). IM distortion and harmonic distortion are two pairs of shoes and must be defined individually as already shown in the short essay about “myth and facts about aliasing” but more on this later on.
The existence of intermodulation components can affect the performance of an audio production in various ways. In the best case, IMD components are a desired artistic effect e. g. to obtain heavily crushed audio effect signals but in the worst and rather common case, they are one of the contributing factors which deteriorate the overall audio quality and might ruin a production.
“Intermodulation is caused by non-linear behaviour of the signal processing being used. The theoretical outcome of these non-linearities can be calculated by generating a Volterra series of the characteristic, while the usual approximation of those non-linearities is obtained by generating a Taylor series” – wikipedia.org
Since intermodulation is caused by non-linear behaviour (in fact, this is a necessary premise to obtain IMD – a linear system can’t produce intermodulation components at all), we have to carefully watch out for all kind of audio effects which actually causes non-linear behaviour. The most prominent audio effects of this sort are saturators, distortion boxes like amp simulators as well as all kind of dynamic processors. In the case of dynamic processors, the most common form of interference which leads to IMD is the amplitude modulation which typically takes place when the main audio path is modulated by the sidechain signal in a voltage controlled amplifier (VCA).
In digital systems, things are getting even worse. Since within a bandwidth limited system IMD can – and in most cases does – generate frequencies outside the by the DSP system represented frequency spectrum, aliasing occurs. Oversampling the DSP system is a common answer to this but in fact this does not solve the actual root of the cause: the IM distortion itself. A better answer would be to use a signal processing which tends to avoid the IM distortion right from the beginning.
In practise, the IMD side effects might appear very differently. Gently applied on rather dull signals they can even produce a sort of “exciter” effect and brighten the overall signal perception. The other way around, most implementations which do not try to avoid IMD tend to have such brightening effects but the other side of this coin is the perception of harshness, unfortunately.
Just as an example, most digital compressor implementations are still suffering from this and are typically featuring some mid-high frequency exaggerations which brightens up the signal but also making them sounding harsh. I’m pretty much sure for now that the fact that we can still spot the difference between some analog circuit designs and its digital model counterpart that much easily, is fundamentally related to the insufficient handling of IM distortion in the digital domain.
Related links:
Variety Of Sound 
Interesting read as always. A bit too techy for my level, but it’s good to have explanations about all that stuff.
“I’m pretty much sure for now that the fact that we can still spot the difference between some analog circuit designs and its digital model counterpart that much easily, is fundamentally related to the insufficient handling of IM distortion in the digital domain.”
Why? Analogue designs suffer from IMD too. It’s not simply a feature of digital processing.
thats what I’m saying in the text. But in the digital domain we have the bandwith limit on top and so the IMD causes aliasing.
Ah, gotcha. 🙂
THANK YOU GREAT PRODUTS