Digital audio processing implies nonlinear processing. The reason is that the latter’s foundation is in linear time-invariant systems. Many technical systems involve both linearity and time-invariance, especially those where input and output signals fix to a specific amplitude range. Many analog audio processing devices have nonlinearities, including analog effect units, valve amps, analog tape recorders, loudspeakers, and the human ear. Nonlinearities usually need nonlinear signal processing as well as a solid background regarding nonlinear systems.
Nonlinear Processing and Compressors
Compressors are a perfect example of nonlinear devices. To state things simple, when a signal passes through a nonlinear system, a certain level of distortion appears. It’s important that the user understands that when the system is less linear, the distortion is proportionally more profound. Harmonic distortion is typical; the user can add it to an existing frequency content of an audio signal. Moreover, analog components cannot be completely linear. In contrast, digital systems tend to be totally linear and free of harmonic distortion. Total Harmonic Distortion, or THD, is a measure of the harmonic distortion content by an analog device.
Analog Distortion
Analog distortion comes in many shapes and sizes. The ratio between low-order harmonics generated by a tube is completely different from those that produced by a transistor. The ratio is a key factor regarding the diversity of sounds that solid-state equipment can tune and produce. Technically speaking, it’s better when the distortion level falls to a lower level because harmonic distortion is an essential element of the analog sound.
Inter-Modulation (IMD)
As stated above, there are many types of distortion. Another example is inter-modulation, commonly known as IMD. It is a measurement that involves a variety of frequencies that are not always harmonically related to sound. Although inter-modulation is essential within every nonlinear system, it can also be harsh and undesirable.
Nonlinear Processing: Phantom Partials
Nonlinear mixing is another typical example of nonlinearity in the audio processing universe. Engineers frequently use it when producing phantom partials in piano tones. Regarding this, phantom partials are anomalous overtones in the spectrum of the piano sound. This spectrum occurs at sum and difference frequencies of the natural overtones of the string. Within this definition, nonlinear mixing can happen because the tension varies during a transverse vibration. In turn, this generates longitudinal string forces of phantom-partial frequencies.
Phantom partials are present in almost every plucked-string or struck-string musical instrument. You can find them in piano tones, the motion of a piano bridge, the longitudinal vibrations of a monochord string, the acoustical output of a soundboard coupled to a monochord string, and in the acoustical output of a guitar. These are just some examples. They are essential when trying to distinguish the timbre of tones that at a given frequency.
