As I was preparing a recent OptEQ workshop, a few things crystallized for me, the most fundamental being “what is equalization?” If you ask 10 audio practitioners this simple question, you’re likely to get 10 different answers.
In the most literal sense, equalization means to “make equal.” But make what equal? Here are some thoughts for consideration.
In the strictest use of the term, equalization applies to the correction of minimum phase aberrations in the transfer function of a device (usually electro-acoustic) using minimum phase filters. This could be called “corrective” equalization, because the anomaly is removed, as though it never existed.
This is an “audio world” definition. Unfortunately, we don’t have exclusive rights to the term. Even so, some practitioners want to limit the term “equalization” to this use of filters only.
The filter banks, analog or digital, that precede the power amplifiers are called “equalizers.” Graphic or parametric, they’re used for spectral shaping of the response. These filters may be used with no regard as to whether they’re “corrective.”
The channel strip of a mixer includes filters for “program equalization.” These can be technical, such as a high-pass filter used to band-limit a microphone, and they can also be artistic, used to change the sound of a vocal or instrument microphone. This is always done by ear, and it is subjective, yet “program equalization” is a very important part of the operation of a sound system. It gives the system operator some control over the way things sound without touching the “house EQ,” which should only be used to establish a neutral system response.
Some equalizers have intelligence and can establish their own response curve based on an ongoing measurement. “Adaptive equalization” is very much a part of modern communications, and is used to compensate (not correct) for the dispersion (time smearing) of signals in a communications channel. This may be the most common form of equalization that most of us encounter, given the proliferation of cell phones and digital communication links.
Dr. C. Paul Boner applied notch filters to compensate for room resonances. A resonance is a frequency that a room stores longer than others, providing some natural amplification. This in turn produces tonal coloration for the listener. The use of room resonance compensation filters is probably the origin of the term “room EQ.”
Of course, these filters are not corrective, and those that see equalization as only corrective are greatly distressed by the suggestion of electronic room equalization. Since the term “room EQ” isn’t going away, expanding the definition of equalization to include “compensation” makes the term work.
In light of these facts, a general definition is in order. Audio equalization is the application of filters, either analog or digital, to the audio signal.
“Equalization” can be made more specific with a descriptor, such as corrective, program, or adaptive equalization. So, there are many forms of equalization. This is one case where Wikipedia gets it mostly right: “Equalization is the process of adjusting the balance between frequency components within an electronic signal.” This “balancing” can include magnitude, phase, or both. The target response may be flat or it may be a curve, such as the “EQ” setting on a media player.
To limit the term equalization to any one of these specific applications causes confusion, as we are then required to create new terms for the others.
SynAudCon recognizes the “application of filters” as equalization, and filter banks as “equalizers.” We teach the specific uses of equalizers in our seminars, and we encourage audio practitioners to make their intended meaning of the term clear by the context in which they use it.
More info on SynAudCon’s OptEQ course is available here.