Equalizers (both graphic and other) are Baked. They're just filter arrays.
Also, there are wide variations in humans ability to hear sounds.
Now, for the serious audiophile, BorgCo offer a semi-automatic self-tuning equalizer unit.
The unit is installed between the sound source and the main amplifier. The user then minimises ambient noise as much as practicable, dons headphones plugged into the jack socket in the equalizer, and presses the start button on the remote.
A tone (Middle C, probably) is played into both ears simultaneously and the listener adjusts the volume to a comfortable level.
When they press "Next", the equalizer plays a sequence of volume-ramped pure audiometric tones, depending on which programme is selected. Some programmes may play as few as eight tones, others have dozens of frequency steps. In each case, the user is expected to press the "Enter" button when the tone is first audible, which stores the value.
At the end of the process the equaliser has a map of the user's hearing response, which can be stored in a named memory.
The equaliser then applies an inverse of the map to the sound source, theoretically producing a "flat" response.
Since the overall output volume is so variable depending on the amplifoer settings, the equalizer doesn't need to be calibrated in dBA; the map is relative rather than absolute.
The user can tweak individual portions of the map using the remote control to allow for room accoustics, music style, personal preferences etc.-- 8th of 7, Mar 06 2010 Equalization http://en.wikipedia...r#Graphic_equalizerThe basic technology [8th of 7, Mar 06 2010] Bose AdaptiQ http://worldwide.bo...echnology/page.htmlOne of several [csea, Mar 06 2010] Fletcher-Munson equal loudness contours http://en.wikipedia...%80%93Munson_curvesFrequency response of human hearing varies widely with level (perhaps not so for Borgs?) [csea, Mar 06 2010] I think somehting similar is baked. AFAIK, the sound they use is noise (white?/Every frequency)-- Dub, Mar 06 2010 [IT], that's to calibrate the listening environment. This is to map the listener's personal audio response too.-- 8th of 7, Mar 06 2010 Most likely, with a whimper rather than a bang.
But at least the whimper will be accoustically neutral ...-- 8th of 7, Mar 06 2010 Well, if you don't like that particular van Gogh, we've got a spare 'ere ...
<Python>
"Look! I'm the bloody pope, I am ! May not know much about art, but I know what I like ! "
</Python>-- 8th of 7, Mar 06 2010 I had my hearing response plotted by an audiologist using this method, I think it peaked around 3k which is fairly normal. Anyway - the problem is that your hearing response at the level of audibility (1db) is very far from flat and much more skewed than a-weighting (which gives a vague indication of equal loudness for normal listening levels).
I think the best thing to do is redefine the problem and make your machine an equalizer for people who like to listen to their music on the bounds of audibility. Come to think of it, there's a lot of music that I prefer that way anyway.-- wagster, Mar 06 2010 //the map is relative rather than absolute.//
As pointed out above, this could work at a fixed listening level at higher frequencies, but not for arbitrary listening levels due to the nonlinearity of human hearing frequency response vs. level [link].
And listening to loudspeakers in a room is fraught with difficulties at low frequencies where room modes can produce 30dB or deeper notches in response. Equalization _cannot_ be used to remove such notches, as they are caused by acoustical cancellation, and adding more energy generally makes the cancellation greater.-- csea, Mar 06 2010 random, halfbakery