In Mark Vail’s ‘the synthesizer’ I found a description of the Serge Resonant ten band Equalizer which puzzles me:
Its 8 middle bands are spaced at major-seventh intervals to avoid the very common effect of an accentuated resonance in one key, as will be the effect from graphic equalizers with octave or third-octave spacing between bands…
The text continues with:
The resonant Equalizer’s band spacing produces formant peaks and valleys that are similar to those in acoustic instrument sounds.
Can anyone explain why there would be the effect of “accented resonance in one key” when using octave or third octave bands in stead of major-seventh spaced filters?
If you consider the Moog 914 fixed filter bank it’s based not on octaves or third-octaves but half-octaves, the same issue. The two lowest frequencies it uses correspond roughly to the notes B natural and F natural, a half octave apart. Moreover the higher pitches correspond to B natural and F natural in higher octaves since they’re multiples of half octaves. So pitches near those two would tend to be emphasized while pitches in between (around D natural and A flat) would be de-emphasized. But there is only one major key that includes both the notes B natural and F natural, namely C major — everything else has either B flat or F sharp. So if those two pitches are emphasized, it tends to impose a perception of C major.
On the other hand the Serge Resonant Equalizer’s second and third lowest frequencies (the lowest seems to be anomalous) correspond to about B and B flat (or A sharp). These are present in both B major and F sharp major. However, the next frequencies correspond to A, G sharp, G, F sharp… There are no major keys with even just the first three of these, let alone all of them. Since they’re not repeating at octaves they’re not reinforcing the same notes in all octaves, and the notes being reinforced are spread across all major keys. So no perception of a particular key is being imposed.
The idea is to create something akin to the fixed resonant frequencies, or formants, present in the human voice and in acoustic instruments. The formants add character to musical sounds. But normally these formants are not at any sort of regular intervals and certainly not at intervals that repeat in each octave. If a luthier, even one with no explicit knowledge of or ability to measure formants, were to make a guitar that had formants at B and F in every octave I think they’d listen to it, they’d say “No no no, that’s all wrong”, and they’d modify it to sound better — by adjusting the formants to be less regular.
I should have mentioned that I know little to nothing about musical theory. So ‘D natural’ and ‘A flat’ and such do not say much to me. Sorry about that.
I think I understand what you mean, however. It is all about intervals and not wanting to repeat them in all octaves by having filter centerfrequencies that coalesce with the tonal components of chords. Thx.