Graphic or parametric EQs

[Ell]

Has anyone got any experience with graphic or parametric EQs.

I was tempted to try to build a clone of the Boss GE-7, but the schematic is a bit bewildering. There are times when paths cross with no dots, other times when they cross with black dots, and other times when they pass each other with white dots. Does anyone know which of these options signifies a connection?
It's a pretty complicated schematic at first glance, but I think a lot of it is repeated, and some of it is for the switch.
If I attempt this, it could be cool to figure out a way to change the Q, and maybe even sweep the frequencies, so that it's a graphic-parametric EQ all in one. Am I insane?

[Rob Strand]

The non-dotted crossovers don't connect and the dark dotted connections connect.   That's all pretty standard stuff.

The thicker curved/dashed lines are not wires,  they delineate boundaries of PCBs.   The open/white dots indicate wire connections, either from a plug or between the two PCB boards.  The thin solid lines to those open/white dots are wired connections.

[Ell]

The non-dotted crossovers don't connect and the dark dotted connections connect.   That's all pretty standard stuff.

The thicker curved/dashed lines are not wires,  they delineate boundaries of PCBs.   The open/white dots indicate wire connections, either from a plug or between the two PCB boards.  The thin solid lines to those open/white dots are wired connections.

Superstar! Thank you so much Rob!

[Vivek]

If I attempt this, it could be cool to figure out a way to change the Q, and maybe even sweep the frequencies, so that it's a graphic-parametric EQ all in one. Am I insane?

No not insane !!!


I would think that in case you made a "Semi parametric EQ"

You would need much fewer bands  (Much fewer is an oxymoron ?)

Maybe an 7 band GEQ can be collapsed to 3 or 4 fixed and 2 semi-parametric EQ and actually have better functionality


Boss metal zone's tone controls are interesting !!!

Check brother ELECTRIC DRUID's work here:  https://electricdruid.net/boss-mt-2-metal-zone-pedal-analysis/

[Vivek]

Gyrator calcuators
http://www.muzique.com/lab/gyrator.htm
http://tools.griffineffects.com/gyrator.php

Essential reading
https://sound-au.com/articles/gyrator-filters.htm

[duck_arse]

.....  (Much fewer is an oxymoron ?) .....

not so. much being a sizing, fewer being a direction. is that a little more clear?

[antonis]

oxymoron

<spam on>

" οξύ-μωρον" -> sharp/acute - dump/idiot(*)

(*) "ιδιώτης": person who takes care about his own personal profit at common welfare expense

<spam off>

[Mark Hammer]

Graphic EQs are for situations in which the device-maker has no idea where, or how many, any resonances or weak frequencies might be located.  This is why we see not only 7, 10, and 15-band units, but 31-band EQs as well.  If the maker doesn't know the specific application, they try to cover all bases.

When it comes to guitar and bass, there are generally not many resonances one wants to alter, whether to "tame" a misbehaving guitar/speaker combination, or to simply alter the voicing of what results.  Many pedals and amplifiers will address this by providing a sweepable midrange, which is essentially a single semi-parametric stage.  I've been contending for a number of years that the "ideal" guitar EQ unit would have a shelving bass control (i.e., standard boost/cut at a fixed frequency), two semi-parametric sections/stages, and a variable two-pole lowpass filter, with a switch to select between a couple of Q-values for the filter.  This is essentially a 6-knob unit.

I presented this to Joel Korte (Mr. Chase Bliss) over dinner a couple years ago, and he liked it.  He tinkered with the idea, improved upon it, and the CBA Condor was born.  Because of the limits imposed by the form factor of CBA pedals, and likely internal space constraints, two sweepable mid-bands was reduced to one, but the rest was retained.  If a person is not intending to incorporate all the other powerful features of a CBA pedal (including microcontroller, presets, and MIDI capability), then 2 semi-parametric bands becomes quite feasible in a 1590BB or 125-B, and provides an enormous flexibility in a small package.

[Ell]

Graphic EQs are for situations in which the device-maker has no idea where, or how many, any resonances or weak frequencies might be located.  This is why we see not only 7, 10, and 15-band units, but 31-band EQs as well.  If the maker doesn't know the specific application, they try to cover all bases.

When it comes to guitar and bass, there are generally not many resonances one wants to alter, whether to "tame" a misbehaving guitar/speaker combination, or to simply alter the voicing of what results.  Many pedals and amplifiers will address this by providing a sweepable midrange, which is essentially a single semi-parametric stage.  I've been contending for a number of years that the "ideal" guitar EQ unit would have a shelving bass control (i.e., standard boost/cut at a fixed frequency), two semi-parametric sections/stages, and a variable two-pole lowpass filter, with a switch to select between a couple of Q-values for the filter.  This is essentially a 6-knob unit.

I presented this to Joel Korte (Mr. Chase Bliss) over dinner a couple years ago, and he liked it.  He tinkered with the idea, improved upon it, and the CBA Condor was born.  Because of the limits imposed by the form factor of CBA pedals, and likely internal space constraints, two sweepable mid-bands was reduced to one, but the rest was retained.  If a person is not intending to incorporate all the other powerful features of a CBA pedal (including microcontroller, presets, and MIDI capability), then 2 semi-parametric bands becomes quite feasible in a 1590BB or 125-B, and provides an enormous flexibility in a small package.

Call it the Anvil! It's how Mr Hammer shapes his tone!

But seriously, that does sound like the perfect EQ for guitarists. I read a reference to this idea of yours on the R.G. Keen article about EQ, but most of the article is beyond my current level of understanding. I think this is going to take me a long time to digest and properly understand all of this...


So, if I understand this right:

Bass boost/cut control (would a simple 2 way switch be good enough?)
Two semi-parametric sections/stages (in the Geofx article it quotes you as suggesting between 700Hz and 1500Hz)
Lowpass filter with Q controls (I don't understand why this is a low-pass filter and not a high-pass filter, when we already have a bass boost/cut control?)

I have been wondering if using a treble booster for the higher frequencies would be good?

As I mentioned in another post about a volume drop in an Octave Multiplexer, I would also add a looper to all of this. A couple of stereo jacks that can optionally put another pedal before or after the EQ.
For example, I recently built a couple of easy-vibes (one includes a switch that swaps to a phase 90) based on your advice, and there is a very different tone when the easy-vibe is engaged, it's almost mid-scooped. I was thinking that the EQ/looper pedal could fix these sorts of problems, or be used as a sort of second channel with FX.

[PRR]

> figure out a way to change the Q, and maybe even sweep the frequencies, so that it's a graphic-parametric EQ all in one. Am I insane?

Optimistic. Every serious feature you add to an EQ adds at least one opamp; and may lead to features-Squared (3 features 9 opamps per section).

Variable Q withOUT mucking the frequency or gain adds at least one and maybe several opamps. HIGH-Q adds opamps.

RANE has papers on this; I don't recall where the new owners put them.

The BOSS GE7 is fine for its time and place, and hard to beat today with analogy stuff. If you make the hump to climb into a digital processor, you "can do anything"; the limit being knobs and display not signal-hammering. (Even in the year 2000, CoolEdit 2K's EQ functions kicked all my post-process analog EQs to the gutter; and in the decade since the CPUs could do it in real-time (tho not in CE2K because it was post-hoc and also sold-out to Adobe).

[Mark Hammer]

Why a lowpass filter with selectable Q?

A few reasons.  First off, it allows for adjustment of how much "bite" a guitar will seem to have.  It adjusts tonal character MUCH better than the stock Tone control on the guitar does, and the steeper slope provides better "definition" of whatever tone you're aiming for.

Second, it serves as a much better shaper of what comes out of a drive pedal, and will do a better job of removing objectionable "fizz".

Third, one of the major differences between humbuckers and single coils IS the nature of the high-end rolloff.  A variable lowpass, especially one with a reasonable steep rolloff and variable Q, can go a long way toward simulating HBs on a SC-equipped guitar.  Note that increasing the Q will provide a small-but-audible peak around the rolloff.

[Rob Strand]

There some notes here, in section 2.17,

https://archive.org/details/bitsavers_nationaldaAudioRadioHandbook_17034677/page/n71/mode/1up


Notice how the amount of boost implied by a single band is much less when used in a multiband.   Single band design 17dB comes out as 12dB in a multiband.  Then later it mentions, an octave equalizer implies a Q of 1.414 but in practice Q=1.7 works out better.  And many instrument equalizers are even higher.

The design Q is the Q at full boost/cut.  When you back-off the boost or cut from max the Q is much reduced.

Multiband EQ's are best designed using a circuit simulator and for instruments by ear.