Freq Filters Something New I Learned Thought I Would Share

[ansil]

So a lot of you may know i do some wacky stuff with guitar electronics, and well that's just the beginning. However i recently had the chance to measure some interesting things and thought i would share them with you.

here we have two basic filters a low pass on top and a high pass on bottom. For simplicities sake lets say that the low pass filters only allow signals lower than the corner freq to pass through unattenuated. the opposite could be said for the high pass filter. since this is a basic passive design we are looking at 6 decibels per octave. what does that mean? well since this circuit is passive ie no batteries no gain devices it can not add to your signal only remove, although this is not a bad thing it does limit us to this. lets say we have a high pass filter that is set at 1kz or 1000hz. that means that everything below this will get its perceived volume cut at 6db per octave. so at 500hz we will be down 6 db. at 250hz 12db and 125hz 18db ie etc and so on. the same holds true for the low pass filter only again in reverse. a 1k lpf will get you -6db at 2k -12db at 4k again ie etc.

now for those of you who suffered through this here is the interesting part. when designing amps, effects anything that is active these frequencies are crucial and make all the difference in the world to our tone and the parts and the math have to be pretty close when you are shaping your sound for a desired effect. Say controlled resonant feedback that is freq dependent. but when you go passive the amount of signal drop changes the perceived eq shift as well.. i will give you an example. in a guitar per chance lets say we wish to roll off 200 hertz ie a high pass filter because the guitar is way too woofy in the neck pickup. and any adjustments can be compensated at the amp. this trick really helps you balance those pesky dual humbucking guitars without demoing hundreds of pickups to find the right one. perhaps the right one is already installed and you just need to help it out.

what i have found and measured is that because of the 6db drop it causes the exact freq response but to get the correct "ear response" we need to take it 1 octave lower. so by moving our filter range to 100hz will give us a more usable freq shift and allows us to do things inside our guitars to tune their response more to our liking

well hope that helps you out. feel free to comment and please try the experiment

sincerely

ansil

and lastly for those who might have missed it this means you can have a freq based gain control with a little math in our guitars and with complex enough switching you can have both a straight signal and a modded one. lets grab those irons guys [and gals if there are any left here lol]

[Ripthorn]

I built a tele last year with low pass, high pass, band pass, and notch filters all on board with selectable caps using different combos of the caps and inductors. My next idea is to try muli-pole filters onboard just for the heck of it. I think a 3rd order Butterworth or perhaps a Bessel filter would be a lot of fun. Filter design was one of my most enjoyable classes and I find that I use it a lot, especially working on radar by day and guitars and amps by night.

[dude]

Good info Ansil. That will definitely make tuning an EQ to SOUND good much easier.

[ansil]

Thank you I hope it will help you out. Something I remember from a long time ago. This guy (shoot I can't remember his name now but it's unique) anyway he showed me his eq and it wa interesting he had a weird subtle midrange hike that I didn't get ans4 I was like what is this? He was telling me that in recording that anything above the zero line would just give you a boost and4 not doexactly what you think. I was like yeah sure. And then he showed me. And I wa blown away. The difference was mind boggling

[dragonbat13]

I wonder if it would be possible to put a bandaxall type setup in a strat. Something like running a volume pot that will allow more treble through (I think that would be a 500K) and then having a bass and treble for the tone knobs.

[ansil]

yes but not a true baxandal az it eats quite a bit of signal although i am not saying it couldn't be done. but yeah a treble and bass pot on it is simple enough

[dude]

Ansil, I was just re-reading your initial post. The "ear response" bit about choosing an octave beyond the theoretically "proper" cutoff frequency actually makes sense. The cutoff frequency for first-order filters like these is not actually the frequency at which the signal begins to roll off, but is rather the "-3dB point": the point at which the signal has been rolled off by 3dB (about half). Now I believe that because power-as-a-function-of-voltage (signal size) for a passive load is quadratic, and loudness-as-a-function-of-power is inverse quadratic, we can say that loudness-as-a-function-of-voltage (signal size) is linear. Someone will have to check my assumption here, but what this means is that 3dB of attenuation actually equates to half the original loudness. When move the cutoff frequency further out, you are actually ensuring that your original cutoff frequency is actually passed through 100% instead of 50%, which makes more sense in terms of what would sound "correct" compared to what you expect.

Also of note: If you choose a cutoff frequency of 1kHz for a high-pass-filter, your attenuation at 500Hz will actually be 9dB instead of 6dB. Bear in mind that the attenuation isn't linear on the other side of the corner frequency, so the 6dB/octave rule doesn't hold there (though it makes a good approximation in a pinch ).

Lots of math there. Feel free to prove me wrong. Sometimes I say things to sound smart only to have somebody destroy all my claims and make me look foolish

[Prostheta]

Here is a particularly cool and very very useful filter circuit as found in the Aria Pro II SB-1000. It is active, however it sold me on varitones.

The varitone taps off treble with a mild resonant hump where the corner frequency sits on a simple RC circuit - it sounds great on bass. Some of the more extreme settings are very dark and bassy however it is far more characterful than a simple RC filter. I figured that I might make the control continuously variable using an op-amp based gyrator but haven't had the time to dedicate to it. Maybe one day.

[ansil]

yes dude that Is one of the things I really like about it. it has really opened up my neck pickup. although I could stand a little more gain out of it the sounds I get are amazing. it would be interesting to plot out of chart for this. but that will be another time. where I can set my shop upin the new house

[ansil]

Here is a particularly cool and very very useful filter circuit as found in the Aria Pro II SB-1000. It is active, however it sold me on varitones.

I just have 1 question where is the coil is it in the black box able tire section

The varitone taps off treble with a mild resonant hump where the corner frequency sits on a simple RC circuit - it sounds great on bass. Some of the more extreme settings are very dark and bassy however it is far more characterful than a simple RC filter. I figured that I might make the control continuously variable using an op-amp based gyrator but haven't had the time to dedicate to it. Maybe one day.

I just have 1 question. where is the coil. is the pick up the coil. if so then where do we input. if not then is the coil inside of the black box that house is the amplifier. and what would the value be? I am viewing this on my mobile though maybe I'm missing something in the drawing

[dude]

I think I saw one of these in the junk drawer at my local music shop (they have a bin of broken stuff they sell for cheap). My buddy and I could NOT figure out why there was a rotary switch switching between a bunch of filters. Looks like I should have looked closer. Though in this case the "black box" was just a fet.