Guitar 2 sine wave filter question

Started by Marek, January 14, 2005, 08:55:20 PM

Previous topic - Next topic

Marek

Hi there!

Which filter combinations should I use to make the guitar signal as close as possible to a nice clean sine wave?  (apart from a simple low pass).

Peter Snowberg

A simple sine wave contains only a single frequency so there is really no amount of analog filtering that can cleanly isolate the funamental from an electric guitar signal. The fundamental is not even present as a clean sine wave either, just to complicate things.

I think simple lowpass filtering is the right direction to head, but another way to look at the problem is, "what extra information can I get away with leaving in the original signal and still do what I want?".

If you can expand on what you're doing with the filtered signal, more people may be able to add answers. :D
Eschew paradigm obfuscation

Paul Perry (Frostwave)

Accepting that NOTHING will help to extract a sine wave from a chord(!), if you are trying to turn a single note to a sine, working on the assumption that the freq you want is the fundamental, I would run the signal into a low pass filter then into a compressor.

lovekraft0

A clipper followed by an integrator and a steep lowpass filter will get you close for single notes (frequencies), but it's going to be nearly impossible to implement, because both the integrator and the filter are going to have to track the input frequency to keep the results consistent. Maybe you could use a PLL with a digital counter/divider in the loop, and use the VCO output to generate a digital approximation  from another counter with voltage dividers on each output, or a ROM and a DAC. Sounds like a lot of work, though.

Nasse

Years ago there was a project in (maybe) British electronics mag, don´t know if I still have that paper... :oops:

Of course it was only for single note playing, pitch tracking was done with phase locked loop and National MF-10 or something switched cap filter...

Lots of parts if I remember...
  • SUPPORTER

Paul Perry (Frostwave)

Nasse, the trouble with those PLL approaches is the problem of capture time. Also, if the signal isn't "well behaved" (that is, full of harmonics!) then things don't always go as well as you might hope, once you lock in things are OK but acquiring lock can be rrrrough :?  :shock:  :x
The lucky thing with a guitar signal, is that at least there are unlikely to be harmonics BELOW the one you want!! (in the single-note case).

Marek

Very interesting. I was comparing the AC (230V) filtering with a guitar signal. Both are AC... So, how do we go about  filtering the AC line and why cannot we use the same principle for a guitar signal?

BTW, I had no project hiding behind this question. I just wanted a nice clean sine wave from my guitar. Kind of synthish sine tone.

And... If you are really interested, I am currently designing a circuit which will have bass & treble separation @ the input, and then there will be a distortion part which will distort only the treble spectrum, leaving the bass part undistorted. At the output you mix the two. So you get some kind of exciter or enhancer or whatever... :-)

Greetings,
Marek

StephenGiles

Marek - try this for size. I guarantee a smooth sinewave where indicated!

Stephen
"I want my meat burned, like St Joan. Bring me pickles and vicious mustards to pierce the tongue like Cardigan's Lancers.".

puretube

the trick of AC (230V@50Hz) filtering is:
cutting everything off beyond 50Hz;
can`t do that with guitar;
you can cut everything off beyond ~1500Hz for guitar,
but then you still got lots of harmonix (=not sine)
of every note below those 1500Hz - the lower, the more "unsine".

the best trick yet to start (if you don`t wanna go Stephen`s fundamental extractor), is: neck pickup with tonecontrol rolled down...

Ge_Whiz

And pluck the strings exactly mid-way between the fret and the bridge.

Paul Perry (Frostwave)


Marek

Thanks, everybody. I'll stick to the good ol' lowpass then...

Greetings,
Marek

Paul Perry (Frostwave)

Stephen, I can't for the life of me work out what is happening in that EHsynth circuit, is it a phase locked loop with a triangular VCO that goes to a sine shaper??... as for where A to M connect  :?:
It certainly doesn't look like an 'early' EH design..