Guitar string processor?

Hex pickup (one sensor per string).

Sampling: 8KHz

I need to sample the frequency, modify it and output the modified signal with minimal latency.

Would the P2 be up to this?

Comments

  • PublisonPublison Posts: 11,232
    edited 2020-01-07 - 19:32:04
    Looking to do a MIDI interface?
    https://www.roland.com/us/products/gk-3/
    I had a Godin with that kind of pickup, but I hated the 13 pin interface.
  • I once suffered a $150 parking ticket because I lost track of time while in the [then] Carvin store in Hollywood trying one of the guitars with MIDI output -- very cool. I have been watching the P2 and wth the arrival of my demo board will start working with it; specifically for audio-oriented projects.
  • JonnyMac wrote: »
    I once suffered a $150 parking ticket because I lost track of time while in the [then] Carvin store in Hollywood trying one of the guitars with MIDI output -- very cool. I have been watching the P2 and wth the arrival of my demo board will start working with it; specifically for audio-oriented projects.

    Jon,
    What MIDI connector was Cavin using? 5 or 13 pin.
    I just sold my Casio MG-510 recently that had the only 5 pin standard MIDI connector that I know of.
    https://www.ebay.com/itm/352887871058

  • PublisonPublison Posts: 11,232
    edited 2020-01-07 - 19:34:31
    Mickster, I hope we are keeping on topic. It all relates I believe. I think with the smart pins, it is a reality.
  • > @Mickster said:
    > Would the P2 be up to this?

    Without a doubt... 6 strings @ 8 kHz and realtime MIDI out shouldn't make the P2 sweet using a single cog in PASM . Goertzel, asynchronous UART, and ADC on every pin is available in hardware. This would actually be a very good demonstration of what the P2 can do using a single cog. I think you could go a lot higher than 8 kHz for better precision.
  • Did you mean a "common" guitar processor with 6 induvidual channels? Yes that sound be doable as well. I think quite advanced processing at hifi frequencies in a single cog should be possible.
  • Publison wrote: »
    JonnyMac wrote: »
    I once suffered a $150 parking ticket because I lost track of time while in the [then] Carvin store in Hollywood trying one of the guitars with MIDI output -- very cool. I have been watching the P2 and wth the arrival of my demo board will start working with it; specifically for audio-oriented projects.

    Jon,
    What MIDI connector was Cavin using? 5 or 13 pin.
    I just sold my Casio MG-510 recently that had the only 5 pin standard MIDI connector that I know of.
    https://www.ebay.com/itm/352887871058

    As I recall, 13. I don't think Kiesel Guitars makes those anymore.

  • Nothing to do with MIDI.

    That horrible Auto tune that we hear on every new pop song? Well the big name is Antares.

    Well they did the same for guitar. I have 3 of these things and absolutely love them.

    The problem is that the one developer retired and so they killed the product.

    The strings might be perfectly tuned but any pressure on the fret and the note goes slightly sharp. The processor figures what the frequency is supposed to be and corrects it.

    I can also strum all strings open, followed by a button press and the processor compensates for any out-of-tune strings.

    The possibilities are endless; one can be playing in standard tuning and part way through the song, switch to an open tuning and kick-in with the slide.

    I didn't think for a minute that I had a critical ear for this stuff but it's almost painful to go back to a regular guitar because perfect intonation is next to impossible to achieve at any price.

    So, analog in, analog out.
  • But you would need 44100 Hz or better. 8 kHz is telephone quality. Depending on the type of pitchbend technique, I would say it's possible. The easiest way would be a ring buffer with independant read and write pointers using a phase accumulator design. You would have to handle cases where the read or write pointer "drives past" the other in some way, to get rid of the clicks. There are different ways to do this without too much CPU load.
  • Maybe this theoretical paper is for interest.
    I download it years ago.
    Must research the source.
  • I used to make pitch benders. You just need an ADC, a DAC, and some memory and a bit of processing power. Flanging works like pitch bending, with the buffer head moving at a constant rate, while the buffer tail follows at variable rate, with some mixing of the tail data into the head data.
  • Ahle2 wrote: »
    But you would need 44100 Hz or better. 8 kHz is telephone quality. Depending on the type of pitchbend technique, I would say it's possible. The easiest way would be a ring buffer with independant read and write pointers using a phase accumulator design. You would have to handle cases where the read or write pointer "drives past" the other in some way, to get rid of the clicks. There are different ways to do this without too much CPU load.

    The 8KHz was a starting point because this is the sampling rate of the Antares.

    But if I understand correctly, 44.1KHz is required to cover the full range of human hearing but a guitar? Aren't we talking more like 1.2KHz and therefore a Nyquist of ~2.5KHz?
  • potatoheadpotatohead Posts: 9,989
    edited 2020-01-08 - 01:27:27
    There are too many harmonics. The important ones peak at 8 to 10khz.

    For acoustic, 12 to 16khz minimums, for that "open sound", scrapes, touches, knocks on the body and strings.

    Just the range of fundamental sounds can hit 2khz, fwiw.

    Frequency, not sample rates.
  • Ahle2Ahle2 Posts: 999
    edited 2020-01-08 - 12:16:32
    Well, you will get quite severe audible aliasing distortion at a sample rate of 8 kHz. An analog bandwith of 8 kHz is something totally different. You would have to have an almost perfect [s]bandstop[/s] brickwall filter at half the sample rate (4 kHz) on the analog input and output, to be able to get rid of the aliasing. Then you would end up with an analog bandwidth of 4 kHz. But that would sound very muffled. A guitar is full of overtones, even beyond human hearing.
  • Maybe the idea is to capture the fundamental frequency, and feed it into Karplus-Strong?
  • Maybe the idea is to capture the fundamental frequency, and feed it into Karplus-Strong?

    Isn't that more for synthetic reproduction of string plucking (I had to go read-up) :lol:

    This is the Peavey guitar with the integrated Antares Auto-tune.
  • Wasn't there a "closed loop" tuning system some time ago? I mean with servos attached to the string tension screws?
  • ManAtWork wrote: »
    Wasn't there a "closed loop" tuning system some time ago? I mean with servos attached to the string tension screws?

    Yeah, Tronical. I might stick one on my Telecaster. Not as good as Auto-tune though because you can't switch on-the-fly during a song and it does nothing for bad intonation or correct for finger pressure making the notes sharp. Nor does it do the other cool stuff like virtual-capo, string doubling, pickup emulation, etc.
Sign In or Register to comment.