Pitch to Cv converter
- Thread starter tjs2010
- Start date
I don't think that's a very good idea. That unit is not specialized for guitar, does not have the correct filtering for guitar nor any provision for "locking" to the played note/string, preventing the 2nd harmonic that will appear after a while.
Also, being a "general purpose" converter i would have some doubts on it's response speed when playing guitar.
The Roland SPV 355 was a similar unit (a general purpose freq-to-v converter), and it didn't work correctly for guitar. The GR300 (that came just 2 years later) guitar synth was way better.
This web page is required reading
http://www.joness.com/gr300/index.htm
For example look at the service manual for a Roland GR700 (a so-so tracking guitar synth) and jump to page 8:
http://manuals.fdiskc.com/flat/Roland GR-700 Service Manual.pdf
The problem is that they use zero-crossing comparation without having seriously filtered the input waveform to isolate the fundamental. This means the zero-crossings are not reliably representing the period of the waveform. They mention they use some algorithm on the CPU to get a reliable pitch measure, but that means WAITING a little bit until the captured pitch is understood by the CPU as "reliable".
A better idea is to either use a very good fundamental isolating filter, and/or to use a different method than zero crossing, for example, detecting the positive and negative "peaks" of the waveform, but only the highest peaks, ignoring the smallest peaks. This can be done in the analog domain or even easily on the digital domain.
Unsurprisingly the GR-700 is not a very fast tracking guitar synth.
http://www.joness.com/gr300/index.htm
For example look at the service manual for a Roland GR700 (a so-so tracking guitar synth) and jump to page 8:
http://manuals.fdiskc.com/flat/Roland GR-700 Service Manual.pdf
The problem is that they use zero-crossing comparation without having seriously filtered the input waveform to isolate the fundamental. This means the zero-crossings are not reliably representing the period of the waveform. They mention they use some algorithm on the CPU to get a reliable pitch measure, but that means WAITING a little bit until the captured pitch is understood by the CPU as "reliable".
A better idea is to either use a very good fundamental isolating filter, and/or to use a different method than zero crossing, for example, detecting the positive and negative "peaks" of the waveform, but only the highest peaks, ignoring the smallest peaks. This can be done in the analog domain or even easily on the digital domain.
Unsurprisingly the GR-700 is not a very fast tracking guitar synth.
This is a perfect and simple explanation of how the GR300 works, required reading!!
How the GR-300 Works
How the GR-300 Works
Apparently it can be done:Guitar into RS-35
And with regard to the latching function, as it is a module for a modular synth, it's a matter of adding a module. Filtering and compressing is advisable and can be done in front with other modules (although the module has some bandwidth limiting built in). Apparently, adding a slewing module is also recommended.
Haven't tried the RS-35 module myself, but it does seem a very good candidate (for bassguitar as that's my primary instrument).
And with regard to the latching function, as it is a module for a modular synth, it's a matter of adding a module. Filtering and compressing is advisable and can be done in front with other modules (although the module has some bandwidth limiting built in). Apparently, adding a slewing module is also recommended.
Haven't tried the RS-35 module myself, but it does seem a very good candidate (for bassguitar as that's my primary instrument).
Halo Jarno,
It CAN be done, but what i mean is that it is not a very good approach for the reasons above mentioned. Not any "filter" will do, you need a filter that locks to the fundamental.
For bass guitar it's even worse since the period of the waveform is far longer. This means that for bass guitar the response times will be even longer. For bass, an identical approach to the one used in the GR-300 is best. Roland did make a bass version of the GR-300.
The good thing with bass guitar is that the fundamental is always very strong and permanent through the duration of the note, thus the filtering is a bit easier. Unless you do some slapping or let the string ring against the fret. In that case a lot of harmonics will be produced.
Last edited:
Hello Flavio,
I agree it's not easy, and the hexaphonic approach is a lot better. Had a look at the documents you linked, that Roland pickup looks diy-able, do you think it has changed much (wrt. the current GK-xb, not sure which number they're at)?
Hello Andy,
Good info as well, didn't know that forum, but lot's of insight there!
I agree it's not easy, and the hexaphonic approach is a lot better. Had a look at the documents you linked, that Roland pickup looks diy-able, do you think it has changed much (wrt. the current GK-xb, not sure which number they're at)?
Hello Andy,
Good info as well, didn't know that forum, but lot's of insight there!
Yeah, I am familiar with the electro-music forum, lot's of diy going on overthere.
Muff wiggler is a nice forum on modular synthesis as well.
Muff wiggler is a nice forum on modular synthesis as well.
Not sure how actual this thread is, but I've bought a secondhand Analogue Systems RS-30 pitch to voltage converter, and it uses a LM2907/LM2917 (don't recall which one) pitch to voltage IC.
The power supply in my modular is down, so I haven't been able to try it out extensively, but definitely does not feel like a plug-n-play solution.
The power supply in my modular is down, so I haven't been able to try it out extensively, but definitely does not feel like a plug-n-play solution.
It has changed a lot. No other Roland system uses the same guitar-to-synth method as the GR-300. The GR-700 was totally different.
Basically the GR-300 uses a clever method to convert guitar frequencies into pseudo-sawtooth waves of any frequency. It works perfectly but it's limited since you can only get those saw waves.
The hexaphonic approach is a lot better if you can afford doing everything SIX times. I'd rather go for a monophonic converter that is six times more sophisticated...
Read my previous post; it's a tachometer circuit. Seems the guys who designed the RS30 did not do their homework fully. But i don't blame them at all; it took me more than 1 year to have a small glimpse of the history of converting guitar pitch to control voltage. There's not enough information on the web.
Well, We'll have to wait anxiously to see what you come up with. In the meantime, be glad that most parts have various uses, and that an IC meant for an industrial application is useable for DIY audio/music applications as well, or it will be obsolete in the blink of an eye and appallingly expensive (see CEM IC's, SSM IC's etc. etc.).
Hi Pyronet,
That circuit uses a tachometer-type circuit for F-V conversion and thus is not ideal. I would not recommend building it.
New pitch-to-cv solution
Hello,
I know this is an old thread but there's a new solution for pitch-to-cv coming out soon:
Convertor – Sonicsmith
https://www.youtube.com/watch?v=juCM68eOKnw
It doesn't rely on a PLL or any of the old F-to-V chips and offers a number of other advantages as well.
Hello,
I know this is an old thread but there's a new solution for pitch-to-cv coming out soon:
Convertor – Sonicsmith
https://www.youtube.com/watch?v=juCM68eOKnw
It doesn't rely on a PLL or any of the old F-to-V chips and offers a number of other advantages as well.