Roland GK3 and Fishman TriplePlay: low latency pickup access

[Gruust]

Hello!

 

I wonder if there is any way to tinker with the per-string signals picked up by those devices directly, ie. if there is any way to get low latency access (ASIO) to waveform data coming from the individual pickup coils.I basically want to develop my own guitar synths like Boss SY-1000. Because I think I know the answer already (no!), are there other solutions (ie DIY)? The main problem seems to be a somewhat reasonably priced 6-channel ADC with USB connectivity and ASIO support...

[Winston Psmith]

Can't address the Fishman, as I've never used one?

 

With the GK-3, there's no way that I know of to mechanically adjust nor attenuate the signal strength from each individual string. Once you've connected the GK-3 to a 13-pin device, you can assign each string to a separate MIDI Channel, and try to contour your signal as incoming MIDI data. If you're using a Roland GR Guitar Synth as a MIDI converter, your can set the individual String Sensitivity within the GR unit; how effective that would be in addressing your question, I'm really not sure?

 

FWIW, while the Fishman is designed to interface directly with your computer, the GK-3 is not, and would require yet another piece of hardware to connect to your computer. The discontinued GI-20 half-rack sized Guitar Interface doubled as a Hex Pickup to USB-MIDI interface.

 

I have to ask, when you say you want to develop your own Guitar Synths like the SY-1000, do you mean that you're designing hardware devices, or designing your own sounds?

[Gruust]

I have to ask, when you say you want to develop your own Guitar Synths like the SY-1000, do you mean that you're designing hardware devices, or designing your own sounds?

 

Neither. I'd like to write a software synth for PC that does NOT use MIDI but the raw waveform data. For that purpose I'd need the raw ADC input from each of the 6 strings/coils. In principle, I could re-wire the GK3 or TriplePlay and connect them to something like a Behringer UMC1820. However, I'd prefer a more plug-and-play solution. I assume that the Fishman TriplePlay Connect Hexaphonic Pickup does only provide a combined signal via its ASIO interface and not one channel per string.

 

[KuruPrionz]

I have the Fishman.

 

The problem is strings, not the convertors.

 

A string takes time to create enough cycles for a pitch to be detected, a lower frequency string takes longer because the note is lower and the cycles are farther apart.

 

Plus strings go slightly sharp on the initial attack, and "fall' to the note. So it is a moving target.

[Winston Psmith]

I have to ask, when you say you want to develop your own Guitar Synths like the SY-1000, do you mean that you're designing hardware devices, or designing your own sounds?

 

Neither. I'd like to write a software synth for PC that does NOT use MIDI but the raw waveform data. For that purpose I'd need the raw ADC input from each of the 6 strings/coils. In principle, I could re-wire the GK3 or TriplePlay and connect them to something like a Behringer UMC1820. However, I'd prefer a more plug-and-play solution. I assume that the Fishman TriplePlay Connect Hexaphonic Pickup does only provide a combined signal via its ASIO interface and not one channel per string.

 

Ah, thanks for the clarification. I'm not sure either the Fishman or the GK-3 are ideal for what you have in mind, in that case? If function is more of a priority than $$$, I'd look into some of the Guitar-shaped Controllers from Starr Labs. Z-Tar Controllers

 

You might also bring this question over to Craig Anderton's thread on MIDI Guitar.

 

[KuruPrionz]

I have to ask, when you say you want to develop your own Guitar Synths like the SY-1000, do you mean that you're designing hardware devices, or designing your own sounds?

 

Neither. I'd like to write a software synth for PC that does NOT use MIDI but the raw waveform data. For that purpose I'd need the raw ADC input from each of the 6 strings/coils. In principle, I could re-wire the GK3 or TriplePlay and connect them to something like a Behringer UMC1820. However, I'd prefer a more plug-and-play solution. I assume that the Fishman TriplePlay Connect Hexaphonic Pickup does only provide a combined signal via its ASIO interface and not one channel per string.

 

The Triple Play software allows you to asign a string to a sound so you can have 2 bass strings and 4 Hammond clone strings for example.

 

I think the Roland GK synth pickup might be easier to work with. It allows for adjusting to the radius of the bride and the output cable has 13 pins so there are probably individual outputs for each string available for you.

https://www.roland.com/V-Guitar/about.html

https://www.sweetwater.com/store/detail/GKC5--roland-gkc-5-15-foot

 

Craig Anderton also has a thread running in the MPN GearLab forum regarding Jam Origin, which is non-MIDI guitar synthesis software.

https://forums.musicplayer.com/ubbthreads.php/topics/3022228/jam-origin-polyphonic-guitar-to-midi-software#Post3022228

 

Cheers, Kuru

[Gruust]

Okay, so here is a general misunderstanding: I do NOT want the MIDI data. I want the raw signal data *before* the MIDI conversion (drop the MIDI part entirely), and for each single string in the form of a separate low latency ASIO channel input (6 total). What's the best way to get that?

[KuruPrionz]

Okay, so here is a general misunderstanding: I do NOT want the MIDI data. I want the raw signal data *before* the MIDI conversion (drop the MIDI part entirely), and for each single string in the form of a separate low latency ASIO channel input (6 total). What's the best way to get that?

 

I covered the difficulty as quoted in the last quote below. Truly, the best way to do that is to invent a new kind of string that defies the laws of physics. I will mention in passing that a friend has a nylon string Godin guitar with the Roland socket for playing Roland synths and it has less latency than steel strings. Nylon strings do not respond to bending notes as much as steel strings do. They also do not go as sharp when plucked. But the challenges of lower frequencies remain.

 

They are still the same frequency, pitch detection of strings cannot and will not ever be latency free. The best pickup would probably be a light beam pickup, this should deliver true seperation. It does NOT solve the problems that strings present when attempting to track pitches. Even analog octave pedals are "glitchy" on the lower notes. It is simply a matter of lower pitches having longer cycle times and nothing more.

 

Bottom line, the lower the pitch, the longer the latency. A friend recently bought the Boss SY-1 pedal for his bass and it was unusable due to latency and tracking. It works pretty well on a guitar, especially if you play up higher on the neck. Higher pitches can be parsed more quickly, it really is that simple. Cheers, Kuru

 

PS, Craig's MIDI Guitar thread ( https://forums.musicplayer.com/ubbthreads.php/topics/3014072/does-anyone-care-about-midi-guitar#Post3014072 ) has the following quote (see my quote below it):

 

I believe it was Dick Rossmini who strung one (not a Casio, something else) with the same string gauge across the guitar and tuned to the same pitch, so each string would take the same amount of time to figure out what note was being played. Each string had its own pickup output so he just set up the instrument to transpose each string to the correct pitch. Joni Mitchell used a Roland like that for a while, using the MIDI smarts to transpose the string pitch into that of the off-brand tuning she was using.

 

 

I have the Fishman.

 

The problem is strings, not the convertors.

 

A string takes time to create enough cycles for a pitch to be detected, a lower frequency string takes longer because the note is lower and the cycles are farther apart.

 

Plus strings go slightly sharp on the initial attack, and "fall' to the note. So it is a moving target.

[Winston Psmith]

Okay, so here is a general misunderstanding: I do NOT want the MIDI data. I want the raw signal data *before* the MIDI conversion (drop the MIDI part entirely), and for each single string in the form of a separate low latency ASIO channel input (6 total). What's the best way to get that?

 

It doesn't appear that there is a way to do what you want, not with the design of the GK-3, as I understand it.