Guitar-like midi synth controller

[squeal]

OK so we talked about this project a bit in the "simple DSP inside Core Module?" thread, but I wanted to give an overview here.

The plan is to build a guitar-like midi controller for use with midi synthesizers. The strings of the guitar will be essentially ribbon controllers with vhs tape glued to a fretless guitar neck. From these circuits I will extract a pitch value and either quantize it and apply it at note-on or turn it into a cc to be applied to the synth's pitch bend parameter. Other, more exotic uses could probably be tested with ease once I have built it.

The strings will run over graphtech ghost bridge saddles which also act as individual pickups. (If anyone knows of a less expensive polyphonic pickup system, let me know. They are $100) Each string's amplitude will be converted by some means into midi-cc values, with the intent to apply these to a synth's volume parameter.

I have some experience in analog electronics and some c programming, but am new to pics and designing midi controllers, so I appreciate any input you guys have.

[bugfight]

(If anyone knows of a less expensive polyphonic pickup system, let me know. They are $100)

you could probably use regular piezos, which are much cheaper.

a bit more work, though...

[squeal]

I will look into that.

I guess it depends how much time it will take and how successful I feel I could be.

[Jaicen]

You might want to consider looking at the Hagstrom guitar synth approach. That utilised the strings and frets as a keyboard matrix. Each string was isolated, and the frets were wired into a matrix. That way, when the string was pushed onto the fret, it produced a discrete voltage. Obviously this doesn't allow for pitch bend, but it is fast (keyboard fast) and very reliable. It would also allow you to use standard pickups and regular magnetic strings. In fact, it would play just like a real guitar, except that it's not necessary to pluck the strings, just to fret them, to get the sound. I don't know if you've seen it, but the Stepp guitar synth used a similar detection method, with gate signals derived from a seperate set of strings, so maybe that could be a possibility if you use piezos as triggers. 

[cimo]

hey

(hagstrom)

i didn t know that hagstrom guitar.. have you seen his shoes?

and this guy s glasses? (stepp)  8)

we really don t know what style is anymore..  :o

simone

[squeal]

I don't think you could use normal frets and achieve polyphony.

[Jaicen]

.....Because??

I love posts like that. Quite beside the fact that it's already been achieved over 30 years ago......

[squeal]

.....Because??

I love posts like that. Quite beside the fact that it's already been achieved over 30 years ago......

I don't think it has. The Hagstrom was monophonic, The stepp thing had some kind of plastic frets or something.

I don't think there has ever been a polyphonic guitar synth with standard metal frets and metal strings. I could be wrong, though.

[seppoman]

the (diode) matrix approach would work polyphonically. the frets can be in one piece, just you´d need to find a way to have the strings insulated from each other.

Seppoman

[Wilba]

I've been dreaming about getting a Ztar for ages:

http://www.starrlabs.com/

which has been the inspiration for my own guitar-like ribbon controller instrument, still unrealised four years later... man, what have I been wasting my time on?  ;)

[cheater]

I say go for a fretless.

The board will be a resistive element and depending on the place the string is touching a different resistance is created. That's for pitch tracking.

Fretless bass for the win :D

Edit:

Either way, each string should have a separate pickup that goes into a separate synth voice.

What is the best way to create separate pickupping? Maybe take apart a normal pickup and insert separate coils for each string?

There's a DIY pickup page somewhere on the 'net. Pretty good at that.

[squeal]

Cheater:

In the opening post I mentioned the graph tech ghost bridge saddle pickups. They are bridge saddles with piezo elements mounted inside. These were designed with hex output in mind, so I imagine that they would get good individual output with minimal crosstalk.

[squeal]

I haven't really seen the diode matrix design, but it seems that barre chords would frustrate any design using solid metal frets.  ??? 

[Screaming_Rabbit]

The board will be a resistive element and depending on the place the string is touching a different resistance is created. That's for pitch tracking.

... I don't know how you play the guitar. - If I would play it, the resistive plating would be worn down after a few vibratos.  ;)

I haven't really seen the diode matrix design, but it seems that barre chords would frustrate any design using solid metal frets.  ??? 

... I don't see a reason.

Greets, Roger

[squeal]

I bet you're right about the wearing down of the resistive element. Maybe there are more robust resistive materials than vcr tape? Antistatic packaging is supposed to have resistive qualities. But that's also just plastic.

I'm sure the companies that make slider pots would know about robust resistive strips.  ;)

Here's a question: What value pots are suggested for use with the ain module? What values are allowable?

[Screaming_Rabbit]

I'm sure the companies that make slider pots would know about robust resistive strips.

... the company where I work was building their faders by themselfe a few years ago. - The biggest issue was the design of the sliding piece in matter of wearing down the taper vs. cleaning it (scrape away the dirt without wearing down the conductive plastic). - So ... no chance  :-[

Greets, Roger

[cheater]

How about, um, ebonite? no idea really

[Durisian]

What about a single resistive track on the side with each fret attached. Connected to Vs and Vd

Each string connected to an AIN input.

As you press down on a string the voltage travels off the resistive track, through the fret, down the string and to the AIN.

Would this work?

Of course you'll have a live wire problem. :)

[Screaming_Rabbit]

What about a single resistive track on the side attached to each fret. Connected to Vs and Vd

Each string connected to an AIN input.

As you press down on a string the voltage travels off the resistive track, through the fret, down the string and to the AIN.

Would this work?

Of course you'll have a live wire problem. :)

... don't look/think too far. As Seppoman already mentioned: http://www.ucapps.de/mbhp/mbhp_scan_matrix1.pdf

[bugfight]

... don't look/think too far. As Seppoman already mentioned: http://www.ucapps.de/mbhp/mbhp_scan_matrix1.pdf

you need a diode on each switch for this to work.

you would have to separate the frets into 6 parts to achieve this...

[squeal]

... the company where I work was building their faders by themselfe a few years ago. - The biggest issue was the design of the sliding piece in matter of wearing down the taper vs. cleaning it (scrape away the dirt without wearing down the conductive plastic). - So ... no chance  :-[

Greets, Roger

Ya but in this case I wouldn't be really sliding anything against the resistive element, just pressing down the stationary string at different points, assuming I can resist the urge to "bend" notes side to side as on a normal fretted guitar. ;)

[squeal]

the (diode) matrix approach would work polyphonically. the frets can be in one piece, just you´d need to find a way to have the strings insulated from each other.

Seppoman

You can't insulate the strings from each other when you play a barre chord.

[stryd_one]

Who plays barre chords anyway? ;D

[seppoman]

you´re right, the frets would need to be segmented, too :( I missed the part that even while you can wire it up as one (after the diodes), you still need a diode at every single contact point. So this would be a lot of mechanical work. But with separated frets and bridges, it would at least work.

When playing barré stuff with the resistive approach, you need to put into the consideration that the Rs of the different "strings" will also influence each other.

I´d say the easiest solution would be to buy a commercial guitar-to-midi interface. everything else will have some drawbacks compared to it, and having to make swiss cheese out of the neck will cost a lot of time, work, and eventually also money. If you still want to diy, the diode matrix/env follower approach still looks to me as the best solution.

Seppoman

[Jaicen]

I agree, if you want something that just 'works' then get a GR-20 and a GK3. I use mine regularly, it's about as good as it gets if you ask me.

If on the other hand, you like a challenge....