Optical Pickup Midi Guitar. How does it work?

[Fear the Weasel]

Hi there, I stumbled upon this the other day.

http://www.opticalguitars.com/

I was wondering if anyone had any ideas how it worked. 

I'm thinking through the idea of building something similar and I'm a bit stuck on how to do the pickups.  Basically I just need to figure out how to get a velocity from the strings (not pitch).

Thoughts, ideas?

[Screaming_Rabbit]

... could be an opto interrupter. Any 0-passing of the string would end up in an interruption and so you could read out the string's frequency. But I think this would need some kind of a lense in front of the interrupters.

Greets, Roger

[smashtv]

I'm pretty sure Roger is right, I have seen similar systems back in the coinop days.  Special lenses might not be needed if discrete parts (separate emitter/detector) are used. 

Decent dynamic range would only be possible by blocking outside light completely, or tuning the code so the emitter is running at x frequency instead of always on, and reject everything not at that freq read by the emitter.

This would be easiest to handle with a separate MCU that feeds string data to a Core via IIC, due to the critical timing on the optos.

Best

Smash

[Fear the Weasel]

After a bit more looking around on the net I found another similiar design.  From what I gathered it had a UV LED on one side of the string and a detector on the other and measured the 'shadow' of the string.  Your right about the ambient light thing though, seems to be one of the biggest problems with the pickups design. 

Optical aside then, does anyone have any idea how I could convert hitting a string into usable data.  Just the fact the string has been struck to start with but also to get a corrisponding 'velocity' at some point hopefully. I was thinking maybe something along the lines of using the output of a normal pickup? (would need a different pickup for each string though to distinguish)

Thinking out loud..  ;)

[Sinnsyk]

why not building a sort of box around the end of the strings, that their can't be any other light source shining in the sensor. It would be big on the guitar, but would save a lot of time.

[Screaming_Rabbit]

Optical aside then, does anyone have any idea how I could convert hitting a string into usable data.  Just the fact the string has been struck to start with but also to get a corrisponding 'velocity' at some point hopefully. I was thinking maybe something along the lines of using the output of a normal pickup? (would need a different pickup for each string though to distinguish)

... the interrupter doesn't just picture the zero-line of the wave but more an area... it would not be precise enough. - So you actually have a time factor for the lenght of the interruption. The lenght of the interruption is a value for the amplitude, since the higher the amplitude is the shorter is the interruption. I don't know in which area this time difference will be and about it's even measurable.

Greets, Roger

PS: Why would you wanna add a magnetical pickup per string? In this case you could take the whole information out of the analog signal right away. The only advantage I see in the optical way is, that you don't have any magnets influencing the string sustain and that the signal of the optical solution takes less DSP-power to calculate.

Greets, Roger