Optigan...

[untune]

As a sortof a follow up from the Mellotron thread, and being inspired by Gordon's mention of the optigan, I wondered how feasible it would be to make one.

Here are a few videos that I'm very intrigued by:

Motor Keys Optical

opto organ

Optical organ: spindle

The second one interests me quite a bit.  I've just found another page here http://www.umatic.nl/tonewheels_historical.html that details the process quite thoroughly, which I'm now going to go and read (there's also a 'technical' link on the right which has schematics etc.

Any thoughts?

[Gordonjcp]

The middle one is kind of what I was talking about.  Touching the disc when it's spinning looks like a good way to give yourself some astoundingly nasty cuts on your finger.

The last one is interesting - got a link to the original clip?  The sound is probably muffled because it doesn't look like he's using a tight enough aperture for the LED - it really wants to shine through a thin slit, the way it does in a film sound unit.

[untune]

This?

Reading an Optigan disk

[untune]

I'd like to see 3 discs in unison acting as 'oscillators' in a synth

How is it that most of these homemade optigans use pretty basic block patterns to generate sound, yet the optigan's disc tracks more closely resembed real waveforms and are real samples?

[Gordonjcp]

Probably because it's easier to draw a squarewave on a disc than anything else!  I'm not sure how it was done on the original Optigan discs - some cunning optical device, no doubt,

I'd be inclined to write a bit of software that produced an image of the track and mapped it into a circle, with some suitable crop marks for cutting it out once you print it on acetate - another poster kindly provided a link to some acetate suitable for laser printers.  Don't use ordinary non-laser printer safe acetate in a laser, it will melt and stick to the fuser rollers!  You could try inkjet printing but I think it would really need to be a deep toner rather than translucent dye - you can print colour on acetate with an inkjet because it lets light through.  You don't want that, you want to block the light totally.

[untune]

I found a step by step procedure of how they were made somewhere yesterday.  It was a long, complex, costly procedure that involved cutting master recordings from tape and using a photographic process to expose the waveforms on the acetate material.  Used a lot of machinery but could no doubt be done very easily now with a computer and a laser printer

I do have access to a laser printer so this could be fun to experiment with.

[moordenaar]

The "acid machine" by gieskes might be of interest to you aswell.

[untune]

Would a 2-octave keyboard just need a bigger disc to contain all the different waveforms (than the smaller ones in these videos)?

i.e. 25 seperate strips arranged in a circle?

[untune]

DIY Optigan / Orchestron Disc Player

A new development... sounds really cool, apart from the clicking... and the fact that it's so big!

[Circuitbenders]

on a vaguely similar theme:

http://www.youtube.com/watch?v=HibVUaXBYFU&feature=player_embedded#!

http://obsoletetechnology.wordpress.com/projects/turntable-sequencer/

[untune]

That's a very cool idea!  The option to 'draw' your drumbeats is a ace!

[untune]

Hey all...

I've been doodling this idea up for weeks... I don't think the microcassette idea is do-able without some deeper knowledge of the arduino which I don't really have the time to learn at the moment.  But after doing a LOT of reading on the subject of optical synthesis the idea has really caught my interest.

Anyways, I've settled on a basic idea for a micro 'optigan' which uses 5 small discs as opposed to one big 12" record sized one.

Each track on the disc will be 3mm like a regular film soundtrack, so it's going to basically need to have five banks for 3mm LEDS, each one with a 1mm gap between.  Now I'm hoping to do some prototyping, I could probably get some kind of tubing that would act as a collinator to 'focus' the beam into a 3mm spot - heat shrink might be hard to control but I'm sure I could come up with something.  Either way, I'm having trouble finding the right LED's.

They need to be 3mm (and no bigger than 4mm) and I think they should be red.  Thing is I'm unsure right now whether this should be using phototransistors, photodiodes or LDRs as I've seen it done with all of them.  Then the light output from the LEDs ideally needs to be as close to the peak wavelength of the photosensor.  I've been scouring rapid and farnell - ideally needs to have as focused a beam as possible to begin with.  So I've been looking at transparent/water clear red LEDs with a narrow viewing angle.  Seems almost impossible to get anything in the right size under 50 degrees!  Any ideas on what colour/type is best to use for a project like this, and any ideas on focusing the beam?

[Gordonjcp]

Photodiodes or phototransistors will be much faster than LDRs, and have a smaller window.

Try mounting the LEDs in an opaque block with holes drilled in it and the optical disk flying just a little above it (not touching, so it doesn't get scratched).

[untune]

Photodiodes or phototransistors will be much faster than LDRs, and have a smaller window.

Try mounting the LEDs in an opaque block with holes drilled in it and the optical disk flying just a little above it (not touching, so it doesn't get scratched).

Thanks gordon, that's made it easier!  I'm trying to think about the best way of making the circuit... it's going to need a way of driving 5 motors with PWM - at this point I'm not sure if those motors will need to be spinning at different speeds or the same - I'm still attempting to work out the relationship between speed/frequency/size etc.
I was hoping there might be some way of having an ADSR on this - but I'm guessing that it'd need 25 seperate ADSRs if all a key does is completes the circuit and lights an LED.  I've also looked at the idea of using the LED/photosensor as an opto coupler, which would effectively be an ADSR (or an AD at best).  Apparently though, when Mike Walters prototyped that with the melloman, the decay time was around a minute, so he ended up using transistors.

Yeah I was thinking of mounting them just like that, my only concern is the fact that those 3mm holes through the  block would need to have a 1mm gap between them - very precise drilling required and I don't have any way of doing it!   plus, typically, the base of a 3mm led means that it isn't going to fit the whole thing inside the little 'tube' that it creates.  I'm not sure if the LEDs are going to sit on the underside or overside of the disc yet though.

I've also designed the discs so that they're the excact size of a minidisc, and will fit inside the shell of one.  That would protect them and provide ideal storage too, besides looking cool   Nice little recycling aspect there aswell