This weekend I made a proof-of-concept demo of the SyncWalk digital composition interface I’m working on. You can see photos on my flickr set, and a video is up on YouTube (I think Harlo may have a video of it up soon).
As you can see, it’s a little noisy right now—I’m going to add some filter capacitors and mill a PCB this coming week in an effort to ameliorate this.
At this point, I’m still not sure how I will differentiate between a palm-contact or a thumb contact, let alone contact between the two hands. (The video does not show it, but the glove has a conductive pad on the palm as well) One strategy would be to use the PWM outs to give each finger’s pad a different tone, like a DTMF pad, or to use different reference voltages for the thumb and palm contacts. Interference would be avoided if I could make a DTMF-like solution (whereas it would be the main source of noise if I had to read analog values), but I’m not sure frequency-detection is something AVRs can do well.
It’s always nice to have incremental rewards when you do a project, and this weekend has yielded encouraging results. I’m ordering conductive fabric tonight from here and will be trying to design a glove-to-cat5 PCB in the coming days. Let me know in the comments if you have any ideas or suggestions!
Friend, I have been absent due to a disgusting illness.
Nowdays I’m working on the following:
Cameroon Project—yesterday Harlo and I picked up some inexpensive crank-based power supplies from Ocean State Job Lot. Next step is dissecting them and seeing what can be modded to work with the toy. Last week while wheezing and expelling, I mocked up the voice chip’s demo circuit on a breadboard but had a strange problem: I can’t seem to make the mic work. When I press play I can hear that I’ve recorded the noise naturally made by the pins floating, but somewhere in the recording process the signal is either being lost or never being generated at all. Anyone had any experience with this chip? Google yields some good applications and examples but no users have had this issue (makes me think I’m really fucking something simple up here).
South Africa—Dear one, have I even told you about this? Or am I keeping you in the dark? My apologies!
This project involves taking some samples of native South African instruments recorded by my friend Thenji (part of Curious AV) and muxing them into something usable for the score of her upcoming film. So far I’ve been in the cut-and-experiment phase of the project, but I’ve committed to have three or so working models of songs by December first. Here’s the first go at a tune, made with Ableton: Kalimba(draft)
I’m taking this as an opportunity to further flesh out my max/msp interface, the woods, and add some useful features it’s been missing. It wouldn’t be fair to harp on about my own interface without crediting Keith Fullerton Whitman, whose presentation of his “H Mod” interface in 2003 at a Harvard Aritst Workshop was truly inspiring. You really should go give some time to the rest of his site (and, it goes without saying, his music). He is a good man, and he has a good brain. The wonderful part of this interface is that the main images on his screen (the four groovemodule waveforms in rad colors) are literally parts of his RAM - the mapping is that direct, and it allows you to easily see what’s interacting with what, what’s going to make your machine hang, etc. I’m not the fan of external VST modules that Keith is, but I can’t dispute the man’s sound, and that sound is gorgeous.
Grad School App—Well, shit. I’ve secured vacation time to get my online portfolio super-ready, I’ve tried to make friends with some of the awesome people in the Hyperinstrument group (although no luck getting a meeting with Tod, the Leader), but damn it if the professors I had hoped would write my recommendations are un-reachable. Jef Huang, my awesome Physical Computing prof, has always been a bit difficult to get in touch with, but I was really hoping I’d be able to secure something, anything from him. It’s truly unfortunate but in some ways telling that the professors I admired the most were destined never to become full-time Harvard staff; Jef’s in Switzerland at an institution that seems to have a far better handle on what new art is, Peggy Phelan is in sunny Stanford, and Elisabeth Subrin is working all along the East Coast. Dammit!
Well, the flower project hit a snag when the dragon skin I molded turned out to be too thick for the muscle wire to move—in the course of figuring this out, I think I may have damaged/overheated some of the wire segments.
Most of the wire is undamaged, though, and it was easy to retrieve the other wires from the silicone.
Now I’m using a much thinner and lighter strategy involving a plastic flower I got at a craft shop. The downside here is that now I’ve got a shot at making a flower that magically sets itself on fire.
My new form factor also means there’s no inbuilt means to force the wire back into a bent position when it’s unactuated - I’ll have to work out something there once the silicone has cured. It won’t be pretty.
The other bad part: so far, I have not got a circuit working - I initially ordered too few/the wrong parts. Now I’m using a relay from an old robotics project of my brother’s, but I’m concerned about the current drawn by each wire—to get them to react quickly you have to really juice them up, and I doubt my relay can handle the current drawn by 6 of these, not to mention where I’ll find a power source for all this.
The circuit I’ll worry about tomorrow. Tonight I’m not sleeping until I’m confident at least the form factor is correct and workable.
Well, it’s really coming down to the wire here with the flower project. For the next few days I may not be updating as frequently, but I’ll post a final write-up with lots of photos etc when the dust settles.
I have spent too much time working out the mechanics and not enough time planning for the circuitry side of this; I don’t even have a final behavior model yet, so tonight I’ll be hastily curing v2 of my prototype petals and trying to decide what the best operational flow will be, and whether or not that’s feasible. Last-choice scenario is using an Arduino or my flaky CUI board to detect events and actuate the flower.
Meanwhile, my good friend Thenji (a CuriousAV affiliate) has called me in on a film she’s making about South Africa - I’ll be making a small amount of prepared sound from samples she’s recorded. I listened to the first batch of folk instrument samples and they’re quite exciting—mbiras, little flutes, and hand drums are some of my favorite toys!
This weekend I had some encouraging results with Dragon Skin, a silicone molding compound from the pricey folks at Smooth On. Luckily, the silicone we had left over from PS04 was still usable, and it was much easier than I remembered. Some photos of the prototypes are available here.
What I’ll now need to figure out is exactly how to shape the nitinol I’ll be embedding, and how to ensure an electrical connection will always be made regardless of the bend-state of the wire. A journeyman friend is in town tonight but I’ll try to have a more detailed summary by tomorrow.
In the meantime, please have a look at the photos and tell me: how am I going to get the petals to shrivel/contract/close on command and what will have enough force to return the wire to its normal state?
Buddy, last night I fulfilled a commitment I made to a bar in my area. I am not bragging, but I thought you should know.
I have some updates on other projects as well:
I’ll be making more progress on the Cameroon project once I’ve got the dorkbot commitments. Fortuitously, Chelly will be in the States this Christmas, allowing me to push the deliverable deadline back to late December. Shit is going to be bananas at work and getting a grad school app ready, so it is nice to have the extra time (although admittedly, this project has been very slow to gain momentum).
The flower project is going well but will require some real work this weekend making a final decision about the actuators. Either they’re going to be embedded in each petal, which may diminish mobility, or they’re going to be somewhere else and I’ll use monofilament to transfer motion. Using monofilament would be a bit of a defeat, though, since it really obviates the need for muscle wire - I could do it with servos.
Excitingly, there’s a new (and hopefully easy) project that might prove more immediately rewarding—there was a call at the last Dorkbot meeting for an interactive audio exhibit for the kids to enjoy during the opening reception. I’m going to make a small patch of flowers that react to presence and produce (hopefully spooky) sounds. More on that soon.
Friend, I’m making progress on this project. You can always check the flickr set here for the latest images.
Progress-wise, here’s what I’ve got left to do:
Decide on the actual material to use for the final flower
Figure out how the muscle wire will connect on either end, and where this will be - it seems now that molding my own silicone petals will not be feasible in my timeframe.
Design the circuit - I have a photoresistor that would be easy to hook up to a small transistor circuit, but I’m not sure this’ll yield consistent results. Also, there’s a latching problem that I might want to try and solve with a timer.
