interactive controller design '03 :: Grab Pipe ::
** The Grab Pipe now has its own page. This is my class journal (designing/building
the GP)
assignment #1
I decide on a design.

assn#2 - play with a sensor. I've got some noisy footswitches, and magnetic reed switches.

//week4 Sketching out project. Talk to Taku about sliders - he recommends Vestax. Talk to Eric about rotating inner joint on controller.

//week5 Working on controllers from last year - good practice. Coding for the pic (midi) and sharp IR rangers. Fixing my footswitch box. Getting to know Max/MSP very well!


Getting my sketches and ideas together. Research on Vestax sliders - probably $50 each! Also need to figure out mechanical connectors so that tube can rotate well.

Tuesday Oct. 21st, 2003
A lot more sketching. The rotating joint scares me, it will have to be smooth, and yet not break either the potentiometer or the PVC tubing when it rotates. I come up with a few pages of sketches, but none that I particularly like. And I'm still having trouble finding good sliders - I try to buy a mixer off Craigslist, but they never get back to me. I look at other sliders online, and find some cheapo Gemini ones for $19 at some shady DJ place, but that's too expensive. I'll need to take a trip to Chinatown and see what's up at the DJ stores there, but I don't expect much.

Also, John Bergen has offered to give me a 3D accererometer. I'm going to stick it in a drum I have, and use the tilt, force of hit, and angle of hit to do some crazy drum synthesis. I love hand drums.

Tuesday Oct. 28th, 2003
More sketching. Thinking about how to work with that rotating shaft that I rescused from the old sander in the physcomp lab.

Went to Home Depot by me in Queens to figure out more ideas for parts. Played with various PVC tubing and accessories, but found nothing compelling. Was hoping that some off-the-shelf parts would work, and some came close (rotating sewer joints), but none looked better than the rotating shaft I already have.

Think I figure out a final sketch. I have good plans now. Also need to scan my sketchbooks and put them up online!
A final plan. Nothing left to do but finish coding and clean up some soldering.

Also, here's a description of the Grab Pipe:

The Grab Pipe is an audio/visual controller that lets a performer manipulate sounds and images on a computer or synthesizer. The Grab Pipe combines the general form of a saxophone with new controls techniques developed by DJs and Turntablists. Specifically, The controls are a cross between a saxophone and a DJ Battle setup; the top hand has saxophone-like "keys" and rotates (like a record), while the bottom half uses a crossfader straight from a Numark DJ mixer.

//week11 Painted Purple, assembled. A full-size image. Purple PVC paint is a solvent, smells really bad, and takes a few coats to e re sult is great, though, a really brilliant purple. Nothing but good comments on it. Physically, it's done, and the boards are soldered together. The code isn't finished yet. Trouble using ADC on the 16F877. I'll figure it out.

//week12 More soldering. The digital oscilloscope is my best friend. Yes, my friend. No, it likes me better. It told me so.

main board top board finished

The main board has a PIC16F877 chip, and connects to the top section of the Grab Pipe though a Cat5 cable (accelerometers, switches), a 5-pin header (the crossfader, which uses 2 power leads and a ground lead, and sends 2 separate analog voltages, one for each tied potentiometer), and the rotating potentiometer in the neck.
The design is modular, so that the top can easily come off, and the whole thing stuffs into my green duffel bag. This is a real-world, rugged, gig'ing instrument.

Also, Midi pitch bends are great fun. And using separate Midi channels for all of your sensor data greatly improves performance on the computer-side. Sending all of my controllers on the same channel was, stupidly, my problem last time. It will cause your computer to lock up and or/crash.

//week13 I'm finished. I'm using Midi pitch bends, and the range of the 14-bit values sent through pitch bends makes up for the small range of different values I'm getting through some of my ADCs. Yes, the crossfader's resistance is too high, but it works and gives me a good range of data.

Look here for my Grab Pipe code.

There's also my Max patch - using an ioscbank~ object and a bunch of cycle~ objects I'm generating some interesting sounds. More on this as I get it down. I'm also controlling live video (MP4, 6.1MB) via Jitter.

//week14 Done! Live performance for Live Video Processing. Also took the 'pipe on the road and played in a jazz jam session at The Frying Pan with Frylab. The Grab Pipe works great, and people respond very quickly to the controls, yet there's definitely more fine-controls to master through practice, which is the mark of an good instrument.