My penultimate semester at ITP has come to a close, and it was my busiest yet. These pictures show my final products for the Zoraida project: a microphone/vocal processor made out of elephant statues, a machete that maps its movement to sound (via a wireless accelerometer), and two conga drums that control video (xfade, saturation, and awesome color feedback) when played.

These pictures are from the NIME show where the instruments were performed:

And these pics are from the ITP Winter Show where the instruments were subjected to rigorous user testing (and passed with flying colors for the most part).

Despite being jetlagged and massive sinus pressure, I got a good bit of work done today. I’ve received only one of the two Fishman Powerjack endpin preamps I’ve ordered, but I got one conga drum to control max/msp/jitter today. I had to solder the 6-inch piezo film to the preamp, attach a 9V battery, and then secure the film and preamp inside the drum. The result is surprisingly good. This configuration makes a very sensitive and high quality microphone:

I then ran the output of the mic/preamp into an m-audio quattro interface, and into max/msp/jitter. Right now the patch is very simple. I’m using the peak frequency of the incoming audio amplitude adjust crossfading between two images, and things like brightness, color, and saturation:

At least things are talking, the actual effects can be modified later.

Also, I got a new aluminum pipe with a secure base flange to replace the pvc pipe on the old mike stand. I’m going to give up the twisting effect of the main pipe in exchange for structural integrity. The elephant will still twist, but the rest of the stand will be stationary. Other effects will have to be controlled via some onboard pots or something.

The major breakhrough today was getting the machete movements mapped to sound. I did this through rewire, with Max as the master and Ableton Live as the slave. Basically, I took the x,y,z data in from the accelerometer, filtered each stream of data using some javascript functions I wrote, and then used that filtered data to control midi notes which were used to control a synth in ableton live. The sound is not really what I want, but for now at least it’s fluid and fairly well-mapped to movements. Whew.

The patch:

Machete in action:

Click To Play

Also, I ordered aluminum pipe and flanges so i can work on elephant mic stand round 2. I also ordered pre-amps for the drums, which are here now. Here’s some footage of the prototype stand in action. The song being sung is still in the works, and Nancy’s still rehearsing the vocals.

Click To Play

Unfortunately I’ve got to go out of town tomorrow till Sunday the 24th, but all necessary parts will be here when I return.

This week has been pretty good. First off, I got my WiTilt from sparkfun. Following the instructions on the sensor workshop, I had no trouble getting the wiTilt’s data into Terminal. Getting the data into max and parsing it, however, proved not so easy. I could get the data in, but couldn’t get it to send continuously as the messages kept timing out. So then I tried downloading Woon Seung Yeo’s W2O, an app that parses the WiTilt’s data and sends it via OpenSoundControl into Max, but in her example patch she was using a deprecated object for receiving OSC data. I tried using Max’s built-in OSC reciever object, udpreceive, but things wouldn’t talk.

I finally found a patch via the Max/MSP forums that someone had made for the receiving and parsing the WiTilt data. Analyzing it, I realized that I was sending the wrong command (in ascii) to the WiTilt telling it send data. Problem solved. So now I’ve got a patch that takes in the witilt data wonderfully. I attached some oscillators to the incoming x,y,z data from the WiTilt, and the results are below:

Click To Play

Plus, I got the machete I’m going to use with the WiTilt. So one object/instrument is looking very good.

I also received a large elephant I ordered off ebay that is fairly light and i think it’s perfectly drillable for inserting a pot. I also got 20 panasonic WM-61A mics for the elephant after reading about them in “Getting a Bigger Sound” by Bart Hopkin and hearing them used in a live setting on this site.

These mics are really, really small. And there is a hack for the mics that prevents distortion under high sound levels that requires some really intricate soldering. I got one to work, but probably ruined 3 in the process. I’m not sure what to do with these things now. Wiring many of these seems like too much a time consuming task. I’ll probably give it another go though. I really need to put at least two or so on the elephant’s trunk.

I also got 5 or so OPA134 op-amps that I was going to run the mics into max with. They are typically used in DIY headphone amp circuits, but I figured they’d be plenty powerful for my needs. First tests weren’t successful, but I need to sit down and put more time into it.

Next week: Get mics/amps to work and then use them in max. Get hand drum to use with contact mic in max. Be optimistic.

For Meditation 3 (patch), I modified an algorithmic drawing machine so that it draws arcs whose values are based on the golden ratio of phi (1.61803399). The arcs drawn by the algorithm have a nice symmetry to them, and create some pretty, undulating circular patterns.

Here’s some video footage of the patch’s output:

Click To Play

For Meditation number 2, I made a jitter patch that calculates the compound interest of my ITP (principal + interest), and uses that data to modify the brightness, color, and saturation of a photo of a handful of money. The higher the loan value gets, the more “aflame” the photo appears. You can download the patch here.

This week I bought some necessary parts for the project, namely a few Radio Shack phototransistors and infrared emitters, as well as a few condenser mics. I think built an preamp/opamp circuit using an 4049 IC and an LM386 chip to test out the mics (we might use these on our elephant objects) and see what kind of sound values they put out.

I then built a basic audio-in max patch to play with the sound:

The 4049 got the signal nice and crunchy, but the overall output left much to be desired. We are still deliberating as to whether or not we will use this circuit.

We also ordered a statue of three elephants on a wooden board, which might serve as the main object/instrument with each elephant trunk housing a sensor (mic, beam-break emitter, ultrasonic?).

We plan to set it up on a podium-like board, so that interacting with the object evokes giving a formal speech, and hopefully we will be able to rotate it 360 degrees so that we can choose with which sensor to interact. This is better said than done.

Also last week, we recorded footage of Nancy dancing and experimenting movement that will be worked into the piece’s choreography.

This week I need to order some robust 360/infinte/endless pots, figure out a way (prism?) to focus the beam-break emitter (whether IR or superbright LED), and get other elephant statues to modify. Then I need to test, test, test so we can start putting these things together.

Tomorrow, Nancy and I are playing a show at West Nile in Brooklyn under the alias Zoraida. It’s gonna be mostly improv, but it will our first experiment in public with some material related to the project we’ve been working on.

My biggest block right now is finding a prism to focus a superbright LED or IR emitter onto my phototransistors. If this can’t be done quickly, I may buy some laser pointers and crack them open instead (as suggested on the pcomp wiki). Jamie also suggested getting a rotary coupler for each elephant so that they can rotate without getting any wires tangled. So I need to find one of these that is cheap and easy-to-use.

Nime Proposal

My NIME “instrument”/ LIPP performance will consist of a set of electronically augmented objects to be included in a dance/sound/performance project I’m working on with Nancy Garcia. We’ve come up with a narrative about a performer who lives in a post-Castro Cuba 10-20 or so years in the future. A large number of multinational entertainment firms have just evacuated Cuba after a failed attempt at turning the island into a entertainment playground for wealthy tourists, leaving the island with few inhabitants and in a state of technological decay. Our performer, Zoraida, is a one-woman performance act who roams the post-abandonment countryside performing with a set of custom “instrument objects”. These objects are modified heirlooms and trinkets that have been passed down through her family for generations, like these below:

Growing up she learned that the way the objects were arranged and placed inside her house brought about different sorts of fortune and charm to the family and their living space. Now, she incorporates the objects into her performance and music through augmenting them electronically. These objects are, in a sense, her backing band now that Cuba is virtually uninhabited.

I plan to turn these objects into instruments by attaching xbees to them and having their location values sent into Max to control/manipulate an MSP/Jitter patch. Moving an object around will affect the qualities of the sound ascribed to it in the patch (frequency shifting, filtering, scrubbing, etc), as well as affecting a visual backdrop programmed in Jitter. Thus, these “object instruments” will be played by being moved around and relocated. Thus, for each song or section of movement, the objects will be placed a unique location determining the sound enveloping the performer and the space around her.

These objects will be kept inside a briefcase-style container, which will also house the main zigbee receiver circuit as well.

First, I have to figure out how to work/program a zigbee (which won’t be hard since everyone at ITP uses them all the time) and figure out what kind of data the devices send so I can determine how the MSP/Jitter portion of the patch words.

Second, I need to know if Zigbees themselves can transmit locational data so that the objects can let each other know where they are. Then as the objects are moved around, their position can be tracked and fed into max so that data can be used to do something in MSP/Jitter.

If not, I’ll have to hook up some sort of distance sensor that tells the distance of an object in front of one of the zigbee augmented “object instruments,” then send that data to the main zigbee receiver. That would turn these instruments into motion detector type object that are affected by the performer’s movement.

*****UPDATE*********
We won’t be using xbees, as hardwiring these things seems the better option. Right now, we plan to have a stand for each elephant (3 altogether), and each stand will be connected to a potentiometer and a beam-break sensor. These sensors will divide the stage up into 3 spaces, in which we will assign each movements and sounds. There is still much to be worked out.

For my first meditation, I created a patch (download) that slowly and randomly displays each pixel of a image or movie file via a polyrhythmic process. The algorithm divides the display window in half so that the pixels in the left side of the screen are rendered with a “count” of 4, and the pixels in right side of the screen are rendered with a “count” of 3. The 3 count rhythm generates more pixels per cycle than the 4 count, so the right side technically renders quicker than the left, but the difference is just barely perceptible. Below is a photo of the patch, and a video of the process in action (for the input file, I used this colorbars picture):

Click To Play

I was also able to capitalize on a curious glitch I found in jitter. When I increased the speed at which the polyrhythmic beats are triggered, jitter starts throwing large chunks of the file onto the screen creating a weird and wonderful Mondrian-esque sprinkle pattern that I really dig. Below is the video:

Click To Play

Outtakes from my Max/MSP noise generating patch: