Patching with Mattias

Yesterday I brainstormed my ideas for controlling the EMFountian with Mattias, an MA student at the Academy of Fine Art in Bergen, Norway. The aim of the session was to make a plan for developing a simple control system for the fountain that can be developed over time. We ruminated together over using parts of existing max/msp/jitter patches which we have each developed for 2 different projects that used similar/related soft- and hardware elements. I put my patch for the Emotion Organ (made with the help of, amongst others, Piotr Pajchel) on the table, and Mattias brought out a patch he made for When You Hunger Number One, a sound/light installation by Asbjørn Hollerud.

This is what we ended up with.

1. Start by sending the sound from each sniffer through fiddle~ objects
(text taken from documentation file):

“The fiddle~ object estimates the pitch and amplitude of an incoming sound, both continuously and as a streem of discrete \”note\” events. Fiddle~ optionally outputs a list of detected sinusoidal peaks used to make the pitch determination. Fiddle~ is described theoretically in the 1998 ICMC proceedings, reprinted on http://www.crca.ucsd.edu/~msp.

Fiddle’s creation arguments specify an analysis window size, the maximum polyphony (i.e., the number of simultaneous “pitches” to try the find), the number of peaks in the spectrum to consider, and the number of peaks, if any, to output “raw”. The outlets give discrete pitch (a number), detected attacks in the amplitude envelope (a bang), one or more voices of continuous pitch and amplitude, overall amplitude, and optionally a sequence of messages with the peaks.”

2. Let Fiddle find three groups of frequency ranges HIGH, MEDIUM, LOW, and scale them to DMX values.

3. Use the full range of values to control the manifold water pump.

4. Use the full range of values to control the central jet pump, but when passing from one range group to another, the output signal should be set to zero to create a water drop before it rises agian.

5. Asign HIGH range values to R, MEDIUM to G and LOW to B values to control the RGB LED lights.

6. Use the ATTACK signals to open and close the water jet solenoid valves.

Advertisement

Sound analysis

I have been using Aleksander Refsum Jensenius’s max/msp sound analysis software to analyze digital and analogue signals that I have recorded with my detectors. When I see the sound rendered from different perspectives – loudness (spectral energy), brightness (spectral centroid), noisiness (spectral flatness), sonogram, pitch (estimated), onsets and as musical notation – in this way, I get to know more about the character of each of the various field recordings. The screenshot below is one moment of detecting the electromagnetic activity when I make a call on my mobile phone.

From this exercise I hope to glean some ideas about how to treat the data that will come in from the fountain’s detectors, and how this will control the various water valves of the fountain.

The digital signals are on-off and squirty – like premature ejaculations, while the anaolg signals are continuous and surgey and can be more kind of slowly orgasmic. What I am currently thinking of is passing each peak of the digital signals sequential to the small valves close to the circumference of the fountain bowl. As the peaks happen so fast, I figure that this is the best way of signal distribution – it allows for time for the mechanics of the each of the valves to respond before getting new input. Right now it seems like either the pitch or noisiness signal analysis methods would be the most useful way of controlling the valves. Pitch/peak opens a valve and sends a water jet into the air, the height controlled by its value, and then the valve closes fast to get a “drop”.

If you want to try this out, download the sound of my mobile (unedited from dv cassette/7.89mb) and then open/run it in Aleksander’s sound analysis patch. There’s a osx stand alone version, (1.6mb), so you don’t need to have max/msp installed on your mac.