Midterm: Tidal Observation Instrument

For my midterm project I developed the Tidal Oscillator Max/MSP patch described and prototyped in my first sketch.

Looking at what?

The project focuses on four tides around the world, and allows the observer to use the instrument to look closely at each. The tides are in Wood’s Hole, MA; Zhapo, China; Mombasa, Kenya; and the Vernadsky Research Base in Antarctica. The International Oceanographic Commission provides access to this data for free use. I chose these locations for the interesting patterns in their tides (being in a bay or on a certain coast formation can change how the tide is experienced), for the distance separating them, and for the consistency of their data.

How does it look?

This implementation of the instrument also uses audible oscillators (set to 330Hz) so that the changes in the low-frequency oscillators may be heard by a listener. What you hear is actually the month-long tidal frequency repeated 330 times a second, and then modulated by the tidal frequency at a pace much closer to its orginal speed. Max/MSP doesn’t seem to support floating-point math at a small enough value to get a true tidal period. The tides oscillate once every 12 hours and 25 minutes, for a total of 1/745Hz or 0.00002237136Hz.

Regardless, the result you can see and hear above creates an interesting pattern of resonance, in which the irregularities of the tide from one day to the next become audible. The difference in texture between one wave and the next also communicate something about the minute-to-minute shifts in the tide from one wave to the next. Using the MIDI knobs I have set up, an observer can shift between an incredibly slow observation of the change and one that——while still pretty slow for musical applications——provides a much faster view of the tidal phenomenon.

There are a few interesting quirks to this particular presentation of the data. First, my access of the databases through Max is based largely on index, not so much on date. For Zhapo (where there is regularly updated and infrequently missing data) this works quite well. However Wood’s Hole has a much less reliable set of data (although I have also noticed the reliability and consistency of datasets can vary from one month to the next). The result is that the wave representing Wood’s Hole is not technically representative of one month, it’s representative of 10,080 data entries, one for each of the 10,080 minutes in a 30-day month. Its gaps become audible absences in the data.

Next?

I’m very pleased with how much I’ve been able to learn about Max/MSP, about tidal data, and about observation. I do think there are still more avenues to explore in this project. I would be interested in taking on the challenge of having data as live as possible (potentially interpolated), and the challenge of bringing these generated waveforms into an environment more conducive to sonic exploration.