Easily scrape-able sources of varying data on the web, ideally reflecting natural phenomena? I need some help finding sources.
I've just developed (With the help of
NetMidi,
Audiomulch,
Python and
MidiYoke) a system whereby a Python script takes constantly changing inputs from the internet, and manipulates ambient soundscapes accordingly. For instance, at the moment I have the system hooked up to METAR weather data, and increasing temperature increases the frequency of a flanger effect on a rhythm track. Inputs are also taken for air pressure, wind speed etc.
My problem is that this data isn't very interesting on a short time scale. I can sit here for three hours, watch the temperature drop 5 degrees, and this is manifest as only a very minimal change in the soundscape. In particular, METAR data is only updated every 15 minutes or so - if I was able to track individual gusts of wind it would be great, but I don't know where this sort of data may be available.
Does anyone have any ideas, with associated source URLs, of other data I could use as input? I need something that varies on the timescale of minutes, rather than hours, and which changes chaotically but not completely randomly, but I'm having trouble thinking what phenomena might be interesting to track, that fits these criteria. I'd prefer natural phenomena to something like the stockmarket. The data can be binary or categorical as well as continuous.
What would happen if you fed audio data in?
Two suggestions:
(1) Beatles In -> ???? Out
(2) Feed the output back into the input, with a short delay
As to data...
Earthquake data. I'm sure I've seen web pages hooked straight to seismographs.
Webcams. 'net-accesible telescopes.
Position of the planets (calculated, rather than observed, obviously)
Satellite data (sequential tiles from Google Maps)
posted by Leon at 7:04 AM on June 2, 2006