“Timelapse” usually refers to the process of sampling small bits of video or film and piecing them together to form a sped-up version of reality. (Actually, that’s not entirely accurate. Any recording involves sampling small bits of time. Timelapse simply plays back those samples at a rate faster than reality, so that instead of playing back film frames recorded at 30 frames per second at a playback speed of 30 frames per second, you play back film recorded at one frame every ten minutes at 30 frames per second, for example.)

What if you made a timelapse of sound, and not simply image? Reader Andrew Spitz did that, building a sound-sampling app in visual development tool Max/MSP. He’s made the resulting tool available to anyone using Mac or Windows, for free, so you can try it yourself. In the demo video, what you get is a stuttering, rhythmic montage of found sound. But change the material or setting, and perhaps you can get very different results.

I love the word he’s using here: “phonography.”

phonoLapse is a free desktop app for Mac and Windows that lets you create audio time-lapses. For the 2010 Enterferenze New Art Festival I put together a little Time Lapse Phonography piece that followed me over the course of 24 hours (check the video below). I have been receiving emails from people wanting to create their own, and decided to work on a standalone version so you too can create some time-lapse phonography :-) .

Grab it yourself:
phonoLapse {+ software} [sound+design]

By the way, Andrew is responsible for one of my other favorite recent projects:
Voice Messages Become 3D Paper Waveform Sculptures: Paper Note

  • ReidDP

    I don’t think he can claim to have invented the term Phonography!


    • http://pkirn.com/ Peter Kirn

      Yeah, sorry – I adjusted what I say there! It’s a nice term to *use*, even if he didn’t invent it. My mistake, not his.

  • http://www.facebook.com/ersheff Eric Sheffield

    I like the idea, but not so much the execution.
    The magic of time lapse photography is that it’s reassembled and presented as an other-wordly whole.
    This just sounds like a bunch of tiny audio snippets played one after the other. Because that’s what it is (of course).
    I’d be interested in hearing this done in a way that seamlessly stitches the audio snippets together.

    • Martin W

      That is very much easier said than done, in part because of the fundamentally different nature of sound and image, and in part because of the perhaps even more fundamentally different nature of how our eyes and our ears work in conjunction with our brain. ( And BTW, the image here is every bit as jerky and disjointed as the sound is for the most part, especially when the camera is moving, so it doesnt have the “magic of time lapse photography’ in the way I think that you mean anyway )
      However, it would be very easy to put a crossfade on the sound slices which would to some extent reduce the clickiness of the sound, but if a car goes by in one slice and doesnt in the next, you are still going to have specific short loud events jumping out at you that we perceive very differently from the effect of a car flickering into frame in a classic timelapse cityscape … Such is the nature of hearing …

    • http://www.facebook.com/ersheff Eric Sheffield

      The crossfade thing is exactly what I was thinking. Not only is it clicky, but there are minute pauses between sounds that kill the illusion.
      I’m sure there are more interesting (and difficult) ways to combine the sounds, but the crossfade technique would at least start to blur the boundaries of many individual moments (which is what we’ve got now).

    • http://pkirn.com/ Peter Kirn

      Actually, it’d be possible to build on the idea here in this way — it’s doing basically linear buffer sampling (lots of little slices), and what you could do instead is a bunch of sonic grains sampled over a long period of time (effectively just windowing those slices differently). Wouldn’t be quite film timelapse, and wouldn’t even be conventional granular resynthesis, but could be interesting.

    • http://twitter.com/joshgiesbrecht josh giesbrecht

      Maybe decompose the individual samples via some kind of spectral analysis thingummy, and then reconstruct a new blended sound that combines the spectra?

      Just interleaving the slices seems to me like it’d just turn into white noise, whereas reconstructing something based on some kind of fourier analysis would let you combine the actual sound characteristics rather than shred them beyond recognition.

      (Lots of maybe’s in there though)

    • Vanya Perevozov

      It would be cool to record audio with an infrasonic microphone and speed it up to match the time lapse. Normal sounds will become squeaks, of course, but we’d be able to hear infra-sounds and associate them with sources in the video. I think that would create a more consistent visual-audio experience in this time frame.

  • papernoise

    I really have to try this, because the idea is just brilliant!

  • Robin Parmar

    “I love the word he’s using here: ‘phonography.’”

    Then you’ll love the fact there’s a user group called “phonography” just full of people who use the term. :-)

  • John

    i used to do the poor man’s version of this quite a bit – just a quick record on/off with a handycam.  of course that was a different concept than this — i like the slice of ordinary life/routine of this result.  my way, i liked being able to really hone in on images / movement / and even timbre of ambient surroundings.  …you could also try blending the segments of this phonographic experiment with a tiny bit of short reverberation  (or even delay, considering that all the segments are exactly the same length)

  • Erik