The Future of Geotagged Audio

geotagging example

For my inaugural post to this blog, I’d like to write about something I’ve been thinking about lately, and hopefully begin a discussion on it. Namely, what to make of geotagged audio samples and recordings. In case you’re not familiar with the term, geotagging is the practice of assigning geographic coordinates to a piece of media like a recording or photo as a form of metadata. In one incarnation, such as on the Freesound project, geotagged samples are layered over Google maps, allowing one to zoom in on any spot on the planet and potentially find samples tagged to specific geographic locations. As numerous startups and one very large corporation (beginning with a ‘G’ and ending with ‘oogle’) have realized, the commercial potential of geotagging is huge. But we hear less about its scientific potential and, of importance here, its aesthetic potential.

Scientifically, geotagged audio has potential in areas such as the environmental sciences. As one example, imagine taking annual recordings of a section of forest over many years, studying the variations or declines in population of certain bird species via their prominence in the recordings. This has likely already been done, but then imagine putting those incremental recordings into the public sphere via an application like Google Earth.

Of course, as an artist, I am primarily interested in the aesthetic potential of this technology. Currently on Freesound (and hopefully soon on Google Earth too), one can navigate around a map of the world, looking for and listening to geotagged samples, downloading them if one is interested in using them further. However, once the geotagged sample is downloaded and separated from its coordinates, it becomes just another field recording without any accompanying data. For a geotagged sample or recording to be of value compositionally – as a geotagged sample tied to a specific place and not just an anonymous field recording – the metadata must be maintained for compositional use. This is where we apparently reach the edge of current development: tools for working compositionally with geotagged sounds off of a network have not really been developed. There is a multitude of approaches to using this type of material, from composers interested in ecoacoustics to installationists wanting to tap ‘global’ recordings in some improvisatory way. What I’m getting at here is the need for a discussion (hopefully to take place below), about the aesthetic and technical issues surrounding geotagged audio, and tools that composers/artists would like to see available for making the best of this material.

If you were to make use of geotagged audio, what would you use it for? What kind of interfaces into a geotagged audio database would interest you?

Jul 4, 2007
Trackback URL

17 Responses

  1. Peter Swendsen:

    Very useful point of departure for a discussion, Peter. I’ve just returned from a trip to the Lofoten Islands here in Norway, where I was doing some field recordings and thinking quite a lot about issues related to geotagging (it was my first time north of the Arctic Circle–or Polar Circle as they call it here–and I was somewhat obsessed with latitude coordinates in particular).

    The real benefit of geotagging to me is that it has the potential of giving listeners/users (“users” in the sense of someone downloading a sound to use as part of a composition, for instance) a better context for the sound. This is important because we face the interesting conundrum of having access to an increasing array of downloadable environmental sounds while simultaneously having fewer real-world experiences related to our material. While geotagging cannot grant us these experiences, it can at least point us in the right direction.

    For me, the “experience of place”, rather than the gathering of interesting sounds, is the key to creating meaningful sonic interpretations of real-world environments. This may demand, at the very least, the creation and retention of metadata that goes beyond the coordinates to include prose, photos, etc. Even then, one could question whether a site-specfic sound can ever transcend its anonymity if the user has not been on site him/herself.

  2. Chris Joseph » The Future of Geotagged Audio:

    […] Peter Traub blogs about geotagged audio on Networked_Music_Review: […]

  3. jo:

    Hi Peter,

    Though I’m no audio artist, your post poses a number of questions for me. Maps have traditionally tended to be visual, and geotagging–in relationship to mapping interfaces such as Google Earth– seems strongly connected to images. I wonder how one might develop an interface that is primarily audio–where one isn’t looking at a screen? I also wonder about what it is that draws us to online maps in general; I think the answer partially lies in time, “real-time.” Liveness and the ability to tune in to remote places “on-demand” can generate meaningful engagement beyond static compositions. Should “real-time” be a primary ingredient of audio geotagging?

    Finally, how is “place” or “location” conveyed? Images often rely on titles or captions to convey specific geographic information; I imagine the same is true for audio/sound. This type of information matters more when images and sounds are used to document or identify specific places. But artists are often more interested in communicating the “feel” of a place. Is geotagging intrinsically objective/informational rather than subjective/expressive?


  4. peter:

    regarding both peter swendsen’s and jo’s comments, i am also thinking about this from a somewhat visual perspective, insofar as a performance using geotagged audio would likely have to involve some visual display that clues the audience in to the location of the recorded audio. imagine, for example, doing some performance in which the geotag data is used to pull photos off of panoramio (sp?) or flickr that correspond geographically to the audio source you are playing or manipulating. the idea then is that the geotag is a central piece of metadata that could then be used to pull together and correlate numerous types of media such as photos, video, prose, and of course, audio.

    personally, i’m interested in the ability to do this in realtime or at least on the fly. for example, perhaps i want to make an interactive installation in which users can touch a screen some representation of a map or the earth on it. depending on where they touch, an instant collage is created from various pieces of geotagged media at that coordinate. this is just a small slice of what i think is possible (i don’t even want to get into the potentials of the iphone, geotagging, gps, etc… i’ll save that for another discussion)…

  5. jo:

    Hi Peter,

    There are numerous examples of net art projects that, in part, accomplish some of what you’re after. “Simultaneous Translator” allows the user to select and/or add multiple audio streams in real-time, and allows the user to track the streams via Google Earth. Maybe one could add, for instance, flickr or YouTube streams that would pop up at the various nodes along the route?

    Brooke A. Knight’s “Cell Tagging” allows users to locate themselves on Yahoo! Maps via zip code, and to “graffiti” or mark a particular place and record/tag it with an audio message.

    Then there’s Pall Thayer’s gorgeous piece, “On Everything”, that merges various audio, visual, and text streams but is not concerned with place at all.

    You stated in your post that you were concerned with aesthetics so, as always, the question arises: how much aesthetic control can you have when a) random streams from outside sources are part of the piece and b) users are asked to contribute to the piece?

    Update: I just came across several geotagging posts on Unfortunately, it hasn’t been updated in a while.
    Update 2: Satellite Jockey.
    Update 3: SoundTransit


  6. peter:

    hi jo,

    thanks, those pieces/sites look interesting. i suppose i’m also wondering how geotagged audio can be used not just in web-based interfaces and pieces, but in realtime performances, or location-specific performances: for example, perhaps one does a live performance in brussels and the software pulls geotagged media from the surrounding area to mix into your work. or, perhaps you have an interactive piece written for the next generation of iPhones (once people can actually seriously develop on them and they have gps built in), that changes depending on your coordinates, collages location-related media, and so forth…

    regarding your comment about aesthetics, that is always the hard part. i’m not sure if i equate ‘aesthetic’ control of my work with ‘content’ control. for example, in my first net piece bits & pieces from 1999, i grab random sound files off the net and never know anything about their content before collaging them. however, if you were to listen to several half-hour runs of the piece, you’d soon hear the compositionally imposed components as a common thread between the randomly juxtaposed found sounds. so in this sense, i feel like i exercise a great degree of aesthetic control through how the system works, how the collages are made, what processes are used, but have little to no control over the content upon which that system works.

    on another totally different note, i was just reading in one of the posts on about geotagging data being embedded in the EXIF data for jpegs. maybe one day we’ll have a portable field recorder, like the edirol r-09, that has gps built in and automatically geotags field recordings when you make them. that would be useful…

  7. jo:

    Hi Peter,

    I feel a bit dumb asking these questions, but here goes:

    Does the metadata matter to your audience? Why?
    How will your audience know whether the materials you are using are location-specific?

    Thanks for your patience!


  8. peter:

    hi jo,

    i suppose i’m envisioning pieces that have a visual component in which the metadata would be useful from a compositional standpoint. for example, the metadata could be used to pull random photos from adjacent/nearby coordinates off of flickr or some other site. or, perhaps i have an installation like the SoundTransit site you mention above in which i want to upload soundfiles and have them automatically attach to a map or group with other pieces of tagged media. or perhaps i want to use that metadata in a max/msp patch to group samples together based on some common locational connection (essentially using the metadata as another descriptor to organize/group things). does this make more sense?


  9. Delfo:

    I would like to explore metaAudioSpaces
    linking similar sounds in different places of planet
    i call it surfing the layers
    i’m sure something amazing could pop up 😉

    keep me informed about you research


  10. eva sjuve:

    Some other people/projects are exploring this questions as well. I was just pointed to Locusonus project.


  11. eva sjuve:

    Interesting question. One can think of it as participatory installation, using the metadat as room location. But I still think the question is much more complex, as it involves aesthetic and technical aspects.

    I’m looking at some issues with mobility in relation with field-recordings. In general I think the question is related to the different methods involved both technical and aesthetic and the context:
    method of recording, real-time or planned recording session
    sampling rate
    participatory/ non-participatory
    if participatory, if a certain task or theme is given
    mixing in realtime/ non-mixing
    composition, and is the location is part of the concept
    listening experience is this participatory or not

    You can take a look at my geotagging piece for mobile phones:

  12. peter:

    hi eva,

    audioTagger is a nice piece and a great example of what i’m talking about in terms of the metadata integration. one obstacle that you have to deal with in the piece is that users have to manually enter the addresses of their recordings after-the-fact, so that you can properly locate them on the map. this is where some sound file standard (primarly for people who make field recordings) that incorporates geographic metadata would be useful. at the very least, having recordings stamped with coordinates/elevation would be useful for different types of searching, classification, organization, etc. i predict that, given such data automatically and easily accessible in their soundfiles, artists would figure out many interesting things to do with it.


  13. Delfo:

    Hi Peter, Eva

    in order to have such position information “automatically” a GPS module is mandatory in the recording device.
    Mobile phone have also the capabilities to locate the user in the network provider cell but the geographical mapping is fully managed by the operator.
    another way could be to address the recorded file to the geographical address like “5th avenue 1155”
    and use server side a mapping tool for such metadata


  14. eva sjuve:

    hi delfo and peter,
    A gps enabled phone with custom made software that records metadata would be possible even today. The only thing is that most people do not have that kind of device, so there is a limitation, and you just get 8 bit low bandwidth. In the very near future we will have technology for this, 16 bit recording, gps (or galileo), real-time recording, high bandwidth, ubiquity.

    Problems with using triangulation, data generated by the mobile providers is possible, as Botfighters did in early 2000. But then again, this is only on a local basis.

    In my piece audioTagger I use software generated goecode. I wanted to use technology that is accessible to most people, even if there is low bandwidth, and yes you do need to put in a street address.


  15. Dave Miller:

    I’m thinking along similar lines, as a development of my “Tense Nervous Headaches?” project – these were exploratory walks around an area of London, mapping local electromagnetic radiation,

    Basically on the walks we measured the local radiation levels and the GPS coordinates. One of the next intentions is to add this data to Google maps, so people can search for radiation levels by postcode. Ideally we’d like to extend the mapping to all over the UK, and build up a national Google e/m radiation map.

    We measured the radiation levels using a device called an “Electrosmog Detector” which converts e/m radiation to sound (noise). So we had to make a subjective decision on the radiation intensity, based on the noise level. If we could record the sound and then put that also into Google maps, that could be very interesting.

  16. Epicanis:

    The major technical problem here is that other than EXIF for JPEG images and GeoTIFF files, there isn’t any kind of consistent standard for embedding geographic metadata into files.


    I am, coincidentally enough, trying to come up with some kind of solution for it though:

    I have a page set up for it here.

    I’m still in the early stages, but I’m hoping to come up with an easily-used format that someone could throw into any available reasonable metadata field for any kind of file (a “comment” field in an ID3 tag in an MP3, or a “Vorbiscomment” in Ogg/Vorbis, or a comment or description field of a video file or whatever). Comments and suggestions welcome, of course.

  17. jo:

    Geotagging: growing trend of labeling online photos with GPS data adds a new dimension:

Leave a comment


Current interview:
Robin Meier, Ali Momeni and the sound of insects

Previous Interviews:


livestage music sound performance calls + opps installation audio/visual radio festival instrument networked audio interactive experimental electronic workshop video participatory writings event mobile exhibition concert live collaboration electroacoustic environment nature reblog distributed soundscape field recording net_music_weekly improvisation software history locative media space public noise recording immersion voice acoustic sonification lecture generative conference body tool sound sculpture net art art + science VJ/DJ light diy remix site-specific perception mapping film visualization listening laptop algorithmic multimedia city urban data wearable architecture open source game virtual biotechnology sound walk spatialization webcast hacktivism robotic image score platform electromagnetic new media cinema ecology found news composer telematic interface streaming residency interviews/other sensor dance circuit bending synesthesia physical political notation intervention object controller broadcasts conversation narrative second life responsive mashup place technology ambient social network symposium motion tracking hybrid intermedia augmented spoken word livecoding text phonography auralization acousmatic upgrade! gesture opera aesthetics mixed reality resource theory processing 8bit orchestra nmr_commission wireless device toy wireless network theater web 2.0 presentation community surveillance p2p 3D copyright soundtrack research podcast sample feedback psychogeography social chance interdisciplinary tactile recycle interview language systems code emergence presence cassette privacy free/libre software media play chiptune newsletter place-specific archives avatar education haptics activist surround sound audio tour glitch hardware tactical identity bioart asynchronous business tv tangible composition animation jazz transmission arts apps tag e-literature collective microsound relational synchronous Artificial Intelligence conductor convergence reuse simulation ubiquitous synthesizers im/material
3D 8bit acousmatic acoustic activist aesthetics algorithmic ambient animation apps architecture archives art + science Artificial Intelligence asynchronous audio audio/visual audio tour augmented auralization avatar bioart biotechnology body broadcasts business calls + opps cassette chance chiptune cinema circuit bending city code collaboration collective community composer composition concert conductor conference controller convergence conversation copyright dance data distributed diy e-literature ecology education electroacoustic electromagnetic electronic emergence environment event exhibition experimental feedback festival field recording film found free/libre software game generative gesture glitch hacktivism haptics hardware history hybrid identity im/material image immersion improvisation installation instrument interactive interdisciplinary interface intermedia intervention interview interviews/other jazz language laptop lecture light listening live livecoding livestage locative media mapping mashup media microsound mixed reality mobile motion tracking multimedia music narrative nature net art networked net_music_weekly new media news newsletter nmr_commission noise notation object open source opera orchestra p2p participatory perception performance phonography physical place place-specific platform play podcast political presence presentation privacy processing psychogeography public radio reblog recording recycle relational remix research residency resource responsive reuse robotic sample score second life sensor simulation site-specific social social network software sonification sound soundscape sound sculpture soundtrack sound walk space spatialization spoken word streaming surround sound surveillance symposium synchronous synesthesia synthesizers systems tactical tactile tag tangible technology telematic text theater theory tool toy transmission arts tv ubiquitous upgrade! urban video virtual visualization VJ/DJ voice wearable web 2.0 webcast wireless device wireless network workshop writings



May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul
Jun | May | Apr | Mar | Feb | Jan


Dec | Nov | Oct | Sep | Aug | Jul

What is this?

Networked_Music_Review (NMR) is a research blog that focuses on emerging networked musical explorations.


NMR Commissions

NMR commissioned the following artists to create new sound art works. More...
More NMR Commissions


"Two Trains" by Data-Driven DJ aka Brian Foo

Two Trains: Sonification of Income Inequality on the NYC Subway by Data-Driven DJ aka Brian Foo: The goal of this song is to emulate a ride on the New York City Subway's 2 Train ... Read more
Previous N_M_Weeklies


Guest Bloggers:


Massachusetts Cultural Council
Networked: a (networked_book) about (networked_art)
New American Radio
New Radio and Performing Arts, Inc.
New York State Council on the Arts, a State agency
New York State Music Fund
Upgrade! Boston

Turbulence Works