Sonic metadata: Difference between revisions
Caseorganic (talk | contribs) Created page with '{{stub}} ===Definition=== Encoding or relating certain actions or events with sounds, allowing data to be compressed into soundscapes. An example would be to use one type of no…' |
Caseorganic (talk | contribs) No edit summary |
||
| (One intermediate revision by the same user not shown) | |||
| Line 1: | Line 1: | ||
===Definition=== | |||
Encoding or relating certain actions or events with sounds, allowing data to be compressed into soundscapes. An example would be to use one type of notification sound for a text message and another for a phone call. An extended use would be in a proximal notification application, where the distance between two users is defined by sounds that change based on the distance. Term created by Jason Wilson. | |||
===Description=== | |||
Sonic metadata treats sound as a data channel rather than as an alert. The distinction is important. An alert interrupts and says only "attend to me"; sonic metadata carries information in the character of the sound itself, so that the listener learns what has happened without having to look. | |||
The bandwidth available here is larger than it first appears. A sound can vary in pitch, timbre, rhythm, duration, loudness, spatial position and repetition rate, and human hearing discriminates these with considerable precision and almost no conscious effort. A single well-designed sound can therefore encode several dimensions at once — what kind of event, how urgent, from whom, how near. | |||
===Properties of the Auditory Channel=== | |||
Sound has characteristics that make it suited to certain kinds of information and unsuited to others: | |||
* '''Omnidirectional.''' Hearing does not require facing the source, which is why sound works for peripheral information while a screen does not. | |||
* '''Continuously available.''' Ears cannot be closed. This is both the channel's strength and the reason it is so easily abused. | |||
* '''Temporal.''' Sound exists only while it is playing. It suits state changes and ongoing conditions, and suits reference material badly. | |||
* '''Backgroundable.''' Continuous sound recedes from attention and returns when it changes — the property that makes ambient soundscapes possible at all. | |||
* '''Socially public.''' A sound in a shared space is heard by everyone in it, which makes any notification an act of disclosure. | |||
===Related Practice=== | |||
Sonic metadata sits alongside an established research literature. '''Auditory icons''', introduced by William Gaver, use recorded everyday sounds whose real-world source resembles the event being signalled — a crumpling paper sound for deletion. '''Earcons''', developed by Meera Blattner and colleagues and by Stephen Brewster, use abstract structured motifs, where a musical grammar builds meaning compositionally, so that related events share a related phrase. '''Sonification''' is the general practice of mapping data to sound, and the Geiger counter remains its clearest example: an instrument whose entire output is a rate of clicks, comprehensible without training and usable while looking elsewhere. | |||
=== | The proximal notification case in the definition — distance between two people expressed as a changing sound — is a sonification of a continuous variable rather than a discrete event, and it illustrates the channel's real advantage. A number on a screen requires looking, parsing and comparing to a remembered previous value. A rising rate of clicks requires none of these. | ||
===Design Considerations=== | |||
Sonic metadata fails in predictable ways. Distinct sounds must be learnable, which limits how many can be deployed before they become noise. Sounds must be distinguishable from one another in the environments where they will actually be heard. And because sound is public, a system that encodes meaning into sound also broadcasts that meaning to everyone within earshot. | |||
===Related Reading=== | |||
* [[Calm Technology]] | |||
* [[Paralanguage]] | |||
* [[Ambient Intimacy]] | |||
* [[Materials]] | |||
* [[Status]] | |||
[[Category:Cyborg Anthropology]] | |||
[[Category:Calm Technology]] | |||
[[Category:Design]] | |||
Latest revision as of 13:14, 25 August 2026
Definition
Encoding or relating certain actions or events with sounds, allowing data to be compressed into soundscapes. An example would be to use one type of notification sound for a text message and another for a phone call. An extended use would be in a proximal notification application, where the distance between two users is defined by sounds that change based on the distance. Term created by Jason Wilson.
Description
Sonic metadata treats sound as a data channel rather than as an alert. The distinction is important. An alert interrupts and says only "attend to me"; sonic metadata carries information in the character of the sound itself, so that the listener learns what has happened without having to look.
The bandwidth available here is larger than it first appears. A sound can vary in pitch, timbre, rhythm, duration, loudness, spatial position and repetition rate, and human hearing discriminates these with considerable precision and almost no conscious effort. A single well-designed sound can therefore encode several dimensions at once — what kind of event, how urgent, from whom, how near.
Properties of the Auditory Channel
Sound has characteristics that make it suited to certain kinds of information and unsuited to others:
- Omnidirectional. Hearing does not require facing the source, which is why sound works for peripheral information while a screen does not.
- Continuously available. Ears cannot be closed. This is both the channel's strength and the reason it is so easily abused.
- Temporal. Sound exists only while it is playing. It suits state changes and ongoing conditions, and suits reference material badly.
- Backgroundable. Continuous sound recedes from attention and returns when it changes — the property that makes ambient soundscapes possible at all.
- Socially public. A sound in a shared space is heard by everyone in it, which makes any notification an act of disclosure.
Related Practice
Sonic metadata sits alongside an established research literature. Auditory icons, introduced by William Gaver, use recorded everyday sounds whose real-world source resembles the event being signalled — a crumpling paper sound for deletion. Earcons, developed by Meera Blattner and colleagues and by Stephen Brewster, use abstract structured motifs, where a musical grammar builds meaning compositionally, so that related events share a related phrase. Sonification is the general practice of mapping data to sound, and the Geiger counter remains its clearest example: an instrument whose entire output is a rate of clicks, comprehensible without training and usable while looking elsewhere.
The proximal notification case in the definition — distance between two people expressed as a changing sound — is a sonification of a continuous variable rather than a discrete event, and it illustrates the channel's real advantage. A number on a screen requires looking, parsing and comparing to a remembered previous value. A rising rate of clicks requires none of these.
Design Considerations
Sonic metadata fails in predictable ways. Distinct sounds must be learnable, which limits how many can be deployed before they become noise. Sounds must be distinguishable from one another in the environments where they will actually be heard. And because sound is public, a system that encodes meaning into sound also broadcasts that meaning to everyone within earshot.
