From sound to a trail: discovering music without simply following recommendations

A question about what you hear opens other paths

Sometimes a fruitful discovery begins with a detail: a brief attack, a slowly shifting timbre or an unusual repetition. Learning Synths offers terms and audio examples that allow such impressions to be described more precisely. This greater precision is useful because otherwise the question of similar-sounding music remains very vague. Someone who can distinguish between pitch movement and changes in volume, for example, can formulate more targeted search ideas. This does not yet allow the instrument used to be identified with certainty. Different processes can produce similar results, and recordings further alter the impression. Terms for describing sound therefore initially serve as working hypotheses. They help us compare consciously, without jumping from an audible characteristic to conclusions about a particular device, person or historical origin of the sound.

Distinguishing recordings, works and releases

The MusicBrainz data structure distinguishes between works, recordings, releases and release groups, among other things. This differentiation is particularly useful for research. A musical work may exist in several recordings, and the same recording may appear on different editions. Anyone who confuses these levels can easily mistake a reissue for a new recording. Relationships between people, places and releases can also open up new directions for searching. The database provides information about music, not the audio files themselves. To listen, other legally accessible sources must be sought out. The real benefit lies in a traceable chain of connections: from a recording, the route leads to contributors, from there to other projects and finally to music that a search based purely on similarity might not have suggested.

A discovery route needs limits

A listening route you build yourself quickly becomes hard to navigate if every new name immediately triggers ten more searches. A simple rule is more practical: for each starting recording, initially follow only two relationships, such as one contributor and another interpretation of the same work. Add a brief note explaining why the connection is interesting. After a few tracks, it is worth pausing. Which differences were actually audible, and which existed initially only on the screen? Asking this guards against confusing research with musical experience. Data can enable encounters, but cannot anticipate them. This creates a form of digital discovery in which information and listening take turns. The route remains open to surprises without dissolving entirely into an endless list of new names.

Keep exploring

• Listen more closely, claim less: a culture of attention
• When Lines Sound: Daphne Oram’s Different Vision of Composition
• Searching with patience: making sense of digital music collections