Segment Time-Series are SongScore's visualisation of how key audio dimensions evolve over the full duration of a track — a dynamic map of the track's structure, rather than a single static average.
SongScore generates time-series charts for: genre confidence (how strongly the dominant genre is present at each point), mood dimensions (how emotional character shifts over time), vocal presence (when vocals appear, disappear, and vary in intensity), valence (how the track's positivity/happiness changes), and arousal (how listener activation varies throughout).
These charts are commercially useful in several ways. They reveal structural problems — a track whose energy drops dramatically in the second minute is likely to trigger skip behaviour at that point. They identify hook timing — precisely when the track's most engaging moments occur. They show intro and outro length — long, low-energy intros and outros are algorithmic risk factors.
For producers and mixers, the Segment Time-Series provides a data-driven structural analysis that complements their ear-based assessment. For artists, it shows exactly which part of the track needs attention before release.