Music IDPrivate beta
Know which track plays where — even under dialogue.
Music ID fingerprints your library and reads any mix: a delivered score, an editor's reference cut or the broadcast copy. Out comes the cue sheet.
Pipeline
How it works
From your library to the cue sheet, in four steps.
- 01
Fingerprint the library
Each track gets a blueprint and a fingerprint. Music ID works on fingerprints alone: nothing is embedded in or added to your original files.
Originals untouched - 02
Run any mix
A delivered score WAV, an editor's reference video, or the broadcast copy with dialogue on top. A source-separation pass lifts the music out from under the dialogue.
Source separation - 03
Identify
Which library file, where it enters and exits (timecode), which source seconds were used, how it was edited (cuts, layers, levels), and identical-sounding duplicates.
File · timecode · edits - 04
Deliver
An automatic cue sheet (CSV), a piece list and a DAW import file.
CSV · piece list · DAW
Measured in-house
- No dialogue99.1%
music identified correctly (up to 99.9% on later episodes)
- Dialogue on top97.9%
with dialogue at the same level
- Music −10 dB95.2%
music 10 dB under dialogue
- Music −20 dB86.3%
music 20 dB under dialogue
- Entry timing98%
entries placed within ±0.5 s (clean music)
- Speed~3min
per episode
In-house bench, 3 Oct 2026: 7 broadcast episodes, ~3.8 hours of music, real dialogue mixed in at each level. Real editor reference cuts with dialogue: 96.9–98.8%.
- With source separation
- Without (shown where it differs)
| Dialogue level | With source separation (%) | Without (%) |
|---|---|---|
| No dialogue | 99.1 | within 0.6 pt |
| Same level (0 dB) | 97.9 | within 0.6 pt |
| Music 10 dB under | 95.2 | within 0.6 pt |
| Music 20 dB under | 86.3 | 78.2 |
Share of music correctly identified. Without source separation the result is within 0.6 points everywhere except music 20 dB under dialogue. In-house bench, 3 Oct 2026: 7 broadcast episodes, ~3.8 hours of music.
Out comes the cue sheet
| # | Track | In | Out | Duration | Source in/out | Use |
|---|---|---|---|---|---|---|
| 1 | Track A | 01:00:14:12 | 01:00:50:08 | 0:36 | 0:08 – 0:44 | Background |
| 2 | Track B | 01:01:06:02 | 01:01:47:15 | 0:41 | 1:12 – 1:53 | Background |
| 3 | Track C | 01:02:31:00 | 01:02:59:10 | 0:28 | 0:00 – 0:28 | Featured |
Placeholder rows to show the format. The real export is a CSV cue sheet, plus a piece list and a DAW import file.
Use cases
Cue sheets for broadcasters and rights societies
Every music use, with timecodes and the source seconds used, exported as a CSV.
Find the exact temp track
Run the editor's reference cut and see which library file sits behind each temp cue.
Verify what actually aired
Read the broadcast copy, dialogue and all, and see which tracks went to air and where.
Honest limits
- Music outside your library isn't named.
- Very quiet music (about 17 dB or more under dialogue) can be missed.
- Exits under dialogue are less precise than entries.
Run Music ID on one episode.
Private beta for broadcasters, studios and post houses.