Lyrics and source handoffs

Identify repeated lyric occurrences by context and position in 调调

Build occurrence cards for two repeated lines in one track edition, keeping neighboring context and audible positions instead of identifying a passage from matching words alone.

English translation of the Chinese original, translated on . Original publication dates, attribution, code and version notes are retained.

Build occurrence cards for two repeated lines in one track edition, keeping neighboring context and audible positions instead of identifying a passage from matching words alone.

1. Choose material with two complete repeated occurrences

Choose an authorised playable edition with viewable lyrics and a complete line occurring in at least two separate passages. This concerns occurrence identity, not a repeated character within one sentence. Establish both through playback/source material and label the common wording R externally without publishing the whole lyric. If the source says only “repeat chorus” rather than supplying another full line, record that representation instead of duplicating text and attributing it to the app. Preserve edition, duration and source; repeated wording need not have identical delivery or timing.

2. Assign external identities to the first and second occurrences

Call the earlier occurrence R-A and the later R-B in external notes, without editing the lyric in 调调. Give each source position, preceding/following-line description and audible cue; copying only R does not create different identities. Without source line numbers use passage/order descriptions rather than invent an app line-number field. If neighbors repeat too, retain wider passage/accompaniment context and state that scope. These references return to this track without assuming one universal song-structure template. Both cards should belong to one beginning-to-end play cycle. A whole-track repeat reaching the same line is the next cycle’s first occurrence, not automatically this cycle’s later second passage. Record cycle identity and any progress reset externally without assuming an app play counter.

3. Practice same wording at different positions on a paper card

Use an explicitly original paper sequence:“门开了;灯还亮着;雨停了;灯还亮着;人回来了。” The repeated“灯还亮着” occupies items two/four, with different neighbors: door/rain, then rain/person. Identical wording does not identify one occurrence. This card supplies no audio and was not automatically installed in 调调; it practices contextual extraction. A real track needs its own context, order and sound evidence rather than treating these five item numbers as player time. As a control, giving both paper occurrences the same preceding “雨停了” defeats preceding-context identification alone. Inspect following or wider context. Name the changed paper card separately rather than alter actual lyrics. Context has scope conditions too; one arbitrary neighboring word need not uniquely identify a passage.

4. Observe the first occurrence from the track beginning

Play from the verified beginning and wait for the first R, recording preceding audible passage, visible context and elapsed position. With timed lyrics record the actual emphasized line; with static text record your reading correspondence without claiming automatic synchronization. R visible on screen alone is insufficient because a view may retain an earlier reading position. Use preceding sound and track identity to establish R-A rather than call a later playhead position the beginning. Preserve this first observation.

5. Check the later occurrence and its following context

Continue to the second occurrence, or approach a known later position through visible progress controls and listen from its preceding passage, recording which method you used. Verify following context/accompaniment or ending rather than R alone. In a paper example with occurrences at twenty/sixty seconds, seeing R near sixty does not establish a return to twenty; use actual track times for real observations. If landing misses the onset, replay from earlier instead of diagnosing missing lyrics from a phrase tail.

6. Separate repeated performance from source abbreviation

If the performance repeats but the source supplies only one body text, record two audible occurrences versus one/abbreviated source representation instead of automatically adding another app-recognized line. A changed word must remain changed: approximate meaning is not exact identity. Keep nonprivate position descriptions or notes rather than whole lyric images. Do not assume tapping a repeated line seeks there. Published scope covers viewing and available timed synchronization, not every unobserved line-jump interaction.

7. Diagnose viewport, occurrence and edition separately

For a mistaken occurrence, check whether playback actually moved backward and whether the text viewport merely stayed on a passage. Repeated visible text and repeated audio are different events. Wrong neighbors call for checking edition/source or a same-name track before editing lyrics. Add R-C for further occurrences instead of merging all later passages into the second. Correct occurrence numbering or contextual notes one at a time and retain the earlier judgment. After seeking, audible context remains necessary; “the later R” alone is not a reproducible reference. If R-A/R-B end up at one position with the same sound/context, check a copied card or whole-track replay. Two records do not establish two distinct occurrences. Reconcile same-cycle order and neighbors, then retain separate cards or mark the second unconfirmed instead of splitting one observation merely to reach two.

8. Reopen and deliver two returnable occurrence cards

Reopen the edition and approach either occurrence from its preceding passage, checking that the record returns to the intended location. Deliver a minimal shared-text identifier, two occurrence labels, their neighbors, actual positions and static/timed status, leaving unseen portions pending. Explain why identical text can represent different occurrences and why one supplied text does not prove one performance. Separate your numbering from source material. Changed translations, edits or server data need remapping; these cards do not establish automatic chorus recognition or song-structure analysis.

Lyric-source checking; Three-anchor timing differences.

Checked on 17 November 2026: Apple’s Russia lookup and China lookup both list 调调 1.0.1, requiring iOS / iPadOS 17.6 or later. The official store release notes include music themes and Home Screen widgets. Users supply audio and servers; identify controls actually present in the installed version. These are repeatable instructions with paper expectations separate from observations, not claimed completed device or private-server tests.

Method references: Official Apple lyric-display screenshot.