See the v1.7.0 release notes for full details about integrated transcription, the major new feature.
Transcription improvements
This release features significant improvements to the transcription flow, particularly around the way subtitles are divided/grouped into lines. Using a mix of sources, each provider was assessed and analyzed for weaknesses, resulting in some global and specific improvements.
Each provider uses its most reliable signal for content. Sentences that could be dropped or scrambled by the processing are more reliably complete and correct, and timing is validated and adjusted if necessary to avoid over-long or over-short lines.
Most improved is Qwen Local, which is now also fully compatible with macOS.
Multi-lingual transcriptions are handled more intelligently, with language-aware timing validation. More aggressive timing adjustments can be enabled with a new timing_correction_factor (from 0 to 1).
Other improvements to the handling of line breaks, dialog markers and filler word removal should produce translations that need no further work more of the time.
Optimised startup time
The initialisation of expensive provider SDK's is now deferred, which drastically reduces start up times. A background task warms the selected default provider once the main window is opened.
Improvements to the settings dialogues prevent freezing when other providers are selected, with background initialization and supported model retrieval.
Qwen Local transcription on macOS
Torch installation was fixed for macOS, and compatibility issues between qwen-asr and PySide6 were resolved. The macOS package was advanced to Python13 to fix compatibility - an existing Torch installation must be on Python13 to be used for Qwen transcription.
Various fixes and improvements
- Selection fixes: Split Batch is available again when one line is selected, and Translate Selection covers exactly what's selected.
- Merged subtitles no longer disappear from the subtitle view, and undoing a merge restores the lines correctly.
- If the translation provider can't be created , translation stops with an error rather than failing for every scene.
- Subtitle files with duplicate line numbers are renumbered when loaded.
- A rejected project file no longer clears the subtitles that are already loaded.
- Fixed an empty proxy setting for OpenRouter and Requesty.
Main changes
- Fix macOS test failures, quiet test noise, and frozen build Qwen startup by @machinewrapped in #443
- Fix merged subtitle display and undo by @machinewrapped in #442
- Simplify subtitle row mapping by @machinewrapped in #444
- Fall back to best available break point when no balanced one exists by @machinewrapped in #441
- Optimise GUI startup and warm used providers by @machinewrapped in #448
- Fixed torch setup for macos by @machinewrapped in #449
- Optimised opening settings dialog by @machinewrapped in #450
- Make provider failures terminal for a translation run by @machinewrapped in #452
- Transcription line assembly: parts first, speaker-aware merging, timing fixes by @machinewrapped in #451
- Make ProjectSelection reflect the explicit selection, and fix Split Batch by @machinewrapped in #456
- Renumber duplicate subtitle line numbers during load by @machinewrapped in #455
- Transcription: derive lines from the transcript for words-only providers, and fix Muse turns over silence by @machinewrapped in #459
- Fix test logging: real integration counts, quiet renumbering warnings by @machinewrapped in #465
- Add a per-provider timing correction factor for transcribed lines by @machinewrapped in #467
- Make transcription chunk bounds provider settings with per-provider defaults by @machinewrapped in #470
- Send app attribution headers with OpenRouter transcription requests by @machinewrapped in #472
- Cut transcription chunks at short or quiet pauses instead of at the cap by @machinewrapped in #471
Full Changelog: v1.7.0...v1.7.1