Changelog#
Warning
Alpha releases
0.9.3#
Fix issue in transcription fields not showing up
Fix issue in transcription errors not highlighting correctly
0.9.0#
Fixed issues when running Anchor on MacOSX
0.8.1#
Fixed keyboard interaction on table views
0.8.0#
Added support for whisper and speechbrain for transcribing utterances
Added support for trimming utterance silence based on VAD
Added VAD model selection to Model menu
Added Language menu for use in whisper transcription
Fixed a bug where alignment metrics were not being computed
Improved search to better handle cases where diacritics were being treated as non-word symbols
Improved transcription window functionality
Compatibility with MFA 3.2.0
0.7.1#
Fixed bug when editing text
Disabled scrolling to the top when editing pronunciations
0.7.0#
Refactored thread workers
Optimized word and phone interval drawing
Added transcript verification button to alignment window
0.5.0#
Further optimize rendering of intervals for long files
Added ability to use native OS theme instead of MFA theme
Refactored icons and resources to optimize space
0.4.0#
Fixed an issue when adding pronunciation for OOV item from normalized text window
Fixed an issue where volume bar would not save volume properly
Fixed an issue where punctuation would incorrectly label items as OOVs
Fixed a display issue on tables in Windows
Updated diarization code to be compatible with file optimizations in 0.2.0
Improved regex search to not rely as much on postgres character classes (which are not unicode aware)
0.3.3#
Fixed a bug in loading saved dictionaries and models
0.3.2#
Fixed compatibility with Qt 6.7
0.3.1#
Fixed a crash on launch
0.3.0#
Added a corpus manager dialog to allow for cleaning up previously loaded corpora and more efficient loading with dictionaries and models
Enabled volume slider control and mute
Removed sync and reset actions on toolbar in favor of export files
Fixed a visual bug in size of text grid area
Added customizable settings for the maximum duration that pitch tracks and spectrograms will be show (by default 10 seconds is maximum for pitch and 30 seconds for spectrogram)
Added setting for disabling audio fade in on playing
0.2.1#
Fixed bug with calculated columns when merging/splitting and then changing time boundaries
Fixed bug when double clicking on an utterance to focus it
0.2.0#
Revamped and optimized utterance loading to improve performance for long sound files with many utterances
Added ability to click on word intervals to highlight those words in transcripts
0.1.0#
Added ability to swap time axis for right-to-left languages to preferences
0.0.11#
Updated documentation and examples for latest version
Polished some UI elements with style sheets
0.0.10#
Updated compatibility with MFA 3.0