RHVoice Versions Save

What's Changed

• Fix NVDA driver availability without installed voices by @kvark128 in https://github.com/RHVoice/RHVoice/pull/443
• add Polish voice by @grzezlo in https://github.com/RHVoice/RHVoice/pull/445
• Warn about the need to install voices in the NVDA addon by @kvark128 in https://github.com/RHVoice/RHVoice/pull/446
• A build on Linux shouldn't fail without Speech Dispatcher. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/454
• mgc with world by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/459
• voice-building-utils correction and convert-audio to py3 migration. by @grzezlo in https://github.com/RHVoice/RHVoice/pull/460
• Partially parallel context clustering. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/465
• Core: handle binary feature flags in the syllabification output. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/466
• Voice training: support additional iterations of tree context clustering. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/467
• Update Ukrainian dictionary by @VovaMobile in https://github.com/RHVoice/RHVoice/pull/468
• SSML: support the phoneme element, only native phonemes are supported. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/469
• Voice-building-utils: save the trim positions in the original recordings. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/472
• New male polish voice: Natan by @zstanecic in https://github.com/RHVoice/RHVoice/pull/473
• NVDA addon: Move Rate Boost setting to the conventional position. by @TheQuinbox in https://github.com/RHVoice/RHVoice/pull/490
• Segmentation, experimental: use vad to initialize the silence model. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/491
• Voice-building-utils: the export-voice command calculates and saves the key parameter. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/500
• Core: sentence level tokenization support. by @louderpages in https://github.com/RHVoice/RHVoice/pull/501
• Native data package manager, Android-only for now. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/452
• Add script to convert text files to SSML format by @cyrmax in https://github.com/RHVoice/RHVoice/pull/504
• NVDA addon: Make RHVoice compatible with NVDA 2022.1 by @zstanecic in https://github.com/RHVoice/RHVoice/pull/507
• Macedonian Language Updates by @louderpages in https://github.com/RHVoice/RHVoice/pull/508
• Fixe segmentation fault (core dumped) on Georgian language by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/516
• Improved support for the reuse of msi package files in SAPI voice installers. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/523
• cmake: Support new speech-dispatcher modules location by @living180 in https://github.com/RHVoice/RHVoice/pull/535
• RHVoice-test: don't ignore the quality setting in the configuration file. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/538
• NVDA addon: correct the project website url. by @Olga-Yakovleva in https://github.com/RHVoice/RHVoice/pull/539

New Contributors

• @savoptik made their first contribution in https://github.com/RHVoice/RHVoice/pull/429
• @kvark128 made their first contribution in https://github.com/RHVoice/RHVoice/pull/443
• @TheQuinbox made their first contribution in https://github.com/RHVoice/RHVoice/pull/490
• @louderpages made their first contribution in https://github.com/RHVoice/RHVoice/pull/501
• @cyrmax made their first contribution in https://github.com/RHVoice/RHVoice/pull/504
• @gerazov made their first contribution in https://github.com/RHVoice/RHVoice/pull/534
• @living180 made their first contribution in https://github.com/RHVoice/RHVoice/pull/535

Full Changelog: https://github.com/RHVoice/RHVoice/compare/1.6.0...1.8.0

• Albanian language
• Voices: add tatiana (#386)
• Voices: add Mikhail
• Voices: add Marianna (#400)
• Voices: add lyubov (#412)
• Voices: add Vitaliy
• Voices: add Albanian voice, Hana, North Macedonian accent
• Update emoji data to version 40
• Android: Clean up and update dependencies.
• Android: fix the version display if there are no voices installed.
• Android: users can now select a dictionary file on their device and import it into the app.
• Android: don't create an initial version of the config file.
• Android: the option to import a config file.
• Android: target Android 11.
• Android: fix a crash.
• Android: correct some translations.
• Core: a language developer may set custom flags on segments in post-g2p and use them in their hts labels.
• Core: g2p.case setting.
• Core: word_break feature.
• Core: more hts features.
• Core: data-only languages.
• Core: data-only languages get their list of graphemes from a file.
• Core: voices may provide equalization filters.
  • Use a cascade of biquad filters.
  • Save the coefficients in 24000/eq.txt.
  • First line is a header with the version number. In case I decide to change the format.
  • Then coefficients separated by whitespace.
  • For each filter: b0, b1, b2, a0, a1, a2.
• Core: enable the limitter if equalization is used by the voice.
• Core: general support for a simple "accented" decision tree.
• Core: Failing to load the eq shouldn't cause the voice loading to fail.
• Core: support comments in eq.txt.
• Core: if the name of a phone starts with /, remove the / after syllabification.
• Thvoice-transcribe-sentences: the -v option now also prints out gpos if not content.
• Brazilian Portuguese: fix the pronunciation of the word backup.
• Brazilian Portuguese: pronunciation of the ordinal indicator characters.
• update Ukrainian dictionary
• honor 'SOURCE_DATE_EPOCH' environment variable if it's set. this helps rhvoice build reproducibility.