Skip to content

Releases: VRCWizard/TTS-Voice-Wizard

v1.6.8.1

09 Apr 20:03
@VRCWizard VRCWizard
v1.6.8.1
8ae6741
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.6.8.1

v1.6.8.1

  • minor fixes to chatgpt integration tab

v1.6.8

  • OpenAI TTS
    • Available with Pro and with your own API key
  • ChatGPT Integration (Text to ChatGPT to Speech)
    • Available with Pro and with your own API key. (with pro you can send a prompt and your current message. With your own key you can send full unlimited context)
  • Code optimization to reduce response times, a couple ms have been shaved off of several TTS and STT options
    • Faster responses from pro tts options and deepgram speech to text
  • VRChat chatbox typing indicator toggle
  • Minor UI changes
    • Deepgram silence scale
    • tooltips added for some ui elements
  • Bug fixes
    • fixed issue when switched stt methods when whisper was running

v1.6.7.3

  • Deepgram Aura TTS (english voices only) available on the dev branch of Voice Wizard Pro
  • Added button to add progressbar for media integration
  • Added tooltips to media integration buttons and some general setting buttons
  • Added Albanian language for TTS though Azure/Pro and Speech to Text via Azure.
  • Added "/TTSVoiceWizard/TextToText" OSC endpoint
  • hotkeys no longer trigger multiple times when held down
  • Removed zh-CN-XiaoxuanNeural voice from azure voice list
Assets 4
Loading

v1.6.8

31 Mar 16:59
@VRCWizard VRCWizard
v1.6.8
182f035
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.6.8 Pre-release
Pre-release

v1.6.8 (current)

  • OpenAI TTS
    • Available with Pro and with your own API key
  • ChatGPT Integration (Text to ChatGPT to Speech)
    • Available with Pro and with your own API key. (with pro you can send a prompt and your current message. With your own key you can send full unlimited context)
  • Code optimization to reduce response times, a couple ms have been shaved off of several TTS and STT options
    • Faster responses from pro tts options and deepgram speech to text
  • VRChat chatbox typing indicator toggle
  • Minor UI changes
    • Deepgram silence scale
    • tooltips added for some ui elements
  • Bug fixes
    • fixed issue when switched stt methods when whisper was running

v1.6.7.3

  • Deepgram Aura TTS (english voices only) available on the dev branch of Voice Wizard Pro
  • Added button to add progressbar for media integration
  • Added tooltips to media integration buttons and some general setting buttons
  • Added Albanian language for TTS though Azure/Pro and Speech to Text via Azure.
  • Added "/TTSVoiceWizard/TextToText" OSC endpoint
  • hotkeys no longer trigger multiple times when held down
  • Removed zh-CN-XiaoxuanNeural voice from azure voice list
Assets 3
Loading

v1.6.7.3 Deepgram Aura

08 Mar 17:44
@VRCWizard VRCWizard
v1.6.7.3
d7e6f62
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.6.7.3 Deepgram Aura Pre-release
Pre-release
  • Deepgram Aura TTS (english voices only) available on the dev branch of Voice Wizard Pro
  • Added button to add progressbar for media integration
  • Added tooltips to media integration buttons and some general setting buttons
  • Added Albanian language for TTS though Azure/Pro and Speech to Text via Azure.
  • Added "/TTSVoiceWizard/TextToText" OSC endpoint
  • hotkeys no longer trigger multiple times when held down
  • Removed zh-CN-XiaoxuanNeural voice from azure voice list
Assets 3
Loading

v1.6.7.1

21 Feb 00:57
@VRCWizard VRCWizard
v1.6.7.1
a57bac4
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.6.7.1
Assets 4
Loading

v1.6.7

20 Feb 22:09
@VRCWizard VRCWizard
v1.6.7
35315ac
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.6.7

v1.6.7 (current)

  • Updated the CoreOSC package to allow for receiving utf8 characters via OSC
  • Changed the elements of the progress bar variable {progressBar E:◯ L:13} to use more fitting utf characters. ┣━━━━┫ vs |-----------|
    • You will likely need to lower the length (L) variable for the change in characters. A length of 13 will fill a line of the VRChat chatbox.
      image

v1.6.6.8

  • 2 optional override bools added for OSC to TTS endpoint /TTSVoiceWizard/TextToSpeech
    • bool useChatbox
    • bool useKillFrenzyAvatarText

The bools allow you to optionally forward text the VRC Chatbox or Killfrenzy Avatar Text without adjusting the in app settings. When sending an OSC message to this endpoint you may still send just a string or both bools (sending only one of the bools will not override the settings).

Example of possible combinations in TouchOSC:
image

v1.6.6.7

  • fixed vrchat counter causing timeout if different counters were activated in quick succession.
  • spotify exceptions should always give a message.

v1.6.6.5

  • minor bug fixes
  • ElevenLabs turbo model added
  • removed voices for TikTok API that no longer functioned (optimus prime voices)
  • New Azure Voices:
es-ES | Spanish (Spain) | XimenaNeural (Female)
fr-CA | French (Canada) | ThierryNeural (Male)
fr-FR | French (France) | VivienneNeural (Female)
it-IT | Italian (Italy) | GiuseppeNeural (Male)
ko-KR | Korean (Korea) | HyunsuNeural (Male)
pt-BR | Portuguese (Brazil) | ThalitaNeural (Female)
zh-CN-XiaoxiaoDialectsNeural
de-DE | German (Germany) | de-DE-FlorianMultilingualNeural (Male)
de-DE | German (Germany) | de-DE-SeraphinaMultilingualNeural (Female)
en-US | English (United States) | en-US-AvaMultilingualNeural (Female)
en-US | English (United States) | en-US-EmmaMultilingualNeural (Female)
fr-FR | French (France) | fr-FR-RemyMultilingualNeural (Male)
en-US | English (United States) | en-US-BrianMultilingualNeural (Male)
en-US | English (United States) | en-US-AndrewMultilingualNeural (Male)
fr-FR | French (France) | fr-FR-VivienneMultilingualNeural (Female)
zh-CN | Chinese (Mandarin, Simplified) | zh-CN-XiaoxiaoMultilingualNeural (Female)
zh-CN | Chinese (Mandarin, Simplified) | zh-CN-XiaochenMultilingualNeural (Female)
zh-CN | Chinese (Mandarin, Simplified) | zh-CN-YunyiMultilingualNeural (Male)

v1.6.5.2

  • uberduck voice selection fix (although some voices in the list no longer function)
  • now using latest version of windows media controller, (report any crashing and include event viewer details)
  • {albumArtist} added for windows media and Spotify media integration

v1.6.5

  • vrchat "/chatbox/typing" indicator added for whisper and deepgram (if VAD is active)
  • fix for whisper timestamps being offset for VAD so super short activations wouldn't be recognized
  • possible fix for group in word replacement tab breaking on some computers
  • hotkeys backend refactored, new hotkeys added for changing voice presets quickly
Assets 4
Loading

v1.6.6.8

12 Feb 23:56
@VRCWizard VRCWizard
v1.6.6.8
b635311
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.6.6.8 Pre-release
Pre-release
  • 2 optional override bools added for OSC to TTS endpoint /TTSVoiceWizard/TextToSpeech
    • bool useChatbox
    • bool useKillFrenzyAvatarText

The bools allow you to optionally forward text the VRC Chatbox or Killfrenzy Avatar Text without adjusting the in app settings. When sending an OSC message to this endpoint you may still send just a string or both bools (sending only one of the bools will not override the settings).

Example of possible combinations in TouchOSC:
image

Assets 3
Loading

v1.6.6.7

08 Feb 23:03
@VRCWizard VRCWizard
v1.6.6.7
d57f772
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.6.6.7 Pre-release
Pre-release
  • fixed vrchat counter causing timeout if different counters were activated in quick succession.
  • spotify exceptions should always give a message.
Assets 3
Loading

v1.6.6.5

03 Feb 18:26
@VRCWizard VRCWizard
v1.6.6.5
284b58e
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.6.6.5 Pre-release
Pre-release

v1.6.6.5 (current)

  • minor bug fixes
  • ElevenLabs turbo model added
  • removed voices for TikTok API that no longer functioned (optimus prime voices)
  • New Azure Voices:
es-ES | Spanish (Spain) | XimenaNeural (Female)
fr-CA | French (Canada) | ThierryNeural (Male)
fr-FR | French (France) | VivienneNeural (Female)
it-IT | Italian (Italy) | GiuseppeNeural (Male)
ko-KR | Korean (Korea) | HyunsuNeural (Male)
pt-BR | Portuguese (Brazil) | ThalitaNeural (Female)
zh-CN-XiaoxiaoDialectsNeural
de-DE | German (Germany) | de-DE-FlorianMultilingualNeural (Male)
de-DE | German (Germany) | de-DE-SeraphinaMultilingualNeural (Female)
en-US | English (United States) | en-US-AvaMultilingualNeural (Female)
en-US | English (United States) | en-US-EmmaMultilingualNeural (Female)
fr-FR | French (France) | fr-FR-RemyMultilingualNeural (Male)
en-US | English (United States) | en-US-BrianMultilingualNeural (Male)
en-US | English (United States) | en-US-AndrewMultilingualNeural (Male)
fr-FR | French (France) | fr-FR-VivienneMultilingualNeural (Female)
zh-CN | Chinese (Mandarin, Simplified) | zh-CN-XiaoxiaoMultilingualNeural (Female)
zh-CN | Chinese (Mandarin, Simplified) | zh-CN-XiaochenMultilingualNeural (Female)
zh-CN | Chinese (Mandarin, Simplified) | zh-CN-YunyiMultilingualNeural (Male)

v1.6.5.2

  • uberduck voice selection fix (although some voices in the list no longer function)
  • now using latest verson of windows media controller, report any crashing and include event viewer details
  • {albumArtist} added for windows media and spotify media integration

v1.6.5

  • vrchat "/chatbox/typing" indicator added for whisper and deepgram (if VAD is active)
  • fix for whisper timestamps being offset for VAD so super short activations wouldn't be recognized
  • possible fix for group in word replacement tab breaking on some computers
  • hotkeys backend refactored, new hotkeys added for changing voice presets quickly
Assets 3
Loading

v1.6.6.1

29 Jan 23:00
@VRCWizard VRCWizard
v1.6.6.1
7e40d21
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.6.6.1 Pre-release
Pre-release

This is a demo of a different whisper library, it has not been thoroughly tested so there are likely bugs

v1.6.6.1 (current)

  • more robust output purification and removal
  • metrics on by default for this version
  • metrics changed to be more visible and distinguishable
  • fixed model download loop for models that were considered too small to be real

v1.6.6

  • Fixed memory leak in previous version (which caused crashing, oopsie)
  • Min VAD Duration (s) and Min Avg Confidence have been added to help fight against hallucinated outputs, you can enable debug text to see everything that gets filtered out

v1.6.5.9

Assets 3
Loading

v1.6.6 New Whisper Library

28 Jan 23:10
@VRCWizard VRCWizard
v1.6.6
77a9105
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.6.6 New Whisper Library Pre-release
Pre-release

This is a demo of a different whisper library, it has not been thoroughly tested so there are likely bugs

v1.6.6 (current)

  • Fixed memory leak in previous version (which caused crashing, oopsie)
  • Min VAD Duration (s) and Min Avg Confidence have been added to help fight against hallucinated outputs, you can enable debug text to see everything that gets filtered out

v1.6.5.9

Assets 3
Loading