v1.5.8.1 (current)

• button to manually output media text
• other minor bug fixes and changes for media integration

v1.5.8

• added style exaggeration slider, speaker boost toggle and v2 model for eleven labs
• you should no longer have to click the spotify connect button every couple days
  • no longer spam refreshes token, it will refresh once per hour
• fixed osc listener starting twice on launch with setting enabled

• added style exaggeration slider, speaker boost toggle and v2 model for eleven labs
• you should no longer have to click the spotify connect button every couple days
  • no longer spam refreshes token, it will refresh once per hour
• fixed osc listener starting twice on launch with setting enabled

v1.5.7.5 (current)

• Azure voice list updated, you will notice a couple new voices
  • More Arabic and Spanish speaking voices added
• Uberduck voice list updated

v1.5.7

• Websocket server for receiving TTS Messages (in heartrate and osc listener tab for now)
• Trying a different method for opening Spotify Connect URl in browser (same method used for getting pulsoid token)
  • Open to display connect URL in log if method to open url in browser doesn't work
• Removed optimus prime voice from tiktok voice list
• Option to turn off same language lock (in azure settings)
  • Primary use is for Azure speech to text since you get 5 speech hours and 5 speech translation hours for free each month effectively giving you 10 hours of speech to text
• Colons and semicolons can now be saved in media presets

• websocket server for receiving TTS Messages (in heartrate and osc listener tab for now)
• trying a different method for opening Spotify Connect URl in browser (same method used for getting pulsoid token)
  -open to display connect URL in log if method to open url in browser doesn't work
• removed optimus prime voice from tiktok list
• option to turn off same language lock (in azure settings)
  • primary use is for Azure speech to text since you get 5 speech hours and 5 speech translation hours for free each month effectively giving you 10 hours of speech to text
• colons and semicolons can now be saved in media presets

• Fix for OSC Listener Activate on Start setting not working

• Pulsoid Reconnect Logic Added
  • In the case of the Websocket disconnecting for any reason the reconnection logic with continually try to re-establish a connection

• fix for azure translation not changing button color to making sound on press
• fix for being unable to translate to english with deepL
• fix for azure translation using incorrect language codes
• fix for combobox population when using uberduck

v1.5.6.1 (current)

• Heartrate OSC messages now all output on specified update interval
• minor UI changes

v1.5.6

• Direct heartrate device support for free via Pulsoid
  • Learn how to connect a Pulsoid supported device: https://github.com/VRCWizard/TTS-Voice-Wizard/wiki/Heartrate-Integration
• Taiwanese supported for azure speech to text (3 voices added)
• option to save audio files with unique names (instead of overriding one file each time)
• button added to take you to your audio files folder
• Fix for Configuration Failed to Initialize Error
  • A backup is now stored of your settings file on a successful launch
  • The backup is used in the case of your settings being corrupted
• additional small changes and fixes

v1.5.5

• over 40 new languages added: https://github.com/VRCWizard/TTS-Voice-Wizard/wiki/Supported-Languages
• more azure voices (a couple for each new language)
• switching spoken language auto restarts whisper for you with the new language (this may be made optional in the future)
• Glados TTS Method "rebranded" to locally hosted
• fix for osc send address not saving correctly on close and reload
• fix for proper error message not appearing for configuration failed to initialize

v1.5.6 (current)

• Direct heartrate device support for free via Pulsoid
  • Learn how to connect a Pulsoid supported device here
• Taiwanese supported for azure speech to text (3 voices added)
• option to save audio files with unique names (instead of overriding one file each time)
• button added to take you to your audio files folder
• Fix for Configuration Failed to Initialize Error
  • A backup is now stored of your settings file on a successful launch
  • The backup is used in the case of your settings being corrupted
• additional small changes and fixes

v1.5.5

• over 40 new languages added: https://github.com/VRCWizard/TTS-Voice-Wizard/wiki/Supported-Languages
• more azure voices (a couple for each new language)
• switching spoken language auto restarts whisper for you with the new language (this may be made optional in the future)
• Glados TTS Method "rebranded" to locally hosted
• fix for osc send address not saving correctly on close and reload
• fix for proper error message not appearing for configuration failed to initialize

• switching spoken language auto restarts whisper for you with the new language (this may be made optional in the future)
• Glados TTS Method "rebranded" to locally hosted
• more languages (Documentation WIP)
• more azure voices (a couple for each new language)
• fix for osc send address not saving correctly on close and reload
• fix for proper error message not appearing for configuration failed to initialize

• many path locations changed from relative to absolute to prevent a small portion of users apps not selecting the correct working directory

• fix for Auto Replay KAY break settings on reload of app
• fixed bug with advanced media settings
• relative paths for button sounds made absolute
• typo fixes

• Error Loading Settings popup gives additional info and closes app for the user
• fix for whisper getting stuck listening if user switched stt method while it was starting up
• fix of DoSpeechToText parameter not working

Notes

• Pro Users should update to this version if they want their pitch and rate setting to be accurate when using the Voice Wizard Pro API.

Changelog

v1.5.1.7 (current)

• Imports and Exports added for Emojis Tab and Azure Dictionary
• Fixes for Imports and Exports not working as intended
• additional small changes / fixes

v1.5.1

• Media Presets now editable similarly to Voice Presets
  • Media presets will not transfer over to this version from a previous version
• Voice Presets, Media Presets, Word Replacement Lists and Voice Command Lists can now be imported and exported with ease
• API Branch Selection for Voice Wizard Pro added
  • Currently there are only eastus and dev as branches
• Pitch and Speed Sliders are now on a scale of -100 to 100
  • Pro users on this version will have to switch their branch to dev for this feature to work correctly until the eastus branch is updated
• Mute button has been added for Whisper due to instability when starting and stopping
• Whisper now can now not be stopped during the start up process to help prevent crashing :)

v1.5.0

• auto-breaths, soft and whispered styles(effects) have been added for all standard Amazon Polly voices
• Over 15 new Azure English speaking voices available

v1.4.9

• Over 30 new voices from Tiktok
• Tiktok voices now load from json (if new voices are discovered you can manually add them here)
• Fixes for button color changing
• Current voice preset is saved and loaded (overrides other voice settings)
• App now stops recognizers before closing
  • Having a recognizer running like whisper and closing the app would cause noticeable errors in visual studio
• Added a debug overlay for whisper current transcription status at the bottom of log
• hide delay 10000 ms by default

v1.4.8.1

• Persian language added
  • Added for Whisper Recognition
  • Added for Azure Recognition
  • Persian voices added for Azure
  • Pro translation
• Fixed bug that caused application to crash when attempting to transcribe certain languages
  • Whisper Package modified
• added correct language code for Norwegian to whisper recognition

v1.4.6

• new play tts sound and speech to text start/stop sounds
• speech to text button changes color when listening
• Whisper will no longer stop if pressed too quickly
• fixed quotes appear around all text when turning off Whisper filter
• re-added config redirect for configuration failed to initialize
• fixed issues with OSC ports for sending and receiving being incorrect on launch from v1.4.5

v1.4.5

• Edited uberduckVoices.json file to match the allowed voices as of 7-12-2023
  • Number of voices heavily reduced :(
• Smart String Splitting, splits strings if you say a sentence longer than 295 characters, finds the closest space to split at.
  • You can edit the limit to be 144 characters for example so all messages fit in the VRChat chat box
• Upon a configuration error you should be redirected to the config location to either delete or remove and add back the config file (still can't figure out why updating breaks the config file like this)
• Webcaptioner and uberduck now show api response codes on error
  fixed bug where uberduck tts was still trying to pull voices from old file location (would only work if you updated from an older version of the app or moved the voices folder)
• port and address for OSC are now pulled from settings instead of always starting as the default

• Media Presets now editable similarly to Voice Presets
  • Media presets will not transfer over to this version from a previous version
• Voice Presets, Media Presets, Word Replacement Lists and Voice Command Lists can now be imported and exported with ease
• API Branch Selection for Voice Wizard Pro added
  • Currently there are only eastus and dev as branches
• Pitch and Speed Sliders are now on a scale of -100 to 100
  • Pro users on this version will have to switch their branch to dev for this feature to work correctly until the eastus branch is updated
• Mute button has been added for Whisper due to instability when starting and stopping
• Whisper now can now not be stopped during the start up process to help prevent crashing :)

known bugs

• exports and imports may not work as expected

• auto-breaths, soft and whispered styles(effects) have been added for all standard Amazon Polly voices
• Over 15 new Azure English speaking voices available

• Over 30 new voices from Tiktok
• Tiktok voices now load from json (if new voices are discovered you can manually add them here)
• Fixes for button color changing
• Current voice preset is saved and loaded (overrides other voice settings)
• App now stops recognizers before closing
  • Having a recognizer running like whisper and closing the app would cause noticeable errors in visual studio
• Possible dirty fix for access violation exception (edited comlight dll)
  • This is what caused the app to "randomly" crash sometimes when starting and stopping whisper
• Added a debug overlay for whisper current transcription status at the bottom of log
• hide delay 10000 ms by default

• Persian language added
  • Added for Whisper Recognition
  • Added for Azure Recognition
  • Persian voices added for Azure
  • Pro translation
• Fixed bug that caused application to crash when attempting to transcribe certain languages
  • Whisper Package modified
• added correct language code for Norwegian to whisper recognition
• Possible fix for general whisper crashing upon starting / stopping recognition

ChangeLog

v1.4.6 (current)

• new play tts sound and speech to text start/stop sounds
• speech to text button changes color when listening
• Whisper will no longer stop if pressed too quickly
• fixed quotes appear around all text when turning off Whisper filter
• re-added config redirect for configuration failed to initialize
• fixed issues with OSC ports for sending and receiving being incorrect on launch from v1.4.5

v1.4.5

• Edited uberduckVoices.json file to match the allowed voices as of 7-12-2023
  • Number of voices heavily reduced :(
• Smart String Splitting, splits strings if you say a sentence longer than 295 characters, finds the closest space to split at.
  • You can edit the limit to be 144 characters for example so all messages fit in the VRChat chat box
• Upon a configuration error you should be redirected to the config location to either delete or remove and add back the config file (still can't figure out why updating breaks the config file like this)
• Webcaptioner and uberduck now show api response codes on error
  fixed bug where uberduck tts was still trying to pull voices from old file location (would only work if you updated from an older version of the app or moved the voices folder)
• port and address for OSC are now pulled from settings instead of always starting as the default

• Edited uberduckVoices.json file to match the allowed voices as of 7-12-2023
  • Number of voices heavily reduced :(
• Smart String Splitting, splits strings if you say a sentence longer than 295 characters, finds the closest space to split at.
  • You can edit the limit to be 144 characters for example so all messages fit in the VRChat chat box
• Whisper should no longer crash at all, please report any crashing with what you were doing before it happened.
• Upon a configuration error you should be redirected to the config location to either delete or remove and add back the config file (still can't figure out why updating breaks the config file like this)
• Webcaptioner and uberduck now show api response codes on error
• fixed bug where uberduck tts was still trying to pull voices from old file location (would only work if you updated from an older version of the app or moved the voices folder)
• port and address for OSC are now pulled from settings instead of always starting as the default
• (Warning, this update contains issues with sending and receiving OSC)