-
Notifications
You must be signed in to change notification settings - Fork 0
Change Log
+ Add Acapella TTS Service
+ Add CereCloud Voice TTS Service
+ Fix speaking permission for subscribers
+ Load channel rewards when connecting to Twitch
+ Fix speaking of first emote (was saying the last)
+ Fix removal of emotes
+ Add support for SevenTV emotes
+ Fix BTTV emotes not being ignored if configured to
+ Fix removal of cheermotes
+ Fix minimum cheer amount not being used
+ Harden Twitch token validation
+ Add new command (!tts permissions) to enable/disable specific permissions
+ Update TTS service libraries (AWS, Google, Azure, Watson)
+ Improve shutdown time
+ Add option to be able to silence !tts command output
+ Update VIP/Mods when connecting to Twitch
+ Update cheermotes to be case insensitive (cheer5, Cheer5 would both be caught now)
+ Update networking code to support running on Linux
+ Misc fixes/tweaks+ Some fixes/changes
+ Fix the Twitch host event, there were a few issues causing crashes
+ Add %input% variable to Channel Reward messages
+ Fix voice fall back selection, was not working correctly
+ Misc fixes/improvements+ Enable and fix Twitch Host event
+ Misc fixes/tweaksWill re-organize these in a descending order
+ Fix not being able to connect after disconnect
+ Fix everyone permission
+ Few misc fixes/tweaks+ Will actually ignore a user now if flagged as ignored
+ Misc fixes/tweaks+ Fix an issue w/ gift sub event strings not saying the tier
+ Add ability to disable an event string, so you don't have to delete it to not use it
+ Fix removing an ignored voice (on TTS Settings tag)
+ Misc tweaks/fixes+ Not allowed response being to aggressive, gave it a stern talking to
+ Added !tts reg [add|del] [username] to add/remove regular status of user
+ Events Settings: Forgot to clear Enabled check box in the message area when picking another event type
+ Ignore exclimation was also being to aggressive, gave it a stern talking to as well+ Forgot to add the ability to force a user's speech, so, you can now force a user to always have there chat spoken, maybe tweaked a bit still (i.e. you could force a user, and disable all permissions, and their text would still be spoken)+ Bump SAPI5 sampling rate+ Fix crash Thelete reported
+ Update some default strings
+ Add subscription related info back in, this requires a new scope, so you'll need to re-auth, app will inform you
+ Misc fixes/tweaks+ New TTS Engine! IBM Watson, adds another 50 voices, with default install of windows and all engines active, you now have 376 voices
+ Separate ignore emote setting into the 3 services (Twitch, BTTV, FFZ), be default, BTTV and FFZ are not ignored, so be sure to toggle those to return to default behaviour
+ Misc fixes here and there
+ Update viewer display names when possible
+ Start tracking character counts, for now this is just put to a file, there will be a display area somewhere eventually
+ New Gag button to outright silence TTS
+ Other stuff I can't remember at the moment+ Deadlock resolved, I think
+ Watson is now auto enabled if configured to
+ AWS Keys are masked
+ Toys were put away for now
+ Implement some checks on settings files so they're not wiped
+ Error handling for engine inits, hopefully it won't crash now when aws/google/watson has issues on initial setup
+ Auto connecting option added
+ Other misc fixes/bugs+ Misc bug fixes+ Misc fixes
+ UI cleanup a bit, finally switched scaling in VS, so controls shouldn't overlap/etc anymore
+ Fix text input dialog focus
+ Ability to edit the nickname of a user directly, this is tested as best I can, it should be ok, its a delayed saving
+ Umm stuff I probably forgot?+ Small fix to the delayed text save+ Fix !tts assign last <user>, it should now actually work
+ Fix random sticky voice option, it should also now work
+ Add community goal contribution and finished events to events that can be spoken.Default event strings for contribution is:
Well, %name% decided to contribute %amount% towards %title%, it's %percent% finished, getting closer!
Default event string for goal finished is:
Amazing, %title% was just finished, %amount% points were used to fund it!
These will not be added to existing settings, so be sure to configure them if you would like to use them.
- Full on 4s, thought I'd give them back now.+ Audio output is now using WASAPI instead of WaveOut for, reasons.
+ Added new option to select the audio output device, this effects all speech; may expand this to be more granular, not sure yet. Option is on TTS Settings page
+ Some misc fixes/tweaks
- NOTE: Only WASAPI supported output devices will be listed.
- And yes, I know, the UI is still fugly, it's on my ever growing list.+ Harden some Twitch API calls to hopefully not cause as many crashes, this is an on-going thing
+ Add a check to prevent follow spam, during a session (open/close of program) a follow will only be fired once for a user and any further ones will be ignored
+ Misc fixes/tweaks+ Clicking on edit voice will bring speech preview in focus, and enable the next feature
+ Ability to assign a voice right from the speech preview window.
+ Add grace period (in seconds) for saying username
+ Fix community goal % contribution, I think?
+ Handle some google voice statup exceptions, this is still ongoing
+ Add option to say username only if the previous message was from a different user, can't be used with the grace period
+ Channel Rewards can now have random weighted messages, like events
+ Added new randomization code, will see how it works
+ Add option to swap name with nickname if one is set for a user
+ Misc cleanup and bug fixes
+ Add a UDP listener, will outline the commands later
+ Add ignored voice profiles
+ Add regex replacement, don't make me regret this.
+ Re-organize settings tabYou can now setup profiles for ignored voices/locales.
This adds a new moderator command !tts profile, !tts profile list, !tts profile activate
I know a lot have wanted this feature, and I've been very reluctant on adding it. Flat out, if I get wind of it being used in a purposefully hateful/hurtful way, I will pull all zips that have this version, and remove the code for it.
This replacement will occur for ## all spoken text (message, events, etc), in the chain of events, this is the last thing processed.
An example replacement:
Replace: [_\-!@#$%^&*()]+, With: a space, this will replace 1 or more, of any of the characters with a space, so a username that has _ in it, will get spaces instead.
Replace: \d{3,}, With nothing, this will remove all occurances of 3 or more numbers
Order of replacements is also something to keep in mind
+ Handle regex errors gracefully, and provide feedback when entering regex in replacement+ Fix audio device not being selected on startup (woops)
+ Toggling enabled caused a crash, fixed, but not sure why, need to investigate
+ Fix small issue w/ UDP+ Fix Twitch FUCK UP
+ Misc tweaks/fixes+ Misc fixes/cleanup
+ Twitch chat client is running on a new parser, small speed improvements
+ Better audio device enumeration (you'll see disconnected devices now to)
+ Added Voice Aliases
+ Addde ability for deleting a users message, timing a user out and banning a user to stop/skip speech (enabled by default)
+ Fix a bug where using stop would hang the application+ Squahsed a bug+ Add days left to the Community goal contribution event, use %daysLeft% to get this value
+ Add new !tts set nickname <username> <nickname> to allow a mod to set a users nickname, setting it to nothing will remove the nickname
+ Fix text replacement UI, ctrl+up/down wasn't behaving properly to change ordering
+ Minor fixes/tweaks
+ Updated twitch authorization mechanisms, it may re-ask for authorization, will also no longer complain constantly if there's an issue with your login and allow you to re-auth properly+ So, you can now adjust volume in the app, I've set a ceiling at 200% for now; I'm not sure how well this works overall yet, and you can also apply an override to the volume on a per voice basis.
+ Fixes/tweaks to underlying twitch connections, might mean less issues, time will tell?
+ Misc fixes/tweaks+ Fixed issue with buffers, so, it shouldn't repeat itself anymore+ Fix a potential crash with subs, seems twitch changed data (yay)
+ No longer listen to sub events via pubsub, was pointless, they're obtained by other means+ Fix issue w/ twitch library and cancelling raids
+ Fix issue when setting certain options via twitch chat
+ Add ability to use !tts commands via whisper, these will provide no feedback, and can be enabled/disabled via a setting
+ Added auto backup capability, on application start, it will auto zip up your config files into a time stamped zip file within a backups folder+ Added some more logging
+ Fixed audio device selection, guess I missed something with that
+ Updated HypeTrain end event for future possibilities+ Removed OneCore voices, this means there is no longer the hard requirement of windows, and may potentially work on Linux with Wine now
+ Added a new cloud service, Microsoft Cognitive Services, this can add upto another 220 voices.
+ Tweaked/fixed the cloud service voices use of pitch/rate/emphasis
+ Fixes/tweaks to underlying twitch codeI've altered the way the files are, so I would ## highly recommend extracing to a new folder and copying your settings over. (settings.json, voicealiases.dat, users.dat, usage.dat)
Rate can now be adjusted from -20 to 20, this equates to a 5% change in rate from baseline, pitch can now be adjusted from -9 to 11, this is a 5% relative change from baseline.
Pitch is ## NOT supported on Neural voices
Emphasis is supported on Standard voices only.
Rate can now be adjusted from -20 to 20, this equates to a 5% change in rate from baseline; Pitch can now be adjusted from -19 to 21, this is a 5% relative change from baseline.
Emphasis is supported on Standard and WaveNet voices.
Rate can now be adjusted from -20 to 20, this equates to a 5% change in rate from baseline; Pitch can now be adjusted from -19 to 21, this is a 5% relative change from baseline.
Emphasis is ## NOT supported on any voices.
Rate can now be adjusted from -10 to 10, this equates to a 5% change in rate from baseline; Pitch can now be adjusted from -9 to 11, this is a 5% relative change from baseline.
Emphasis is supported on Standard and Neural voices (might be hit or miss on neural voices).
+ Updated twitch libraries
+ Performance improvements
+ Update handling of emotes
+ When speaking everything, replies will be treated as normal messages
+ When speaking everything, allow different prefixes to not speak anything
+ Fix when saving audio files, if enabled
+ Add auto reconnect to twitch
+ !tts assign last <user> has returned
+ Misc fixes and tweaks/cleanup+ Fix duplicate emote issue, yay?
+ Guess it helps if the streamlabs token box is editable+ Fix ignore prefix, guess I overlooked something there+ Fix BTTV/FFZ emote handling
+ Fix Cheer event, seems I rushed this in the last version
+ Fix for emotes not being completely removed (yay typos?)
+ Add StreamElements support for Tips+ Make it a bit more friendly to those who aren't affiliate yet :)+ Some fixes for streamlabs.com socket handling+ Fixes for StreamElements
+ Added a new option to assign a voice; right click user, assign voice, dialog will popup where you can choose voice, assigning will auto-create alias and use it
+ Only 1 instance of TTS can run now
+ Small tweaks to UI
+ Misc fixes/tweaks
+ Other stuff I probably forgot+ Fix some issues with 3rd party emotes
+ Better handling of Twitch API instabilities
+ StreamElements fixes
+ Twitch Authorization now forces the Authorize button so you can switch accounts. Whenever you need to re-auth+ Add some error handling to UDP
+ Misc fixes/tweaks