todo

[MysteryPancake]

• add mic input and output (with auto transcription using speech api) AND GREY TEXT FOR BITS OF SPEECH THAT ARE PROCESSING STILL
• Volume and pitch matching/humanizing
• Add reverse phonemes at some point
• Fix punctuation matching so it actually works again
• add wav and mp3 export and play options in the results
• add result speech editor - and options in this to disable diphones and triphones and change settings like that and add new stuff and delete stuff
• add auto-completion telling you whether or not the word will work
• add multiple input for combined transcripts
• get all broken website things to work
• add a way to use emphasis (CMU dictionary uses 0 = none, 1 = most, 2 = a little)
• convert numbers to words
• add Final Cut type XML export option maybe
• audio effects like normalization or silence removal
• add pitch modification somehow
• add crossfading
• add phone alternatives to avoid OOV
• add automatic click/pop removal
• add a way to set the audio transcription microphone
• add a way to set the recording microphone
• midi parsing support
• video playing support

After it works properly:

• maybe do something where you could have a twitch or youtube livestream where you type something and it says it in another voice and the voices rotate every now and then