j3sven / alt1-dialog-tts

Addon for the RuneScape Alt1 Toolkit that parses ingame dialog boxes to TTS

J3

Hi there!

I'm Sven but online I go by J3. I am primarily a web developer but eager to expand my horizons with any kind of dev work that piques my interest at any given time.

Fun facts about me:

• 🇳🇱 I'm dutch and I work as a frontend developer in the public sector.
• 💻 Developing since 2012
• ❤️ Open source

The best way to contact me is on discord, you can find me by my handle: @j3

alt1-dialog-tts's Issues

Research options for voice effects for supernatural creatures/gods

title

Fix imbalance in volume between various output sources

Make username mandatory

Make having a username set mandatory to prevent cache pollution

Implement system for overruling voice seeds with custom voices

Allow privileged users to overrule voice seeds with custom voices.

Things to consider:

• Create a priority system so that it can fallback on a generic voice in edge cases
• Must have sufficient labs credit to cover at least the gross dialog

Implement ability to store multiple usernames

Observed parsing errors

Collection of observed parsing errors

Make cache the default TTS service

Priority should be

1. Check cache for audio
2. Generate new audio if a key with credit is present
3. If no key is present and no audio is found in cache do backup if a backup is enabled.

Add exceptions for already voice-acted dialog

See: https://runescape.wiki/w/Voice_Over

Fix volume slider

Currently the volume slider does not save it's value in localstorage and resets on each session, it is also incompatible with the elevenlabs implementation.

Have a fallback voiceId solution for overridden characters

Custom voices are only available to the seed account, so for alternate voicelines not done in the custom voice there must be a fallback voiceId

Dialog not detected when chatheads are clipping up to the border

Examples:

• Nasthroth
• Wandering Ramokee

Loadtest api

Implement analytics

Fix I/l/1 bug

Most likely going to be depend on a core alt1 module update

Research automatic name detection

Not sure if possible with current OCR module

Store audio on postprocessing call rather than making a second round trip.