Pluggable TTS in Wyoming

[lusty pine]

I have piper working find on a remote host using Wyoming. What I'd like is a Wyoming TTS wrapper that can be configured with an arbitrary program w/ arguments such that the "text to render" gets passed to the program over stdin and the program is expected to emit a wave file (or some common format) over stdout. This would allow me to trivially plug in new local TTS engines into HA. Has anyone done this or know how to do this?

[alpine furnace]

Wyoming is an open protocol. you can build a wrapper for other projects as you wish.

same as piper was developed into wyoming-piper

The code and information about the protocol is at this repo - https://github.com/OHF-Voice/wyoming