How I got speech to text working on my arch machines

speech-dispatcher is good, and should be made available.

spd-say "Hello world" is the test.

but what to use for actually generating the speech?

piper is a great answer! it’s
the choice of Home Assistant, too.

Install speech-dispatcher: pacman -S speech-dispatcher
Install piper aurget piper-tts-bin
And some english voices: aurget piper-voices-common && aurget piper-voices-en-us

edit /etc/speech-dispatcher/speechd.conf as per the instructions:

 append to the end of the file:
 AddModule        "piper-generic" "sd_generic" "piper-generic.conf"
 DefaultModule    piper-generic
 DefaultVoiceType "MALE1"
 DefaultLanguage  "en"

see also /etc/speech-dispatcher/clients/piper_as_default.conf

make sure alsa-utils is installed, and aplay works

Test piper directly:

sample_text="Kubernetes is a container orchestrator that is way too complex, but solves real challenges that arise when you consider the requirements to achieve high-availability deployments. Is that complexity worth it? I think so, for many businesses want five nines of uptime which means they will need to consider these requirements."
echo "${sample_text}" | piper-tts --model "/usr/share/piper-voices/en/en_US/ryan/high/en_US-ryan-high.onnx" --output-file - | mpv --quiet -