Just installed VOCA (npub doesn't resolve; you need to set a handle) via nostr:npub10r8xl2njyepcw2zwv3a6dyufj4e4ajx86hz6v4ehu4gnpupxxp7stjt2p8. Looks promising, but doesn't work for me on nostr:npub1235tem4hfn34edqh8hxfja9amty73998f0eagnuu4zm423s9e8ksdg0ht5 — play does nothing. I'm assuming it's a TTS service problem.
Has anyone gotten it working?
nostr:nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqclk6vc9xhjp8q5cws262wuf2eh4zuvwupft03hy4ttqqnm7e0jrqdasl9j nostr:nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpq0r8xl2njyepcw2zwv3a6dyufj4e4ajx86hz6v4ehu4gnpupxxp7shn6f9s For TTS on GrapheneOS, you need to install a TTS app. There are a bunch of both open source and proprietary options. Some are based on old school hard-wired text-to-speech data while others use LLMs. eSpeak NG is an example of a more old school one with basic functionality. If you want something better there are options. LLM ones improve quality but hurt latency.
It works but the pauses between sentences is from 3-15 seconds depending on the length of the following sentence (at least on my pixel 8 pro running Graphene)
Wish it was to the point of being able to let it play like a podcast but it’s not quite there yet
Get Sherpa on zapstore and download the en_GB-southern_english_female-medium model.
Currently use Sherpa. Previously used RHvoice.
