This is what I've been waiting for. This is why I built the translate and text to speech views in noStrudel.
DVMs are getting better and I don't have to have to do any work. The interoperability nostr has is awesome
Discussion
Yes. Hard to believe.. But it works.. Let me know for other DVMs you see fit.. We will keep building..!
nostr:npub1mz3vx0ew9le6n48l9f2e8u745k0fzel6thksv0gwfxy3wanprcxq79mymx voice quality is awesome. Will test this on some more posts.
nostr:npub1ye5ptcxfyyxl5vjvdjar2ua3f0hynkjzpx552mu5snj3qmx5pzjscpknpr I had selected "English" in the dropdown (Default was German from my Browser settings) but the DVM Reuqest showed "Deutsch (German)":
The voice output was still in perfect english. Not sure if it is an actual issue - just confusing.
Also: It might be good to still show the full posts if the user wants to read while listening or look at images. Currently you only see the audio player.
Thanks for testing. I am using Elevenlabs text2speech API which is best among all. I got a lot of free credits.. so hit it hard and let me know for issues. On Text2Speech there is a To-Do on my end to translate before speech if say text is in english and you want German voice. Will push that by tomorrow.
It also might be good to write a few regexp to remove links and nostr notes or replace nouns with usernames when reading. I noticed when I tried it it was trying to read the links and npubs 😀
Once the DVM has translated it once, everybody inherits that translation automatically without having to request it again?
Yup, that’s why the input event id is indexed
nostr:note1j2l0fqwr5r9pfeshysljmtt7xj7f4uzwmy863rktn6km4wkcf8zqfstqq5