this post was submitted on 12 Dec 2023
2 points (58.3% liked)

PC Gaming

8401 readers
615 users here now

For PC gaming news and discussion. PCGamingWiki

Rules:

  1. Be Respectful.
  2. No Spam or Porn.
  3. No Advertising.
  4. No Memes.
  5. No Tech Support.
  6. No questions about buying/building computers.
  7. No game suggestions, friend requests, surveys, or begging.
  8. No Let's Plays, streams, highlight reels/montages, random videos or shorts.
  9. No off-topic posts/comments.
  10. Use the original source, no clickbait titles, no duplicates. (Submissions should be from the original source if possible, unless from paywalled or non-english sources. If the title is clickbait or lacks context you may lightly edit the title.)

founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
[–] [email protected] 3 points 10 months ago (2 children)

...after being feedup with elevenlabs (popular TTS service) I choose to use a different route.

TTS is made with ttspeaker... and then added a bit of "flavor" by passing the output to a RVC model to give the newscaster a more poignant flavor. Let me know what you think about.

(also I am planning to change the newscaster model)

[–] [email protected] 2 points 10 months ago* (last edited 10 months ago) (1 children)

Have you looked into doing TTS on your own hardware? I recently managed to get this running (free !!).

It's pretty good imo, i found that the trick is to generate the voice with Bark and use "RVC Beta Demo" on top after. Coming from image generation RVC feels like a Lora.

Now i barely played with it, but i'm sure you can end up with some crazy level of details and customisation.

Have fun ^^

[–] [email protected] 2 points 10 months ago

The current problem with running these AI on local hardware is that, as supposedly tiny tools, they require huge packages to download and often require specific version of Python (3.8 while most modern Linux distro come with 3.10+) and most of the time you're required to make these massive download (+6GiB of libraries, pip packages and various dependencies/sdk)... just to give one single try. If you mess with something, it's all over again.

What I've found more useful, is using huggingface.I was forgetting about Bark! thanks for remind me... luckily is already available on hugface here