Speech Synthesis Limitations

0 favourites
  • 7 posts
  • I was going to put speech synthesis in my game as a joke for the "voice acting" option because it can turn out hilarious. In Chrome I had access to 20+ voices and I assumed NWjs would have them, but it only has the two that Windows comes with and ones I specifically installed. I'm not too surprised because NWjs is a stripped down version of Chromium, it just was a little disappointing to find out speech synthesis doesn't have as much functionality as I thought it would. I would go through the trouble of adding voices to my game, but I don't think this is possible.

  • I believe Chrome speech synthesis is network-based, there is a noticeable delay while it downloads the audio and it won't work if you are offline.

    NWJS uses the default speech synthesis service built in the OS and it works offline.

  • Try Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Try Now Construct 3 users don't see these ads
  • I believe Chrome speech synthesis is network-based, there is a noticeable delay while it downloads the audio and it won't work if you are offline.

    NWJS uses the default speech synthesis service built in the OS and it works offline.

    I guess that explains why everything I made it say took 2 or 3 seconds to start haha. I was OK with that, but yeah. I just wish speech synthesis could be more of a per-app thing like if I include voices with my game then the OS would be able to use them without the user having to bother installing them all so they get the same voices I intended.

    Ultimately not a big deal because the option was just to make the game silly on purpose.

  • TTS is all over the board - device, OS, user-determines OS settings, browser, user-determined browser settings, etc. It's very tough to implement with predictability. Least common denominator is about all you can count on. I work alot on content that is both English and Hindi TTS. It's a bit of a nightmare.

  • I work alot on content that is both English and Hindi TTS. It's a bit of a nightmare.

    Sounds like a thing that would make me quickly ragequit with how I can't even do a silly thing with it. lol

  • > I work alot on content that is both English and Hindi TTS. It's a bit of a nightmare.

    Sounds like a thing that would make me quickly ragequit with how I can't even do a silly thing with it. lol

    The Golden Ticket for TTS is to pay Google/Amazon/MS cloud services a fairly large amount of money, then you get more cross-platform predictability and higher quality and more varied voices. Implementation is still not easy though, it takes a skilled developer. Construct makes TTS and speech recognition incredibly easy by comparison to any other method I've seen.

  • I'm sure it does. I've seen a plugin that does TTS without Google/Amazon/MS stuff, though its ability to produce intelligible speech was really lacking even if I only used it as a joke since iirc it isn't free.

Jump to:
Active Users
There are 1 visitors browsing this topic (0 users and 1 guests)