Speech synthesizer in 1K of Javascript


34 Responses to “Speech synthesizer in 1K of Javascript”

  1. Jose says:

    I would be nice if it worked. All it does on my computer is output random broken sounds ” ah uhm eh ah um otlotl”…

  2. jtropp1 says:

    It didn’t handle this nearly as well as I’d hoped:

    “Work It Make It Do It Makes Us Harder Better Faster Stronger”

  3. Gnatcatcher says:

    Great beat box synthesizer:

    puh puh puh ts ts ts puh puh puh ts ts ts puh puh puh ts ts ts puh puh puh ts ts ts

  4. xzzy says:

    You can sort of figure out what it’s saying if you already know what it’s going to say (upon launch, it begins spouting out the line of text. Click on it with the mouse and you get a type cursor, allowing you to delete everything and start over).

    In any other cases it fails miserably however.

    That said, doing it at all in under 1K is pretty impressive. 

    (and it is really good at doing drum beats)

  5. technogeekagain says:

    Not sure whether it barely fails or barely succeeds.  Cute attempt anyway.

  6. Adjam Oliver says:

    po1 is a fucking genius!

  7. Terranex says:

    oo oo oo ot oo ot oo ot oo os oo ot oo os oo ot oo os ooost ooost ooost

    There is potential in this.

  8. Paul Renault says:

    Wow, it’s the Beagle Brothers 256-byte Apple Basic program contests, for the 21st Century!

  9. I got it to say “I am the Gorfian Empire”

    • Jonathan Badger says:

      Yep. People are so spoiled now. They actually want synthesized speech that actually sounds comprehensible. In the 1980s we were just happy that an arcade game *tried* to speak.

      • Steve Taylor says:

        On one karate game we released for the Sinclair Spectrum (‘Way of the Exploding Fist’) there was no spare memory for sound effects, so we experimented with playing different sections of the machine’s ROM through the speaker, looking for stuff that sounded vaguely Bruce Lee-ish.

        But kids these days…

        • Glen Able says:

          The speccy also had the “Currah microspeech” hardware add-on for speech synthesis.  However, that used a full 2K of ROM, so this new technology is showing good progress.

      • Paul Renault says:

        Ah, yes, good ‘ole Castle Wolfenstein on an Apple ][.


        /It’s really important that you watch it in 720p, so that you get all the audio quality.

  10. AirPillo says:

    I was surprised to be able to get a decent rythm going by just repeating “nyan gao gao gao”

  11. robbersdog says:

    m s m s m s m ssm s m s m s m ssm ssm ssm ssm ssm m m m m m m m a i a i a i a aia aim aia aim ai ai o i a i o i aoaoaoaoaoaoaoaoa e m s m e m s m e m e m e m s m e ee m e ee s m e ee m ee m

    This is far too much fun!

  12. Kernit the Font says:

    Lol, great minds think alike. Here’s a bit of acid: (P.S. If you hit return twice, you get a cool delay effect)

    tikitikitikitikitikitikitikitiki.Bm. tsh… bm s. Bm. tsh… bm s. Bm. tsh… bm s. Bm. tsh… bm s. Bm. tikka tikka Bm. tikka tikka Bm. tikka tikka Bm. tikka tikka Bm. tikka rikee Bm. tikka rikee Bm. tikka rikee Bm. tikka rikee. oooowow dada h. oooowow dada h. oooowow dada h. oooowow dada h. dooweedoowee o. dooweedoowee a. dooweedoowee e. dooweedoowee u. dooweedoowee o. dooweedoowee a. dooweedoowee e. dooweedoowee u.

  13. awjt says:

    I wrote a song.


  14. SamSam says:

    a b c d e f g h i j k l m n o p q r s t u v w x y z

    sounds pretty sweet

  15. jjatria says:

    Hah. Some of these are a lot of fun.

    uhkgas  u   es  uhkgas  u   es  uhkgas  u   es  uhkgas  u sses  uhkgas  u sses  uhkgass u sses  uhkgass aasses  uhkgass aisses  uhkgass aasses  uhkgass aisses puhkgassiiasses puhkgassiiasses p hkg ss   ss s p hkg ss   ss s 

    Or this one

    sss gssss gssss gss ss sss gssss gssss gssss gss ss sss gssss gssss gssss gss ss sss gssss

  16. coop says:

    I bet Adobe could achieve the same effect with 1 GB of code.


  17. Robert Holmen says:

    E I E I O

  18. Mathieu Henri says:

    ^___^ I didn’t expect people to have fun with my entry for JS1k, let alone write beats with it. Well done guys!

    PS: I know the quality of the speech is meh but Hey! It’s 1k of JS. I’ll try to optimize this puppy a little and see if I can squeeze a 3rd formant filter to the mix.

    • xzzy says:

      Give it a “save to file” feature so we can export our beats into a track editor. :D

    • awjt says:

       Do you have a tutorial or a link somewhere to formant filters?  I would like to learn some of this programming.  I followed the link to the original tiny synth, but that site is like written in Klingon.  I’m looking for something that a mere human can understand, if you have any refs.  THANKS!!!  And nice work!!!

  19. Mark Dow says:

    iem sari dav  i kant do dat

  20. hadlockk says:

    Seems to have a thick Indian accent.

  21. ppdd says:

    Decidedly less irritating than Siri.

  22. Pietari Kokko says:

    domo arigaa.tomister robo to

  23. Jason Aeschliman says:


Leave a Reply