O'Reilly recently snuck a wacky speech synthesizer into our blogs: Clicking the "listen" link above will play back these words with a robotic voice. As a speech synth enthusiast, I immediately started looking for phrases that would produce funny rhythms. I found the first in Peter Drescher's recent blog about the Game Developers Conference:
I love the Game Developer's Conference! the lights, the cameras, the action, all the best and brightest coming together for meets and greets and foods and drinks, it's exciting, exhilirating, and completely exhausting!
(Robot photo by AZAdam.)
I captured the synth's output with Ambrosia Wiretap, imported it into Ableton Live, and cooked up the following ditty:
- Meets and Greets Groove (720KB MP3)
The stuttering effect halfway through is from Ableton's excellent Beat Repeat processor, which applies random gating, echoing, and pitch changes to the audio. Other sounds were detritus from my hard drive, such as a Novation K-Station arpeggio, the looping output of a crashed soft synth, a binaural recording I made with a pocket voice recorder, and assorted drum loops.
One of the earliest speech synthesizers, Joseph Faber's Euphonia (1846), reportedly sang "God Save the Queen."
For more speech-synth entertainment, see my article "Sneaky Tricks for Speech Synthesizers" and podcast "Singing Computers."
Yo.
Categories
Audio
Comments (2)
Read More Entries by David Battino.
PDX: Supposedly our speech synth is trainable, so perhaps I'll give it another shot at that text in a few months. And agreed: It's the imperfections and surprises that make speech synths interesting.
when posting blogs, you hope that readers will find your words interesting, entertaining, informative, possibly enlightening, maybe even inspiring ... but used as lyrics for speech-synthesized vocals? that's a first!
it does seems like a useful sound design tool for creating for curious audio, or injecting literal meaning into music without recording a singer. on the other hand, the o'reilly speech synth does makes some, uh, interesting decisions when generating pauses, pitch, and pronunciation; maybe that's part of what makes it enjoyable to listen to.
hmm, wonder what it sounds like reading shakespeare ...