Ego sold seperately
November 24, 2003 3:54 AM   Subscribe

Computer generated singer, $200. Vocaloid software, which is due to be released to consumers in January, allows users to cast their own (or anyone else's) songs in a disembodied but exceedingly life-like concert-quality voice. Vocaloid will be able to "sing" whatever combination of notes and words a user feeds it. The first generation of the software will be available for $200. [NYTimes link]
posted by Outlawyr (23 comments total)
This is cool. Reminds me of the excellent Idoru by William Gibson.
posted by spazzm at 4:06 AM on November 24, 2003

Found some samples of vocaloid singing here on the Yamaha web site. But you have to have the Mid Radio Player installed--I can't read the language on the download page. Dern it.
posted by josephtate at 4:48 AM on November 24, 2003

there's another story and a sample in mp3 format here. I couldn't check it, though, since I have no sound card at work. Interesting, nevertheless
posted by swordfishtrombones at 5:04 AM on November 24, 2003

They probably just hire imitators to sing it.
posted by angry modem at 6:27 AM on November 24, 2003

And then on the seventh day, I read the rest of the article and said 'oops, that was a dumb comment.'
posted by angry modem at 6:29 AM on November 24, 2003

After listening to the "Amazing Grace" sample (mp3s here) "exceedingly life-like" sounds like a wee bit of an exaggeration.

Nonetheless, as a musician, I'd buy this product in a second. It would be great for prototyping. You could set up a quick synth-vocal track to aid in mixdown before the real vocal is recorded. Or use the synth vocal to tell the real vocalist how you want it sung.

Or hook it to a vocoder and make some weird sh**.
posted by mmoncur at 6:49 AM on November 24, 2003

Ahem. Mp3s here.
posted by mmoncur at 6:51 AM on November 24, 2003

Kinda reminds me more of Spinrad's Little Heroes (I think that was the name of it)
posted by geoffaw at 7:29 AM on November 24, 2003

fuck me. I want one. I really, really, really want one. Not for lead vox, but for backing vox - or to fill in for high notes that I can't reach. This is, after all, what the big name Britneys, Spice Girls et al have been doing for years on gazillion dollar machines.
posted by Pericles at 7:39 AM on November 24, 2003

Wow, that's cool. May not sound completely authentic, but pretty damned impressive nonetheless.
posted by PigAlien at 7:59 AM on November 24, 2003

I'm interested to see what new forms of music will come out of it - look at what people have already done with dictionary-oke and mixing the vocals/music for songs...
posted by whatzit at 8:15 AM on November 24, 2003

But how much is it?
posted by hama7 at 8:15 AM on November 24, 2003

yea, well I'm sure DJ Spooky or Shadow or Coldcut would know what to do with it but --- "realistic"? er, no.

But how much is it?

The player: $200.
The "fonts": ???
Authoring environment: not being released.

(disclosure: is my site)
posted by victors at 8:27 AM on November 24, 2003

Zdnet wrote a better article about this back in March. Sound Generator also delves a bit more into the techie side.
posted by tapeguy at 8:31 AM on November 24, 2003

Interesting, but I think it's only making the news because of Yamaha's big name. Singing voice synthesis is a very active field: see LYRICOS; Flinger; Burcas; and so on.
posted by raygirvan at 8:38 AM on November 24, 2003

Cher is OUT OF A JOB!
posted by brittney at 9:46 AM on November 24, 2003

See also VocalWriter.
posted by kindall at 11:41 AM on November 24, 2003

Direct Link to midradioplayer:
posted by milnak at 2:22 PM on November 24, 2003

That's damned spooky. The Japanese singing had a couple of clinkers when the voice shifted pitch, but the Amazing Grace sample - if I had heard it without knowing the context, I doubt I would have even considered it wasn't real.
posted by RylandDotNet at 3:40 PM on November 24, 2003

I'll wait for the Jeff Buckley module.
posted by yonderboy at 2:50 AM on November 25, 2003

Aren't MOST modern pop singers computer generated voices? Ever hear one of them in their unprocessed voice?
posted by HTuttle at 2:39 PM on November 25, 2003

Aren't you guys all computer generated? ..Or am I the only one?
posted by ZachsMind at 2:59 PM on November 25, 2003

Actually, I thought the Japanese was more likelike--the "amazing grace" sample had a lot of problems with consonants. There were some halting spots in the Japanese, but they seem to have nailed that better (which isn't terribly surprising).
posted by adamrice at 8:42 PM on November 25, 2003

« Older This is the cardinal sin of urinal etiquette:...   |   No, Ma, this is the best turkey ever! Really! Newer »

This thread has been archived and is closed to new comments