Red Carpet Regular Member++

"VOCALOID by Yamaha (Synthetic Singing)" , posted Sun 23 Nov 16:32

It's a new piece of software for recording studios, set for release in early 2004, that simplifies the process of recreating voices to have them synthetically sing. A person first has to record a series of clips where they pronounce different combinations of consonants and vowels. Once a library of vocal clips from that person is completed, it's inserted into the Vocaloid program, where users enter in the lyrics and notes plus inflections. From there, the program selects the voice clip and pitch to play at a certain moment to create the impression of singing.

Yahama press release on VOCALOID, March 4, 2003 (Samples in Yamaha Mid Radio Player format)
Vocaloid Official page (Samples in ASX format)

Kimi no Uwasa (MP3, 28 sec.) [1 min. 35 sec. clip found on Vocaloid Sample Page]
Amazing Grace (MP3, 49 sec.)

As the FAQ on the Vocaloid Official page notes, this technology is primarily for recreating the impression of singing -- trying to recreate natural speech with the vocal libraries is a more difficult process.

---

Though there seem to be several website mentions of this technology dated after Yamaha's press release, it was just recently covered in the New York Times article.

New York Times, Could I Get That Song in Elvis, Please?, Bill Werde, November 23, 2003

I do recommend reading the article (requires account, free and seemingly spamless), although it exaggerates the authentic quality of the voices: if you'll listen to the samples above, you hear that robotic-nasal tones and instant pitch jumps are still prevalent in the recreated voices. Still, it's impressive that Yamaha seems to have greatly simplified the process of recreating voices to sing -- I can imagine that, through MIDI software or MOD trackers, inputting and queueing each of the individual voice clips would be a long and arduous process.

---

If only we would've had something similar to this technology (along with the necessary space) for the Final Fantasy VI Opera Scene: how many of you were initially fooled when the voice sample matched the opera script's first sung vowel?

(... of course, maybe it would've been more time- and money-effective to have it recorded by actual singers than to to through this process...)

---

"I'll punish you with my burning fist of justice!!" | CCT: NEXT STAGE IS 15 |