Yamaha PLG 100 SG Plugin Card a FS1R in a nutshell

[javelin276]

Hi there, Soundinett.
I tested out the ChristinaAguilera1 voice patch on a George Gershwin piece, 'Summertime'. I haven't had time to try out scripting lyrics yet, I just used an 'oooh' vocal sound to sing the tune with. I'm afraid my pianist needs to sharpen up his playing a bit, sorry about that.
I haven't used any dynamics, voice controllers, or anything fancy here, I just adjusted the melody notes a bit and let the SG card sing it.
I think it sounds pretty good for a first try.

You can download the mp3 test off my webpage.
Thor's PLG100-SG Web Page

[soundinett]

[javelin276]

Well, I can't get the card to sing in English very well at all, it sounds like chop suey. Even when it sings in Japanese it still isn't all that convincing as a singer. Maybe a different approach is needed, because I can get a really nice Aguilera 'ooh' out of it. Maybe a better way to utilize the card would be to have it sing simpler oohs and aahs rather than trying to speak. Maybe something more akin to background singing is more appropriate. I 'll try that next.

[joey]

Pretty sure this is done with vocaloid

[video][/video]

[javelin276]

That's a creepy video...
I agree with you, I'm pretty sure it's Vocaloid also. The phoneme transitions are smoother than I can get with the PLG100-SG card. Only one phrase, though. I'll bet I could get somewhat close to that, but would you want to? Yikes.

Vocaloid can sound really good in the right hands.
<Vocaloid Song as an Example>
Vocaloid seems to be limited to use in Anime, though. The voices they've created for it are mostly higher pitched adolescent voices, you don't have the choice of a mature Jazz singer, or any deeper male voices. It's done by splicing and blending recordings together. The SG card uses formants to create the tone from scratch.

I'm going for something more musical with the SG card, which is why I'm going to try a backup singer approach for my next test.

[javelin276]

I've ported one of the Christina PLG100-SG patches I created over to the FS1R for comparison. The FS1R patch sounds better, more mellow, less mechanical, more human-like over a wider range of notes. I'm in the process of setting up controllers so I can morph the vowel sound from one vowel to another on the FS1R.

I uploaded the song onto my webpage, you can listen to the Christina Aguilera voice patch sing here:
<Misty, sung by the FS1R>

[funkel]

Anyone know where to find the SG Lyric Editor? Seems as though it is not on the Yamaha site anymore.