h a l f b a k e r y
The Out-of-Focus Group.
idea:
add, search, annotate, link, view, overview, recent, by name, best, random
meta:
news, help, about, links, report a problem
account:
Browse anonymously,
or get an account
and write.
Login
Create account.
|
|
|
I think that theres a possibility of this working. If you record a voice, any voice say some random guy on the road saying "get away from me you crazy loon" and then transfer that data onto a computer software program and copy the pitch, tone, volume etc and then record that data onto an input/output
device and then whenever you speak through that device, you sound like the random guy off the street!
AT&T Natural Voices
http://www.research.../NaturalVoices.html
Baked 2 years ago by AT&T, quoting: "Additionally, the AT&T Labs Natural Voices Icons is a fully customizable TTS voice product that lets customers develop a custom corporate voice, perhaps duplicating the voice of a celebrity or signature spokesperson to extend corporate image and branding. The customer selects a voice talent to generate live recordings. The AT&T Labs TTS team uses these recordings to create a one-of-a-kind text-to-speech voice that gives the customer a distinctive sound for its voice-enabled applications." [krelnik, Oct 05 2004]
[link]
|
| |
Nb hope this hasnt already been baked. |
|
| |
Sorry, been done. Actually, no, wait. Not sure about that. |
|
| |
Why would you want one of these? |
|
| |
There's certainly pitch harmonizers and voice disguiser products on the market. Both for telephony and online gaming (X-Box Online). I doubt that there are systems for transmogrifying one person's voice into an exact replication of another person's voice, though; at least not directly through sound processing. |
|
| |
One possible approach, though, is to use voice recognition to capture the recording voice as speech, rather than sound, data and then use that captured data to drive speech synthesis to recreate the voice as another person's voice. AT&T's speech synthesis work has already been demonstrated emulating celebrities voices with a surprising amount of accuracy. |
|
| |
Except in "Diamonds are Forever" and "Face/Off". |
|
| |
I have a program that can exactly replicate prof stephen hawking's voice... |
|
| |
I doubt very much this is possible. |
|
| |
Doubt away, but this is in fact baked by AT&T. They need more than just a short clip of the person to achieve the replication, however. See link. Combine this technology with a speech-to-text on the input side, and you have this idea exactly. |
|
| |