h a l f b a k e r yQuis custodiet the custard?
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Please log in.
Before you can vote, you need to register.
Please log in or create an account.
|
This is an application that would first record/capture the
voice of a user. Perhaps the user would be instructed to
repeat a few phrases. The software would 'learn' the
user's normal tone, fluctuations, depth, accents, volume,
etc.
Next, once the user's voice has been captured and
understood, the voice can then be applied and adapted
to original songs and/or speeches. The application would
also understand the distinct qualities of an original
specified song or speech. As a result, a user may then
listen to him/herself sing a particular song or give a
certain speech.
For instance, I could hear myself sing Madonna's "Like a
Virgin" or an opera. Also, let's say I want to hear myself
give one of JFK's speeches. After learning my voice, the
software could then play JFK's speech, with the
appropriate fluctuations/volume/depth/speed, with my
voice.
This application could be described as 'automatic
karaoke' or 'voice syncing.'
Rob Donovan's Trainable Speech Synthesis
http://svr-www.eng....donovan_thesis.html
This sounded pretty realistic. Rob's working for IBM the last I heard, but haven't heard anything from Big Blue... [rmutt, Jan 08 2002, last modified Oct 04 2004]
[link]
|
| |
This would also be really fun hooked up to a vocorder or other voice synthesizer worn around the neck. Imagine giving yourself a Sean Connery or Kathleen Turner voice- or Bugs Bunny, Pee Wee Herman or Betty Boop, if you so desire. |
|
| |
On the downside, such a device could also be used to imitate a person's voice for nefarious purposes as well. Imagine if someone called in a few dozen bomb threats using your voice.... |
|
| |
PeeWee Herman imitating JFK and vice versa-now that I'm for. |
|
| |
Only if I can do [PeterSealy] saying "Baked!". |
|
| |
The current state of the art in text-to-speech can (it is claimed) copy intonation, pauses, volume and pitch fluctuation from pre-recorded speech. (So you can have an excited voice for some speeches and a calm one for others.) Systems tend to restrict basic quality of the speaking voice (probably so their TTS system has a recognisable voice) although some systems let you choose from a library of a dozen or more different speakers. |
|
| |
I don't think there's anything quite this advanced, available. You can however do it in a more painstaking way by editing audio files on a PC, but that requires manipulation of individual words or even lower-level work. I don't think an instantaneous translation is possible, just because people speak at different speeds and put pauses in their speech in different ways, |
|
| |
In season 2 of "24" they use a
technology like this as a plot
device. I believe that was after the
post. |
|
| |