[bookport] Re: Amazing Voice
- From: "Sarah Cranston" <cranston.sarah@xxxxxxxxxxxxxxxxx>
- To: <bookport@xxxxxxxxxxxxx>
- Date: Wed, 11 Jan 2006 09:53:06 -0600
The big tip off is the different inflection the same word. I can't think of a
good example right off the top of my head, but a speech synthesizer will say
the same word the same way reliably, whereas a human being always varies
his/her inflection, even if just a tiny bit.
-----Original Message-----
From: bookport-bounce@xxxxxxxxxxxxx
[mailto:bookport-bounce@xxxxxxxxxxxxx]On Behalf Of Walt Smith
Sent: Wednesday, January 11, 2006 9:47 AM
To: bookport@xxxxxxxxxxxxx
Subject: [bookport] Re: Amazing Voice
Right--no speech synthesis system on earth can get contextual pronunciation
correct 100% of the time.
----- Original Message -----
From: <bookport-bounce@xxxxxxxxxxxxx>
To: <bookport@xxxxxxxxxxxxx>
Sent: Wednesday, January 11, 2006 10:45 AM
Subject: [bookport] Re: Amazing Voice
No, I checked, and it's Kerry all right. There's way too much humanity in
her voice to confuse it with a synthesizer.
-----Original Message-----
From: bookport-bounce@xxxxxxxxxxxxx
[mailto:bookport-bounce@xxxxxxxxxxxxx]On Behalf Of Peter Torpey
Sent: Wednesday, January 11, 2006 9:42 AM
To: bookport@xxxxxxxxxxxxx
Subject: [bookport] Re: Amazing Voice
This definitely seems like synthetic speech. I would bet that it is AT&T
Naturally Speaking or something like it which uses a similar technology to
perform text-to-speech based on sampling from a human subject. I recently
heard voices with similar quality from the samples provided on the Freedom
Scientific web site for their Sarat reading system. Quite amazing - Too bad
this can't be done in a small unit like the Bookport with minimal processing
power. This type of speech reproduction probably requires a fair amount of
horsepower. Some day, we'll be there though!
-- Pete
Other related posts:
