computer generated voiceover's at DVinfo.net

Corey Williams · January 18th, 2008, 05:37 PM

Out of curiosity, are there any programs that allow you to do a voiceover without actually having a person to read it. It would be similar to http://www.vocaloid.com/en/features.html but without the singing feature.

Chris Harris · January 18th, 2008, 10:49 PM

AT&T has a site where you can try out a Text-to-Speech technology and download the .wav files, although you can only use it for non-commercial purposes according to the usage policy.

http://www.research.att.com/~ttsweb/tts/demo.php

Also, on the PC, there's this program called Natural Readers, although I've never used it, it seems you have to pay if you want it to export.

http://www.naturalreaders.com/

Another free Windows program is Sh**talker, which doesn't record to mp3 or anything, but you could use some other program to record the output.

Bill Davis · January 19th, 2008, 01:21 AM

Sure there are.

And they all sound like a computer reading copy.

The human voice is a pretty complex and variable thing.

You can say "How are you" about a thousand ways.

Happy, sad, scared, confident, sarcastic, silly, goofy, angry etc. etc. etc.

Without a brain to decide which is the best way in THIS situation, you're left with flat.

Apple's latest text reader (Voice Over - part of the Easy Access suite of automation tools) is about state of the art. Male, female, even a nice British accent. But the timing and inflection are still clearly computerized.

FWIW

Colin McDonald · January 19th, 2008, 02:50 AM

Quote:

Originally Posted by Bill Davis

Sure there are.

And they all sound like a computer reading copy.

The human voice is a pretty complex and variable thing.

You can say "How are you" about a thousand ways.

Happy, sad, scared, confident, sarcastic, silly, goofy, angry etc. etc. etc.

Without a brain to decide which is the best way in THIS situation, you're left with flat.

Apple's latest text reader (Voice Over - part of the Easy Access suite of automation tools) is about state of the art. Male, female, even a nice British accent. But the timing and inflection are still clearly computerized.

FWIW

Oh, I don't know ... it would go along very nicely with the mindless plastic "call on hold" type music that so many people seem to add to their videos.

:-)

Allen Plowman · January 19th, 2008, 03:17 AM

are we thinking Max Headroom remakes?

Jim Andrada · January 19th, 2008, 11:06 PM

Re computer generated voice---

A couple of years back I went to the big graphics show - SIGGRAPH - where George Lucas was the keynote speaker. One of the questions he answered related to what breakthroughs were needed in computer animation.

His answer was pretty interesting. he said that he didn't think any breakthroughs were required, just a lot of improvement on the tools that were already available. His rationale was that they could already deceive the eye pretty well.

Where the breakthrough was required, he said, was in the area of emotionally convincing voice generation.

Martin Catt · January 20th, 2008, 06:53 PM

I've been forced to sit through a couple of hours of online training videos with the computer-generated voice-overs, and quite frankly would like to choke the ever-lovin' crap out of whoever made the decision to NOT hire a real person to read. The droning quality and background sibilance could put a three-pot-a-day coffee drinker to sleep. About the only time I'd consider using a synthesized voice is when the script called for one.

Martin

January 18th, 2008, 05:37 PM	#1
Corey Williams Regular Crew Join Date: Jul 2007 Location: McKinney,TX/New Orleans, LA Posts: 104	computer generated voiceover's Out of curiosity, are there any programs that allow you to do a voiceover without actually having a person to read it. It would be similar to http://www.vocaloid.com/en/features.html but without the singing feature. Last edited by Corey Williams; January 18th, 2008 at 06:30 PM.

January 20th, 2008, 06:53 PM	#7
Martin Catt Major Player Join Date: Jun 2007 Location: The Colony TX Posts: 327	I've been forced to sit through a couple of hours of online training videos with the computer-generated voice-overs, and quite frankly would like to choke the ever-lovin' crap out of whoever made the decision to NOT hire a real person to read. The droning quality and background sibilance could put a three-pot-a-day coffee drinker to sleep. About the only time I'd consider using a synthesized voice is when the script called for one. Martin __________________ Canon XF300, Canon 5DMkII, Canon XL2, Rolls MX422 mixer, Zoom H4N, AT899 lavs, AT2020's, Azden SGM 1X shotgun, Manfrotto 501 head on 351 tripod

January 18th, 2008, 10:49 PM	#2
Chris Harris Major Player Join Date: Mar 2006 Location: San Francisco, California Posts: 487	AT&T has a site where you can try out a Text-to-Speech technology and download the .wav files, although you can only use it for non-commercial purposes according to the usage policy. http://www.research.att.com/~ttsweb/tts/demo.php Also, on the PC, there's this program called Natural Readers, although I've never used it, it seems you have to pay if you want it to export. http://www.naturalreaders.com/ Another free Windows program is Sh**talker, which doesn't record to mp3 or anything, but you could use some other program to record the output.

January 19th, 2008, 01:21 AM	#3
Bill Davis Trustee Join Date: Jan 2004 Location: Scottsdale, AZ 85260 Posts: 1,538	Sure there are. And they all sound like a computer reading copy. The human voice is a pretty complex and variable thing. You can say "How are you" about a thousand ways. Happy, sad, scared, confident, sarcastic, silly, goofy, angry etc. etc. etc. Without a brain to decide which is the best way in THIS situation, you're left with flat. Apple's latest text reader (Voice Over - part of the Easy Access suite of automation tools) is about state of the art. Male, female, even a nice British accent. But the timing and inflection are still clearly computerized. FWIW

January 19th, 2008, 03:17 AM	#5
Allen Plowman Major Player Join Date: Oct 2007 Location: Paradise, california Posts: 353	are we thinking Max Headroom remakes?

January 19th, 2008, 11:06 PM	#6
Jim Andrada Inner Circle Join Date: Feb 2007 Location: Tucson AZ Posts: 2,211	Re computer generated voice--- A couple of years back I went to the big graphics show - SIGGRAPH - where George Lucas was the keynote speaker. One of the questions he answered related to what breakthroughs were needed in computer animation. His answer was pretty interesting. he said that he didn't think any breakthroughs were required, just a lot of improvement on the tools that were already available. His rationale was that they could already deceive the eye pretty well. Where the breakthrough was required, he said, was in the area of emotionally convincing voice generation.