Skip to main content
Participating Frequently
December 26, 2011
Question

Text to Speech

  • December 26, 2011
  • 1 reply
  • 4811 views

I keep hearing wonderful presentation that proport to be created using text-to-speech as opposed to a live voice.  Mine all sound like a computer generated voice, which all my test groups hate.  What is the secret to creating text-to-speech that sound like a "real" person?  is it even posssible?  BTW, my preference would be to have a middle age female voice.

Any assistance would be greatly appreciated

This topic has been closed for replies.

1 reply

Anjaneai_Srivastava
Adobe Employee
Adobe Employee
December 27, 2011

Hello,

If you have problem with proper diction and pronounciation of certain words then a trick could be to spell the word the way you would like it to be pronounced . For Example -- "VOICE" can written as "VOYC" to improve pronounciation and also use DOT '.' for providing pauses appropriately. You would always have an option to record your voice and then you can decide the accent, emotions, prononciation and rate of speech. 

There are a total of 6 Voices (3 female) available for Text to Speech as part of Neo and Loquendo. If  you have not tested all, download the Captivate 5 voices from here --

http://kb2.adobe.com/cps/908/cpsid_90804.html

For Captivate 5.5 --

https://www.adobe.com/cfusion/tdrc/index.cfm?product=captivate

Sign-in with your Credentials -- Click on the OS next to TTS and it would download Voices.

See if it helps?

Thanks,

Anjaneai

rshindelAuthor
Participating Frequently
December 27, 2011

Thank you Anjaneai,

Can I infer from your comments that the speech I hear with natural cadence and inflections are really recorded voices, and not created with the text-to-speech application?

<mailto:rshindel@carosh.com> rshindel@carosh.com

c: 720-320-1618

p: 312-239-0529