KnowBrainer Speech Recognition
Decrease font size
Increase font size
Topic Title: Text to speech
Topic Summary: Different versions of Dragon?
Created On: 11/08/2007 10:33 PM
Status: Post and Reply
Linear : Threading : Single : Branch
 Text to speech   - boblou - 11/08/2007 10:33 PM  
 Text to speech   - Lunis Orcutt - 11/09/2007 02:23 AM  
 Text to speech   - boblou - 11/09/2007 05:52 AM  
 Text to speech   - Lunis Orcutt - 11/09/2007 03:03 PM  
 Text to speech   - Chucker - 11/09/2007 08:30 PM  
 Text to speech   - boblou - 11/10/2007 05:10 AM  
 Text to speech   - fl - 11/10/2007 08:07 AM  
 Text to speech   - Chucker - 11/10/2007 09:19 AM  
Keyword
 11/08/2007 10:33 PM
User is offline View Users Profile Print this message


boblou
Junior Member

Posts: 4
Joined: 11/07/2007

My brother-in-law and I have installed DNS 9.5 Preferred on our computers.  In the Text to Speech area my version provides 5 voices and allows me to alter the pitch.  His version has only to voices (Jennifer and Jane) and the pitch cannot be accessed.  We have tried the update command on his computer and we get the Error 13003 message.  How do we get the pitch and other voices?
 11/09/2007 02:23 AM
User is offline View Users Profile Print this message

Author Icon
Lunis Orcutt
Top-Tier Member

Posts: 25091
Joined: 10/01/2006

The English versions of NaturallySpeaking only have 2 text-to-speech voices; Jennifer (US Midwest) and Jane (Queens English).  The only thing we can think of is that your additional voices are either coming from a third-party download that you forgot about installing or you installed a non-US English version of NaturallySpeaking.  The NaturallySpeaking options include a pitch control but you cannot use it on the default voices because the Nuance developers locked these 2 voices out but as you pointed out, they can work on other voices.  We have experimented with Microsoft's Sam, Mike and Mary voices in the past and we believe that AT&T offers some additional third-party voices but for our purposes, we find that Jennifer does a pretty good job.
 
You cannot update DNS 9.5 because there is no update available.


-------------------------


Click KB 2012 REV D to Download a 30 Day Evaluation of KnowBrainer 2012
Click SpeechStart+ to Download a 15 Day Evaluation of SpeechStart+

 11/09/2007 05:52 AM
User is offline View Users Profile Print this message


boblou
Junior Member

Posts: 4
Joined: 11/07/2007

Thanks for the quick response.  The only speech recognition software I have on my computer is the two disks I installed that I received from the online store.  As far as I know, I ordered the U.S. English version.  The three extra voices are Adult Male #8 American English (TruVoice), Adult Female #1 American English (TruVoice) and Adult Female #2 American English (TruVoice).  They are not clear and seem to have a lisp or "s" sound after each word.  Changing the pitch does not resolve this sound, thus they are of no value.  I have noticed that I can improve Jennifer by changing the speed and using a different headset.  I am presently searching for headsets which would result in greater clarity.  Any suggestions? 
 11/09/2007 03:03 PM
User is offline View Users Profile Print this message

Author Icon
Lunis Orcutt
Top-Tier Member

Posts: 25091
Joined: 10/01/2006

We don't believe the TruVoice text-to-speech readers are part of NaturallySpeaking.  We believe they are third-party products or possibly something left over from an L & H speech recognition product. 
 
As far as switching headsets is concerned, have you considered utilizing the text-to-speech engine through your computer's external speakers?  We're only asking because this option would be absolutely free and KnowBrainer 2006 makes ample use of this feature.  We love text-to-speech playback but found it to be somewhat annoying when playing it through our headset speakers.  Now that we have given you our free recommendation will move on to headset speaker options…
 
The best wired headset we can recommend is theBoom Quiet microphone which includes stereo high fidelity speakers.  The headset is a little heavy so if you're not planning on using it for music playback it might not be your best choice.  The ANC 750 microphone would fit the bill, includes low fidelity stereo speakers and would cost considerably less.  Unfortunately it is also somewhat less accurate than theBoom line of microphones.  If you don't mind going with a single speaker design, you'll find lots of choices on our Microphone Comparison Matrix where you'll be able to compare noise canceling, pricing, accuracy and comfort.


-------------------------


Click KB 2012 REV D to Download a 30 Day Evaluation of KnowBrainer 2012
Click SpeechStart+ to Download a 15 Day Evaluation of SpeechStart+

 11/09/2007 08:30 PM
User is offline View Users Profile Print this message

Author Icon
Chucker
Top-Tier Member

Posts: 10912
Joined: 10/10/2006

Quote:

We don't believe the TruVoice text-to-speech readers are part of NaturallySpeaking. We believe they are third-party products or possibly something left over from an L & H speech recognition product.



L&H text-to-speech was always RealSpeak 1.0.  All versions of DNS up to 9.5 use RealSpeak 2.0 speakers.  These TTS engines are built into DNS and cannot be adjusted except for speed.  These are special case versions of RealSpeak 2.0 (i.e., the current version of RealSpeak is 4.0).  Actual RealSpeak TTS engines are very expensive (i.e., $5,000 plus for each language version).  TruVoice text-to-speech is a third-party application, never belonged to, wasn't developed by, and was never used by L&H.  The likelihood of being able to incorporate third-party text-to-speech applications (voices & engines) is extremely limited with regard to DNS and requires that any such speech engines be SAPI 4 based.  SAPI 5 TTS (Text-To-Speech) voices will not work with DNS (i.e., internally).

Regardless, even if you are able to get third-party TTS voices and engines to work with, they will work with limited reliability and will be very unstable, including de-stabilizing DNS.

You are certainly welcome to try, but if you have a problem with them or if DNS fails to function properly, don't call Nuance technical support because they will not help you.  They do not support third-party TTS engines.

I tend to agree with Lunis (KnowBrainer) that the RealSpeak voices included with DNS are adequate after adjusting the speed and listening to them through the soundcard speakers vs. headset speakers.  The only way that you can really (i.e., effectively and reliably) use third-party TTS engines is independent from DNS.  That is, to engage them through their own interface and not attempt to get DNS to recognize them.

Chuck Runquist
Former Dragon NaturallySpeaking SDK & Senior Technical Solutions PM for DNS

"If problems can be created by knowledge, then they cannot be solved by ignorance."  Isaac Asimov

 



-------------------------

 11/10/2007 05:10 AM
User is offline View Users Profile Print this message


boblou
Junior Member

Posts: 4
Joined: 11/07/2007

I am not aware of any text to speech being added to my computer, so I have to believe these came off of the Naturally Speaking disks.  Since this program can be installed on more than one computer, I will do this on my brother-in-laws and see if the extra voices show up.  If they do, then these disks have to be the source.  I am not advocating using these because the quality is so poor.  I am just wondering if they are causing me any problems with Jane and Jennifer.  Would it be adviseable to try and remove them, and if so, how can it be done?  Also, if they do show up on the second computer, can we simply uninstall this version?  I see that I can delete Dragon Naturally Speaking through the control panel add/delete, but will this take everything off?  Sorry for all the questions, but I don't want to do something I shouldn't.  Once again, thanks for the help.  You guys are the best monitors I have come across.

 

 11/10/2007 08:07 AM
User is offline View Users Profile Print this message


fl
Senior Member

Posts: 377
Joined: 10/04/2006

The extra voices could have been installed as part of some other software package, without you ever knowing about it. Street Atlas US installs text-to-speech, for example. The extra voices shouldn't hurt. 
 11/10/2007 09:19 AM
User is offline View Users Profile Print this message

Author Icon
Chucker
Top-Tier Member

Posts: 10912
Joined: 10/10/2006

Quote:

I am not aware of any text to speech being added to my computer, so I have to believe these came off of the Naturally Speaking disks. Since this program can be installed on more than one computer, I will do this on my brother-in-laws and see if the extra voices show up. If they do, then these disks have to be the source. I am not advocating using these because the quality is so poor. I am just wondering if they are causing me any problems with Jane and Jennifer. Would it be adviseable to try and remove them, and if so, how can it be done? Also, if they do show up on the second computer, can we simply uninstall this version? I see that I can delete Dragon Naturally Speaking through the control panel add/delete, but will this take everything off? Sorry for all the questions, but I don't want to do something I shouldn't. Once again, thanks for the help. You guys are the best monitors I have come across.



There is absolutely no way that any additional TTS voices or engines could have come from any of the Dragon NaturallySpeaking disks.  Dragon NaturallySpeaking only installs RealSpeak Jennifer (US English) and Jane (UK English) 2.0 voices and the RealSpeak TTS engine.  DNS simply does not include any other Text-To-Speech voices or speech engines on any of its disks.  You can take this to the bank.  Neither L&H, nor ScanSoft, nor Nuance has ever installed any TTS voices other than RealSpeak.  Since Nuance owns RealSpeak, there is no purpose or reason for Nuance to install any other TTS voices/engines in Dragon NaturallySpeaking.  The TTS voices/engines that you're referring to have to have come from another source.

Chuck Runquist
Former Dragon NaturallySpeaking SDK & Senior Technical Solutions PM for DNS

You can fool too many of the people too much of the time. - James Thurber (1894- 1961)



-------------------------

KnowBrainer Speech Recognition » NaturallySpeaking Speech Recognition » Text to speech

Statistics
28318 users are registered to the KnowBrainer Speech Recognition forum.
There are currently 1 users logged in.
The most users ever online was 2028 on 04/05/2013 at 07:36 PM.
There are currently 85 guests browsing this forum, which makes a total of 86 users using this forum.

FuseTalk Standard Edition v4.0 - © 1999-2014 FuseTalk™ Inc. All rights reserved.