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

Well, it's past the point where we can make any changes in the code, but we can still make changes to the Easter Egg!

Rudiger,

How did you get your hands on Windows 8?

MDH

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

Mark,

the consumer preview (evaluation copy) is a download provided by Microsoft and free to anyone. Just let Google show you the way - and help them keep track of your data :-)

Rüdiger

 

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

Well, it's past the point where we can make any changes in the code, but we can still make changes to the Easter Egg!

Interesting Rüdiger. 

It seems the future is upon us and computers are becoming ever more powerful with Microsoft considering 128 bit versions of Windows 8 & 9. 

http://arstechnica.com/microsoft/news/2009/10/microsoft-mulling-128-bit-versions-of-windows-8-windows-9.ars

Cheers

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

Tiger Feet

| DNS 12 Professional | KnowBrainer 2012 | Windows 7 Professional /64 Bit | Intel Core i7 Quad Core 3840QM (2.80GHz) 8MB | 16GB RAM. | 240GB Intel 520 Series SSD Boot Drive | 750GB 7200 rpm Secondary Drive  | Plantronics Wireless Handsfree Microphone CS60-USB. |

Tiger,

That article is almost 3 years old. Given that hardware and software are years behind even making full use of 64-bit operating systems, it doesn't make a whole lot of sense for Microsoft to even be considering (currently) going to 128-bit for Windows 8 and it would still not be practical even in Windows 9. Crunch the numbers. 64-bit is 2 ⁶⁴ bytes and 128-bit is 2 ¹²⁸ bytes.

Microsoft may be considering the possibility of 128-bit Windows for heavy-duty servers, but it will still be at least a decade before consumer hardware catches up to the full capacity of 64-bit operating systems. In addition, going to 128-bits would require major hardware changes because the current state of hardware won't support 128-bits. If only from a practical standpoint, introducing a 128-bit version of Windows wouldn't make a whole lot of sense. If you think about it, going to 128-bit would be like buying a Boeing 787 to go two blocks to the drugstore. To put it more simply, the current state-of-the-art in terms of CPUs, including Ivy Bridge, only support EMT64 (64-bit instructions). Intel would have to introduce EMT128 CPUs. Your current hardware wouldn't even support 128-bit Windows. Besides, the cost of going that route would be prohibitive for the average commercial end user.

The bottom line is that retail 128-bit versions of Windows are a long, long, long way off.

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, We make it better!

"Everything takes longer and costs more."  -- Jerry Pournelle

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

It's the consumer preview and is available here:

http://windows.microsoft.com/en-US/windows-8/download

I wouldn't install on a production machine as it's not finished.  I have it on a 2nd machine in my office and it looks good.  

 

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

Bill
DNS Pro v12.5, KB 2012, Mtech Desktop PC (i7 960 3.2 gHz with 12 GB RAM), Windows 7 Pro 64-bit, 240 GB SSD, Philips SpeechMike 3500, SpeechWare 3-in-1 TableMike, Philips SpeechMike 5274 Classic, MS Office 2010 Professional

Have you checked out the speech rec on there. If it is on there??

R 

To my adventerous colleagues who are experimenting with Windows 8, if you could be so kind -- Rag's question is my burning question, too! As much as I love Dragon, I find WSR indispensible. "Show numbers" is a brilliant feature for us hands-free users. So please report back if you plan on playing with the native speech recognition in the preview build.

Thanks!

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

Eric Wright At work: DNS 12 Pro. At home: DNS 11.5 Pro,  KnowBrainer 2011, and Utter Command by RedStart Systems; Dragon Dictate 3 for Mac

 

Appetite for Dictation - My Blog

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

www.pcbyvoice.com
www.pcbyvoice.co.uk

Lindsay,

My question is fully understanding that this is a preview version and given that I haven't had time to play around with yet, have you found any differences in the features and/or capabilities, commands, etc. yada yada yada between the Windows Speech Recognition (WSR) in Windows 8 and the Windows Speech Recognition (WSR) in Windows 7? That is, is it better or is it the same? I'll get around to setting up my system with the Windows 8 preview release in the next week or so, but I'm just curious as to whether or not you have noted any differences.

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, We make it better!

"Everything takes longer and costs more."  -- Jerry Pournelle

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

Please remove.




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

www.pcbyvoice.com
www.pcbyvoice.co.uk

Lindsay,

Basically what I was referring to, which I guess I will have to test for myself, is whether there are any new feature that don't currently exist in Windows Speech Recognition (WSR) under Windows 7. The "show numbers" is currently a part of Windows 7 Windows Speech Recognition (WSR), so that's not a new feature unless it's better supported under Windows 8 or has some new gimmicks.

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, We make it better!

It is of interest to note that while some dolphins are reported to have learned English -- up to fifty words used in correct context -- no human being has been reported to have learned dolphinese. - Carl Sagan (1934 - 1996)

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

Chuck,

I know what you meant but I just wanted to put some pretty graphics on the screen  :-) as you know Microsoft are using some new artificial intelligence algorithms with their speech recognition and have promised some "breakthroughs".  From what I can see so far this is certainly not implemented in Windows speech recognition with Windows 8, it is pretty much the same as it was with Windows 7.  However I have only tried it quickly.  As it is free and the whole of Windows 8 along with a speech recognition is cheaper than DNS Professional then we can't complain really.

Lindsay

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

www.pcbyvoice.com
www.pcbyvoice.co.uk

Is anyone aware that they are also in for this app monkey business? - Just mentioning this since you have mentioned artificial intelligence. Not so sure about the association in here.

Art

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

Well, it's past the point where we can make any changes in the code, but we can still make changes to the Easter Egg!

Lindsay,

Both Microsoft and Nuance are working on the same thing (concept). I'm not sure that it has anything to do with artificial intelligence, but you can look at the basic research and decide for yourself:

http://research.microsoft.com/pubs/144412/DBN4LVCSR-TransASLP.pdf

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, we make it better!

"Many of the things you can count, don't count. Many of the things you can't count, really count." Albert Einstein

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

Chuck,

That's the same paper as I read from Speech Computing I think.  They are using artificial neural network programming which is what the artificial intelligence guys use to mimic the neural networks in the brain.  They are adaptive networks (both artificial and real neural networks in the brain) which basically adapt to massive complex patterns which aid in memory and logic decision-making.

Lindsay



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

www.pcbyvoice.com
www.pcbyvoice.co.uk

Can someone PLEASE do a test re the WSR accuracy???

Rag,

Accuracy has never been the issue with Windows Speech Recognition (WSR), except perhaps in Windows Vista to a certain extent. Because VoicePower is designed for both Dragon and Windows Speech Recognition (WSR), we do very comprehensive testing with Windows Speech Recognition (WSR) in Windows 7 on a regular and ongoing basis. I'm just beginning to experiment with Windows 8, but I do not find the accuracy to be an issue in any case whether it be Windows 7 or Windows 8. In short, the issue is not accuracy as much as it is that Windows Speech Recognition (WSR) is lacking the comprehensive vocabulary support available in DNS. When taking that into consideration, the accuracy and Windows Speech Recognition (WSR) is generally as good as DNS. Keep in mind that I'm talking about accuracy in general, not total functionality. If Windows Speech Recognition (WSR) makes mistakes (misrecognitions) it's generally a vocabulary issue. Nonetheless, DNS does have the advantage of better context analysis by virtue of using Quadgrams vs. the fact that Windows Speech Recognition (WSR) is only using Trigrams. Depending upon your dictation style, this will tend to make DNS slightly more accurate. On the other hand, if you're dictating in short choppy phrases, which is very characteristic of the average user, instead of longer phrases and sentences, then Windows Speech Recognition (WSR) is specifically designed with this limited dictation style in mind.

The bottom line is that when comparing accuracy between Windows Speech Recognition (WSR) and Dragon NaturallySpeaking, they are comparable. However, it generally depends on what you are talking to as to whether or not one is more accurate than the other. This argument has been presented many times over the years relative to comparisons between L&H VoiceXpress and Dragon NaturallySpeaking and ViaVoice vs. Dragon NaturallySpeaking. The conclusions have generally and always been dependent upon whom you're talking to.

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, We make it better!

Be careful what you ask for. What you get might not be what you expect, or want. - Aesop (620 BC - 700 BC)

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

I tried using WSR with Vista.  Dictation recognition wasn't that bad, but the real dealbreaker was that you could not dictate into many applications.  Sure you can dictate into Outlook, Word, WordPad, and Notepad.  But you couldn't dictate into much else.  If the application didn't support Text Framework Services, it was a no go.  You couldn't even dictate into Visual Studio Microsoft's own IDE.  Has anything changed with WSR in Windows 8?


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

Dragon in use.  Occasionally mistakes are made.  I correct the ones I notice... 

cactustweeter,

Unfortunately, you stopped short by virtue of only testing with Windows Vista. There were numerous changes in Windows 7 relative to Windows Speech Recognition (WSR) that made it usable with more applications. However, it doesn't matter whether you're using Windows Speech Recognition (WSR) or Dragon NaturallySpeaking, if a text window does not support Full Text Control (Select and Say), or, as you put it, Text Framework Services, then it's considered a nonstandard window and dictation is limited and/or not possible at all depending upon the text edit controls used by the application.

In Windows 7 Windows Speech Recognition (WSR) introduced something similar to Dragon's dictation box, poorly implemented as it is. Nevertheless, Windows Speech Recognition (WSR) under Windows 7 does allow you to dictate even in nonstandard Windows using their implementation of the "dictation box". The bottom line is that because you've never experimented with Windows Speech Recognition (WSR) in Windows 7, you haven't seen this.

Lindsay and those who are currently experimenting with such might be able to address whether or not Windows Speech Recognition (WSR) in Windows 8 has any better functionality, particularly with regard to Microsoft's Visual Studio. However, Visual Studio was never designed to support Full Text Control (Select and Say) or Text Framework Services, just as Microsoft Access was never designed for such either. Perhaps Microsoft can make an adjustment in the next versions of Visual Studio and/or Microsoft Access and similar applications. That's open to Microsoft both in terms of whether it can be done and/or whether it should be done. Regardless, support for Text Framework Services and/or Full Text Control (Select and Say) is dependent upon the application developers, not Microsoft or Nuance. Microsoft provides all the technical specifications that are needed for application developers to implement such, but like the saying goes, "You can lead a horse to water, but you can't make him drink." Some application developers refuse, such as the Firefox programmers and some of the EHR/EMR developers. You can't put the blame on Microsoft and/or Nuance because they can only "lead a horse to water." Whether or not Microsoft is moving in the direction of doing the best they can to increase application support given the above is something we can only speculate about. Only Microsoft knows what they or aren't doing. Nevertheless, from what I've seen so far, the preview version of Windows 8 does not contain anything new yet as far as Windows Speech Recognition (WSR) is concerned, even though it is a known that they are working toward this end. Only the final release will tell. On the other hand, Nuance has done everything they possibly can to support what they can in terms of other applications. One example of this is that there is now Full Text Control (Select and Say) support for OpenOffice writer up to and including version 3.3. As far as the Medical version is concerned, Nuance has put out a request to all EMR/EHR application developers to work with them. Some have, but many still have not. So, you can't blame Nuance in that sense for lack of trying.

Regardless, whether you're dealing with Microsoft Windows Speech Recognition (WSR) or Nuance's Dragon NaturallySpeaking, application support can only go as far as those issues over which either Microsoft or Nuance have control. Beyond that, you need to point the finger elsewhere.

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, We make it better!

Be careful what you ask for. What you get might not be what you expect, or want. - Aesop (620 BC - 700 BC)

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

Chucker,

You misunderstood me.  When I mentioned that without an application having Text Framework Services support it was unusable, I was not complaining there was no Select-and-Say.  Sure Text Framework Services enable Select-and-Say, but without Text Framework Services support in Vista you could not even put any text directly into the application.  Did anything change in Windows 7 or 8.  I know Windows 7 WSR has a dictation box, but having to always work through it is a major productivity hindrance.

Forget about Visual Studio, or Microsoft SQL Server Management Studio.  WSR will not let you dictate into Notepad++, Xemacs, or putty to mention a few more unacceptable applications for WSR.  DNS does not offer Select-and-Say in these applications either, but I still can get lots of work done without it using DNS.  DNS is a true productivity tool.  WSR is a toy.

By the way there is a freeware add-on for WSR called Vocola 3 that when added allows limited dictation directly into a non-Text Framework Service application.  Rick Mohr truly deserves much thanks for providing this for WSR.


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

Dragon in use.  Occasionally mistakes are made.  I correct the ones I notice... 

Quote:

Did anything change in Windows 7 or 8. I know Windows 7 WSR has a dictation box, but having to always work through it is a major productivity hindrance.

Cactustweeter,

Some differences in WSR Windows 7 to that of WSR in Vista.  4 examples:

1. You can have the microphone turn off instead of sleeping.  Probably not a good idea for someone with limited accessibility.

2. You can have a dictation box pop up in non compliant Windows as you already know.

3. You now have to say, "Capitalize" instead of "Cap That"

4. The speech meter is not as sensitive as it is in Vista and not as likely to turn on from a sleeping state by the slightest noise.

I can confirm that WSR dictation in Windows 7 does not work in Notepad ++ directly.  WSR text macros do.

The dictation box comes up when you try to dictate anything using WSR in Notepad ++ so you can dictate into the dictation box and then transfer the wording by saying "Insert." 

Dragon NaturallySpeaking dictation works fine in Notepad ++ so my guess is for WSR dictation to work; it needs a plug-in of some sort just as DNS needs a plug-in (Add-in) for use in Microsoft Word.  If there is one available or where you can get it is anyone's guess.  There might not be one available for that particular application in which case you could probably e-mail the makers of the program or use their forums and suggest that there should be one available.

Contrary to what Chuck says, WSR dictation in Windows Vista and 7 is not as accurate as DNS is out of the box so to speak.  If it were, Dragon NaturallySpeaking prices would drop like a stone, especially in the Professional versions simply because of the fact that WSR is free and part of Microsoft's operating system since Vista.  If, as Lindsay has already found out, WSR in Windows 8 is using exactly the same speech engine as Windows 7, then the accuracy thing and WSR's dictation capabilities will not have changed.

WSR, along with the free macro programs that Microsoft provides for it can already, emulate what the Professional versions in DNS can do.  There has never been a problem with WSR on the command side.  It can match DNS here and of course it has ShowNumbers built into its workings.  You can also make your own command and text macros in WSR.

As regard to lack of 'comprehensive vocabulary support,' out of the box, yes this is correct.  On the other hand, you can make a WSR profile as comprehensive or not as you like by adding as many words and phrases as you like to its speech dictionary.  In WSR, to open the dictionary, you just say "Open Speech Dictionary."  Once you have added your word(s), they become part of the WSR Profile vocabulary.  I have to concede though, DNS has much more detail in relation to adding words or phrases and WSR lacks some of DNS' features.  You can add words that even DNS does not possess out of the box in WSR. 

For example; using WSR, just as an experiment, I tried saying the very long word below in blue shown in the next paragraph.  It is supposedly the longest word in the Oxford English dictionary.  It is, according to the Oxford English Dictionary, a factitious word alleged to mean a lung disease caused by the inhalation of very fine silica dust, causing inflammation in the lungs.  A condition meeting the word's definition is normally called silicosis. 

Pneumonoultramicroscopicsilicovolcanoconiosis

It came out as shown below and I do not know if I pronounced it correctly anyway:

Your-month-old running Roscoe Pixley co-wrote an icon I process. 

Trying to say it without adding and training it using Dragon, didn't do much better either.

So, I added that word to WSR's speech dictionary.  After I added it, I recorded a pronunciation of the word in the speech dictionary by breaking it down into sections (syllables) as shown below.  The speech dictionary then told me the word had been successfully recorded and added.  I then listened to my recording in the speech dictionary just to hear how it sounded.

Pne-umon-oult-ramic-rosco-picsili-covol-canoco-niosis

Now, as long as I break it down as shown above without any pauses, the word comes out on screen using WSR every time.

I did exactly the same with the word Antidisestablishmentarianism

Once again when dictated untrained, it came out as shown below:

Ab establishments eeriness

So I went through the same procedure as explained above and broke the word down into syllables again as shown below.

Anti-disestablish-mentarian-ism

Now, when I say that word broken down as I have shown above, it comes out correctly every time on screen using WSR.

There is also an application for managing WSR that is very useful but we won't go there.

Regarding homophones, most of the time, I generally do not have a problem with these in WSR when spoken in context.  That situation is comparable with Dragon, for me anyway.

Having said all that, even with a full vocabulary of added words, WSR has never been as accurate as Dragon on the dictation side.  This has been pointed out many times by the people who actually use it, so from experience, not from someone who never uses it intensively over a period of years or salesmen who are trying to tell you otherwise.  Also, the accuracy issue/difference occurs no matter what microphone you are using when comparing DNS to WSR dictation.

I repeat, if Microsoft got their act together on the accuracy issue, then DNS would have some real competition.

The main reasons why WSR is not a priority I suspect is that programmers can use it and fill in the gaps.  Mainly, WSR was developed for the disabled, people who generally cannot afford to buy Dragon

My reasoning is using Dragon and WSR simply gives me more choices rather than using just one or the other.  I consider them a joint application and with the two of them together, makes for a more powerful one, especially for the disabled but then again for the able-bodied too.

Cactustweeter, have you got details about Vocola 3 and how to add it?

Cheers

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

Tiger Feet

| DNS 12 Professional | KnowBrainer 2012 | Windows 7 Professional /64 Bit | Intel Core i7 Quad Core 3840QM (2.80GHz) 8MB | 16GB RAM. | 240GB Intel 520 Series SSD Boot Drive | 750GB 7200 rpm Secondary Drive  | Plantronics Wireless Handsfree Microphone CS60-USB. |

Tiger,

Don't know about you but I have no problems with dictating into this forum or anywhere in IE 9 with Windows Speech Recognition (WSR). However, if you're using Firefox or Chrome, then that would explain why you're having a problem.

I didn't say they were equal, what I said, which may have been misleading, was that accuracy has not been the [main] issue relative to Windows Speech Recognition (WSR). I'm sure that can be interpreted several ways and for that I apologize. You can't compare Windows Speech Recognition (WSR) to Dragon NaturallySpeaking out-of-the-box anyway because Dragon doesn't require any training and even gives you reasonably good accuracy using the, more or less, speaker independent Acoustic Model. However, Windows Speech Recognition (WSR) has to be trained using at minimum two training sessions.

Windows Speech Recognition (WSR) also has a limited Active Vocabulary, which means that many of your initial misrecognitions are vocabulary issues (i.e., missing vocabulary). Many of the other initial accuracy issues are a result of the fact that Windows Speech Recognition (WSR) parses your dictation in terms of trigrams. That is, it is extremely fast because it only looks at trigram utterances. In short, it takes much longer for Windows Speech Recognition (WSR) to achieve a high degree of recognition accuracy.

The long and the short of it is that there is much more to overall accuracy than simply the speech engine. The Acoustic Model, the vocabulary, and the Language Model are significantly more well developed in Dragon NaturallySpeaking by a long shot than in Windows Speech Recognition (WSR). The biggest problem with Windows Speech Recognition (WSR) is its lack of productivity features. However, one of its main strengths is that once you make a correction, Windows Speech Recognition (WSR) learns immediately from that correction and doesn't make the same mistake the second time around.

Comparing Windows Speech Recognition (WSR) to Dragon NaturallySpeaking is like comparing DNS 4.0 to DNS 11.5. Windows Speech Recognition (WSR) is still in its infancy, which is why I was hoping that they had implemented their new algorithms in the consumer preview version of Windows 8, which they have not.

The bottom line is that if you're going to compare accuracy in Windows Speech Recognition (WSR) to accuracy in DNS, then you have to do it point by point. Otherwise, as you can see from your confusion relative to my point, it's easy to overgeneralize.

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, We make it better!

Management is doing things right; leadership is doing the right things. - Peter Drucker (1909 - 2005)

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

Quote:

Don't know about you but I have no problems with dictating into this forum or anywhere in IE 9 with Windows Speech Recognition (WSR). However, if you're using Firefox or Chrome, then that would explain why you're having a problem.

Chuck,

Now who is the confused one?  If you read my post correctly, I was confirming Cactustweeter's question that Windows Speech Recognition's dictation in Windows 7 does not work in Notepad ++.  Nothing to do with IE 9 whatsoever.  I'm sorry if it confused you but I'm sure he knows what I meant as do others.

Once again just for you, dictation using WSR in Windows 7 does not work in Notepad ++.  When you try to dictate in this application using WSR, the dictation box comes up and you have to use that.  DNS on the other hand does work in this application as Cactustweeter pointed out in his post and tried to make you aware of the same.  I have never had a problem using WSR dictation within IE 9 and never said I did.

Quote:

I didn't say they were equal, what I said, which may have been misleading, was that accuracy has not been the [main] issue relative to Windows Speech Recognition (WSR). I'm sure that can be interpreted several ways and for that I apologize.

This is what you said in your above post:

Quote:

Accuracy has never been the issue with Windows Speech Recognition (WSR), except perhaps in Windows Vista to a certain extent.

How else is anybody supposed to interpret that?  I disagree.  Accuracy has been exactly the issue with Windows Speech Recognition.  I have been using it since it first came out in Vista even with a full user profile vocabulary with many words trained and added to it.

And how many times have you said it in your post above?  Just take a look for yourself below.

Quote:

I'm just beginning to experiment with Windows 8, but I do not find the accuracy to be an issue in any case whether it be Windows 7 or Windows 8. In short, the issue is not accuracy as much as it is that Windows Speech Recognition (WSR) is lacking the comprehensive vocabulary support available in DNS. When taking that into consideration, the accuracy and Windows Speech Recognition (WSR) is generally as good as DNS. Keep in mind that I'm talking about accuracy in general, not total functionality. If Windows Speech Recognition (WSR) makes mistakes (misrecognitions) it's generally a vocabulary issue. Nonetheless, DNS does have the advantage of better context analysis by virtue of using Quadgrams vs. the fact that Windows Speech Recognition (WSR) is only using Trigrams. Depending upon your dictation style, this will tend to make DNS slightly more accurate. On the other hand, if you're dictating in short choppy phrases, which is very characteristic of the average user, instead of longer phrases and sentences, then Windows Speech Recognition (WSR) is specifically designed with this limited dictation style in mind.

It does not matter how you spin it above, bottom line is DNS is more accurate than WSR at the moment.  I also disagree with your point about speaking in context.  I have found WSR just as adequate as DNS where using homophones in context are concerned if that's what you mean.

Again, from your post above:

Quote:

The bottom line is that when comparing accuracy between Windows Speech Recognition (WSR) and Dragon NaturallySpeaking, they are comparable.

No they are not.  If they were, Dragon sales would drop.  They are your words above not mine.

Now compare what you have said above to what you have said below in your most recent post.

Quote:

You can't compare Windows Speech Recognition (WSR) to Dragon NaturallySpeaking out-of-the-box anyway because Dragon doesn't require any training and even gives you reasonably good accuracy using the, more or less, speaker independent Acoustic Model. However, Windows Speech Recognition (WSR) has to be trained using at minimum two training sessions.

You are contradicting yourself all the way.  Perhaps not meaning to I suspect.

Quote:

Windows Speech Recognition (WSR) also has a limited Active Vocabulary, which means that many of your initial misrecognitions are vocabulary issues (i.e., missing vocabulary). Many of the other initial accuracy issues are a result of the fact that Windows Speech Recognition (WSR) parses your dictation in terms of trigrams. That is, it is extremely fast because it only looks at trigram utterances. In short, it takes much longer for Windows Speech Recognition (WSR) to achieve a high degree of recognition accuracy.

More spin and I have addressed the vocabulary issue already.

Quote:

The long and the short of it is that there is much more to overall accuracy than simply the speech engine. The Acoustic Model, the vocabulary, and the Language Model are significantly more well developed in Dragon NaturallySpeaking by a long shot than in Windows Speech Recognition (WSR). The biggest problem with Windows Speech Recognition (WSR) is its lack of productivity features. However, one of its main strengths is that once you make a correction, Windows Speech Recognition (WSR) learns immediately from that correction and doesn't make the same mistake the second time around.

I'm not disagreeing with you here but at the end of the day, the average person who uses WSR is not interested in all the technical jargon.  All they are interested in is how accurate dictation is.  They soon find out that it is not as accurate as DNS where dictation is concerned.  The best people to ask are the people who use it.

Quote:

Comparing Windows Speech Recognition (WSR) to Dragon NaturallySpeaking is like comparing DNS 4.0 to DNS 11.5. Windows Speech Recognition (WSR) is still in its infancy, which is why I was hoping that they had implemented their new algorithms in the consumer preview version of Windows 8, which they have not.

Well it's still likely to stay in its infancy if Microsoft have not improved it in Windows 8 don't you think??

In no way am I confused Chuck, just stating the facts and the difference between DNS and WSR's dictation accuracy.  Something you seem to have backtracked on from your original post.

Do not get me wrong, I use WSR alongside DNS because WSR has its uses over DNS in my view and considering it's free, it is a dam fine piece of software that can be used alongside DNS.  It's just when dictation is needed, DNS is more accurate than WSR at the moment period.  That's how any new user will evaluate it even if WSR had a full vocabulary because I already have, for one.

Cheers

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

Tiger Feet

| DNS 12 Professional | KnowBrainer 2012 | Windows 7 Professional /64 Bit | Intel Core i7 Quad Core 3840QM (2.80GHz) 8MB | 16GB RAM. | 240GB Intel 520 Series SSD Boot Drive | 750GB 7200 rpm Secondary Drive  | Plantronics Wireless Handsfree Microphone CS60-USB. |

Tiger Feet,

Thanks for confirming my suspicions about Windows 7 WSR and NotePad++.  You can find more information about Vocola 3 here:

http://vocola.net/v3/

 

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

Dragon in use.  Occasionally mistakes are made.  I correct the ones I notice... 

cactustweeter,

Not sure why you're using Notepad++. If it's for creating and editing HTML documents, or writing programming code, then I understand.

However, if you're just using it as a Notepad replacement for creating and editing text documents, you might want to try "Notepad Tabs".

Just a suggestion FYI because Notepad Tabs is fully supported under Windows Speech Recognition (WSR). If it's not relevant to your situation, just ignore this post.

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, We make it better!

"Kindness is the language which the deaf can hear and the blind can see." -- Mark Twain

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

Bill,

have you figured out what happened to the control panel?

Rüdiger

 

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

Well, it's past the point where we can make any changes in the code, but we can still make changes to the Easter Egg!

Double-click on the Windows Explorer folder next to the start menu.

Right click on Computer and create a shortcut, stick it on your desktop

Double-click on computer and you will see OPEN Control Panel on the menu at the top of the window.

Once you open it then choose category as you would with Windows 7 and then choose small or large icons to see the classic Control Panel.

Bill Gates

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

www.pcbyvoice.com
www.pcbyvoice.co.uk

Thank you so much, Watson. So wouldn't you agree that they put a lot of effort into it to make the whole thing look like an iPod? I never wanted one, but now I have got it, and it came for free.

Bjarne Stroustrup

 

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

Well, it's past the point where we can make any changes in the code, but we can still make changes to the Easter Egg!

I never wanted an ipod/ipad/itouch

If I ever have to go to Windows 8, I'll do what I can to make it look like XP, unless by some miracle it runs better as is.

Even in Vista and Windows 7 you lose a lot of performance if you want all the "eye candy".

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

----

 .

 .

 .

 .

Help! I can't see vowels!! (and sometimes "y")

Yes very iPod or more specifically very Windows 7 Mobile.  With the Control Panel you can just click on Computer on the left-hand pane of Windows Explorer and then the Computer menu at the top of the display, no doubt there is much quicker way.






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

www.pcbyvoice.com
www.pcbyvoice.co.uk

There is a much easier way to get to the control panel: move your mouse to the lower left-hand corner and you will see the "new" start button.  Instead of left-clicking, right click on that and you will be shown a list of functions, one of which is the control panel.

It does not matter if you are in the "classic" desktop, or the new panel display.

And, WSR works great, albeit with a limited test.  I have my old xTag hooked up to the computer (intermediate processor, X64, 4 GB) and I was quite surprised how responsive WSR was.  It even works on the new panel desktop.

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

Bill
DNS Pro v12.5, KB 2012, Mtech Desktop PC (i7 960 3.2 gHz with 12 GB RAM), Windows 7 Pro 64-bit, 240 GB SSD, Philips SpeechMike 3500, SpeechWare 3-in-1 TableMike, Philips SpeechMike 5274 Classic, MS Office 2010 Professional

Thanks Bill, glad to see someone knows what they are doing.  To be honest I did not know that Windows 8 was at anything like the stage it is at until Rüdiger sent me some screenshots last night and then I couldn't resist having a look.

I like the look and feel  I think.

Lindsay

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

www.pcbyvoice.com
www.pcbyvoice.co.uk

I seem to have a problem with the video system built into the VM because the minimal resolution obviously available in Windows 8 won't allow for resizing any further so as to make the window fit into my main 15" screen, which is something new considering I could always resize any other virtual Windows system down to as far as it could get. By the same token, in this situation it is hard to appreciate all the functions you would expect, for instance getting the start menu decently, and I could be wrong, but consequently, what method can be used to shut down other than using the keys Alt + F4?

I am pointing this out just because to me it seems you minimally need to have an appropritate screen with a reasonable amount of video RAM to run it.

Rüdiger

 

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

Well, it's past the point where we can make any changes in the code, but we can still make changes to the Easter Egg!

The best resolution I can get a 1600 by something else, but then I am running Windows 7 on Bootcamp and then virtual box on top of that with Windows 8 so I am not complaining.  In terms of the start menu it is now full screen feature, so I being a hands-free user just say "start menu" and the whole screen is filled with the start menu.  Alternatively you can say or press the keys "control key escape key".  If you right click on the full-screen start menu then it will list all the apps (one day they will wish they stuck to applications).

I seem to be able to resize windows with virtual box but I think you are using VMware?

Angela M.

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

www.pcbyvoice.com
www.pcbyvoice.co.uk

Well Sarko, you don't seem to think, like always, you just know. I can also resize the windows in VMware no problem but with Windows 8, the minimal resolution doesn't allow for showing the entire desktop when going too far. I will have to ask some Greek to buy me a larger screen I bet.

The waiter spilling beers

 

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

Well, it's past the point where we can make any changes in the code, but we can still make changes to the Easter Egg!

I am sure that some 45-year-old Greek retiring on a 5 million Euro pension, paid for by Angela, will be happy to buy you a new screen to show you all the iPod icons on your Windows 8 display.

In the meantime we have to put up with a Prime Minister sitting smugly on his throne thinking he is well clear of the Euro crisis, which is fine until it isn't a crisis any more which won't be long.

Sarko

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

www.pcbyvoice.com
www.pcbyvoice.co.uk

It will be a disaster for ALL!

History repeats itself.

-------------------------
Jomark

Bill,

What do you mean by "WSR works on the new panel desktop?"

If by that you mean that WSR works with the icons on the desktop, WSR has always worked with the icons on top of the desktop, right back to Vista.  This is nothing new.

The burning question is as Chuck mentions, has any work been undertaken on the accuracy issue in WSR?

Cheers

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

Tiger Feet

| DNS 12 Professional | KnowBrainer 2012 | Windows 7 Professional /64 Bit | Intel Core i7 Quad Core 3840QM (2.80GHz) 8MB | 16GB RAM. | 240GB Intel 520 Series SSD Boot Drive | 750GB 7200 rpm Secondary Drive  | Plantronics Wireless Handsfree Microphone CS60-USB. |

Yes, it has always worked with the 'typical' icons on the classic desktop.  What I was surprised at is that WSR was recognizing the names of the panels on the Metro desktop, to my surprise because this was an early consumer review.

I also was surprised at the speed of WSR and its accuracy using my old xTag microphone. But I did not really give it a real scientific examination and really could not tell is if Microsoft has updated the capability of WSR at all.

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

Bill
DNS Pro v12.5, KB 2012, Mtech Desktop PC (i7 960 3.2 gHz with 12 GB RAM), Windows 7 Pro 64-bit, 240 GB SSD, Philips SpeechMike 3500, SpeechWare 3-in-1 TableMike, Philips SpeechMike 5274 Classic, MS Office 2010 Professional

The recogniser with Windows 8 is version 8, the recogniser with Windows 7 is version 8, in other words it is the same speech engine so there won't be any changes in accuracy.

I also asked Microsoft Redmond if there would be any major changes to the speech engine or the functionality of WSR with the eventual release version of Windows 8, the answer was:

"Effectively no. The changes for WSR in Windows 8 are primarily oriented around making it play well with the Metro environment"

Lindsay

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

www.pcbyvoice.com
www.pcbyvoice.co.uk

Hopefully, WSR will work in Windows 8, vs. Windows 7. On both my  Win7-64 computers, WSR won't start at all.  I'm not alone, many similar problems reported (without any solution) if you Google it.   So I've have to wait for Win8 to try WSR.