Microsoft Speech Recognition - A Review of Windows Vista Speech Recognition

Page content


The face of accessibility users is changing. The stereotypical accessibility user used to be someone over 60 whose farsighted eyes needed extra large fonts to read the screen. This person’s motor skills were not used to small, delicate motions like those used to maneuver a mouse. In fact, they would complain and ask their kids why that thing was called a mouse anyway. They often mutter “This confounded thing,” when referring to their computer, the Internet or their email. The new face of accessibility users looks much different.

The stereotypical accessibility user today is much younger, doesn’t need to puff up the font any, has a solid understanding of technology and has probably done self-searching on Google. This person has been playing video games since they were 15 years old, using a mouse for the last five years, sending out text messages in unlimited quantities each month, wearing wrist guards to work, typing papers for school, or doing data entry at their last two jobs. These people are never over five feet away from some electronic device. The reason they are an accessibility user is because they’ve spent so much time on the computer. Their wrists and hands hurt because of Carpal Tunnel Syndrome (CTS) or Repetitive Stress Injury (RSI).

I was only 33 years old when I was diagnosed with CTS and tendonitis in both of my hands, elbows and wrists. I was playing bass in a band, working as a network engineer at a small insurance company and making my way through grad school papers at 30 pages a pop. After I got over the initial embarrassment of getting a “sissy injury,” I pledged to fight back against the tyranny of the mouse and keyboard. This became my mantra: “If technology gave me this problem, then there must be existing technology to fix this problem.” I’ve used ergonomic keyboards, ergonomic mice, touch pads (, a Wacom digital pad ( ), a hands-free mouse (, and voice recognition software (

I currently use a foot clicker ( coupled with Windows Vista Speech Recognition ( I point with my mouse, click with my foot, type with my voice, pick with my finger and slurp with my mouth. Central to my regiment of computing is Microsoft Windows Vista Speech Recognition.

The voice-to-text engine found in Vista Speech Recognition is powerful, intuitive and extensive. This handy application is Microsoft’s answer to the ever-increasing requests for alternative ways to input data into a computer. This program accepts commands for working with windows, dictates into mainstream word-processing and email programs, fills out forms on the Web, and allows you to click anywhere on the screen with only your voice. Keystrokes are outdated.

The average purchaser of Windows Vista may not even know that Speech Recognition shipped with their operating system. It is not on the Start menu. It is not in the All Programs menu. You have to dig for Speech Recognition. You start Speech Recognition from the Control Panel [see Control Panel Start screenshot.] From there, you can open Speech Recognition, set up your microphone [see Microphone Setup 01-04 screenshots,] train your computer, change settings and get help. With Windows Speech Recognition, your voice is heard right from the start. A guided setup and tutorial familiarizes you with key concepts and commands. Windows Speech Recognition also features an innovative, intuitive user interface that assists you in controlling many operations of your computer by voice. Let’s look some of the available features:

Dictation: this is what saves keystrokes on the keyboard. You can dictate into Microsoft Word, Microsoft Outlook, Notepad, and other various applications to produce emails and documents. You can then use commands to save your document and open and close programs that you’re using.

Commands: this gives you dictatorial leadership over your computer. You can open and close applications, format and save documents, and click anywhere on the screen [see Mousegrid screenshot.] You can switch between applications that you are using, copy and paste text and browse the Internet. You can also print out Common Commands in Speech Recognition [see Speech Recognition Commands 01-06 screenshots] from the Speech Recognition control panel interface [see Voice Recognition Control Panel screenshot.] This will help you to know what to say when using this program.

Correction: you can fix words that were incorrectly interpreted easily by selecting that word using the correction feature in Speech Recognition. It normally gives you different options to choose from.

Simple User Interface: the taskbar can either be “docked” or “undocked” [see Docked At The Top and Undocked screenshots] depending on your preference. When you right click on the taskbar, it gives you many options to choose for correction as well as settings for Speech Recognition.

Constant Learning: you can set up Speech Recognition to learn more about your style from documents that are already on your machine. You can also do more training on your own to teach Speech Recognition more about your speaking style [see Speech Recognition Training, Speech Recognition Training 02 screen shots.] The more that you use Speech Recognition, the better it understands you. The better it understands you, the faster it will run.

Tutorial: there is a tutorial embedded in the program on your machine [see Tutorial screen shot] as well as in online tutorial to teach you how to use this program.

Multiple Languages: Speech Recognition is available in English (United States,) English (United Kingdom,) German (Germany,), French (France,) Spanish (Spain,) Japanese, Traditional Chinese, and Simplified Chinese.

Windows Vista Home Basic can be purchased for $184.99. Voice recognition software from a competitor with all of the options that are available in Windows Vista Speech Recognition costs much more. And you don’t even get this groovy operating system from the competitor for that price. When shopping for voice recognition software, you’d be wise to consider Microsoft Windows Vista with Speech Recognition. By the way, this entire review was written through the power of Vista Speech Recognition.

Price to Value (5 out of 5)

What’s Hot:&nbsp

For around $185, you get a solid, sexy, user-friendly operating system from Microsoft. On top of that, included in the price, is a fast and intuitive voice recognition application. When lined up against the competition, Windows Vista Speech Recognition is a fine wine on a beer budget.

Most computers use Microsoft Windows. A large percentage of computers running Windows have some sort of Microsoft Office equivalent on them. Every job I have worked at uses Microsoft Outlook for email. With that being said, it seems almost imperative that the voice recognition software will need to work easily within a Microsoft Office environment. The competition (Dragon Naturally Speaking 9) does have a version of voice recognition software that will work with all Microsoft Office applications. It retails on their website for $899 ( You have to pony up to their Professional package in order to get all the features that you find in Vista Speech Recognition, again without an operating system surrounding it.

What’s Not:&nbsp

You only spend $185 for the operating system but you may spend a lot more to get a machine that will efficiently run Vista as well as Speech Recognition. In checking out to see how much processing power this Speech Recognition uses when I am speaking and it is interpreting, I found that it looks to be using about 8-10 percent of a quad core processor. It also uses about 75MB of RAM. You need good hardware to run Speech Recognition without going insane waiting for the program to respond to your voice. More on this in the Performance section.

Performance (3 out of 5)

What’s Hot:&nbsp

When coupled with good hardware, Vista and Speech Recognition perform quite well. These two programs are fast, intuitive and powerful. Speech Recognition responds very quickly to commands and dictation. I can speak well over 100 words per minute with this program.

The Common Commands in Speech Recognition white pages are extremely helpful in gaining fluency with this program. Any question that you probably have concerning how to use this program is in the help file. It is very thorough and well worth the time and paper to print out.

This program recognizes speech very well right from the start. I can set up my microphone and speak out an email without the program doing any training on my voice, and there aren’t that many corrections to do. However, with just eight to ten minutes of initial training, you can save yourself a lot of hassles. Plus, the program “learns” as you go. The more times that the program hears you correct a word initially, the less times you’ll need to correct it in the long run.

What’s Not:&nbsp

Good hardware is essential for this whole Vista Speech Recognition thing to work. I just barely ran Windows Vista on my 1.8GHz P4 processor with 512MB RAM. In order to get a good response from Speech Recognition, you’ll need a fairly beefy computer, which can add a lot to your upgrade budget. Without good hardware, the response time for Speech Recognition can become unbearable. The minimum requirements on the website don’t say this, but my personal experience would tell me that a dual core processor with at least 1GB of RAM is needed for good performance. You should get a good sound card and set Speech Recognition to run through it as well [see Control Panel Start 02 and Sound Control Panel screenshots.] Again, from personal experience, running my microphone through the onboard sound has been less than satisfying as the computer doesn’t “hear” as well.

Until I started using this program, I had no idea that my speech quality decreased so much when I’m tired. If it’s the end of a long day, the amount of corrections that I need to do after each sentence goes up exponentially. I must slur like I’ve just killed a bottle of Jack Daniels. When I’m fresh in the morning, I don’t have nearly as many corrections. I may be able to speak over 100 words per minute but the corrections cut that in half a lot of times. Sometimes, the taskbar just asks “What was that?” That is particularly frustrating when I have repeated myself five times. (Now I know how my wife feels.) For instance, if you ever mess up on the word “that,” you’ll have to use the keyboard to correct it. You cannot say “Select ‘that’” because it will select the entire line that you just said. Here are a few other examples that happened while I was writing this review:

What I said: “The more that you use Speech Recognition.”

What the computer heard: “The more they use speech recognition.”

What I said: “I can speak well over 100 words per minute with this program.”

What the computer heard: “If wrong thing I can speak well over 100 words permit with this program.”

What I said: “The correction dialogue box [see Correction Dialogue Box screenshot] is great at giving alternates for quick correction.”

What the computer heard: “The correction dialog box [C correction dialogue box screenshot] is great and giving the alternates for quicken correction.”

Because I have used voice recognition software for so long, I’ve come to expect that correction is all par

Product Features (5 out of 5)

What’s Hot:&nbsp

Microsoft Windows Vista Speech Recognition is filled with features that not only let you dictate emails and documents, but let you control your entire computer. Since this program is proprietary to Microsoft, nearly all Microsoft products work with it well. The dictation function is world class. It is a fast and easy-to-learn alternative to typing. The correction dialogue box [see Correction Dialogue Box screenshot] is great at giving alternates for quick correction. Note: to more quickly and more effectively correct sentences, say a phrase of words to correct rather than just one word. It will be easier for Speech Recognition to locate and fix a phrase rather than just one word. Because of the built-in commands, you can conceivably sit down at your computer and hardly use the mouse or keyboard. Also, the tutorial is quite helpful for getting you started in this world of voice recognition.

What’s Not:&nbsp

I wish there was a shortcut to open up Speech Recognition. You have to go into Control Panel and start it from there every time you want to use the program. There is the option to have it start every time you start your computer. I’m a minimalist though when it comes to my startup programs. I’m not into buying a lot of RAM and then seeing it all used up by leeches in my startup.

Also, whatever language that you installed on your computer is the only language that the Speech Recognition will recognize. You cannot use Speech Recognition in some other language. If you wanted to teach yourself to speak Spanish using Speech Recognition and your computer is laid out in English, you are out of luck.


Control Panel Start

Control Panel Start 02

Sound Control Panel

Voice Recognition Control Panel


Microphone Setup

Microphone Setup 02

Microphone Setup 03

Microphone Setup 04

Speech Recognition Commands

Speech Recognition Commands 02

Speech Recognition Commands 03

Speech Recognition Commands 04

Speech Recognition Commands 05

Speech Recognition Commands 06



Docked at the top

Correction Dialogue Box

Suggested Features

As stated before, it would be great to have a shortcut on my Quick Launch bar to open Speech Recognition. It’d be great to keep Speech Recognition out of my startup programs but make it quickly and easily accessible from my desktop.

There are also some easier-to-use commands found in the competition. For instance, in Speech Recognition to go to the end of the line that you’re working on, you have to say “go to the end of the line”. In the competition, you just need to say “end of line”. Also, in the competition, you can speak a line and say “bold that”. In Speech Recognition, you have to say “click bold”. It then takes your cursor away from the line that you’re working on. These small things add up to one big thing that I would like to see in this program: more personalization of commands. It would be nice if we were able to choose commands to do specific tasks. Other than that, this program really is intuitive.


The features and performance found in Vista Speech Recognition are gold bond compared to the competition. Not only do you get a solid, state-of-the-art operating system with a Vista purchase, you get a powerful Microsoft-native Speech Recognition application. The science of voice recognition is getting better all the time. Microsoft Windows Vista Speech Recognition is at the forefront of that technology. There are some frustrating moments with Speech Recognition but that shouldn’t be a surprise to anyone who spends a lot of time on a computer. Speech Recognition is intuitive to how a person computes and, as stated before, the more you use it the better it gets. For those of us with accessibility needs, Speech Recognition opens up a whole new world. There’s finally an alternative to using our hands on mice and keyboards. Besides, we can now use our hands for important things like popping Pez into our mouths while writing emails. So sit down, take a load off and find your voice with Microsoft Windows Vista Speech Recognition.

Dragon Naturally Speaking, Nuance Naturally Speaking