Software that's fun, easy to use and affordable !
   

Support
Download
Shopping Cart
About us
Contact Us
Links
Blog
Free Speech Recognition!
Click here

TrialPay Referral Program



CoolSoft, LLC is proud to have acheived recognition as a
More Awards
www.Soft3k.com 5 Star Awarded
At CoolSoft, you
can shop with confidence knowing your personal information is secure. Click here to learn more.

Payments processed via the LinkPoint Secure Payment Gateway. 

Click for more information.

Speakeasy Speed Test
Test Your Connection Speed

Welcome to CoolSoft, LLC


SpeakToText™ 2.0 Frequently Asked Questions (FAQ)

 

 

HOW TO:

Q.  What is the difference between Start/Stop listening and Microphone on/off buttons?

Q.  How do I dictate into SpeakToText?

Q.  How do I copy the text I dictated in SpeakToText into Microsoft Word?

Q.  How do I save the text I just dictated?

Q.  How do I use SpeakToText with other programs?

Q.  How do I use SpeakToText to chat hand’s free?

Q.  Can I dictate directly into Microsoft Outlook or Outlook Express?

Q.  What is a user command and how do I create one?

Q.  How do I make SpeakToText read text back to me?

Q.  How do I learn all the speech commands I need for SpeakToText?

Q.  How do I train the speech recognition engine?

Q.  I see that SpeakToText has an update feature.  Is this automatic?

Q.  How can I use the .wav to text feature?

Q.  SpeakToText disappeared.  How do I find it?

 

INSTALLATION AND REGISTRATION:

Q.  I have downloaded the trial version of SpeakToText and installed it.  I speak and it is not working.  What is wrong?

Q.  I purchased SpeakToText from the CoolSoft website.  Now what do I do?

Q.  I purchased SpeakToText on CD from a reseller other than CoolSoft.  Now what do I do?

Q.  I am installing SpeakToText, or the trial version of SpeakToText, and I am getting an error message, “The file “C:\Program Files\SpeakToText v2 (trial)\UNWISE.EXE” could not be opened.  Please check that your disk is not full and that you have access to the destination directory.  The system cannot find the path specified.  What just happened, and how can I install SpeakToText?

Q.  When I try to register STT, I get a message that says "failed registration key."  Why can’t I register SpeakToText?

Q.  What is the difference between the registration key and the CD Key?

Q.  I am using the CD key on the back of my CD sleeve, but I am getting an invalid key message.  Why doesn’t my key work?

 

PURCHASING:

Q.  Does CoolSoft accept Checks or Money Orders?

Q.  Does CoolSoft ship overseas?  If so, at what cost?

Q.  What is the CoolSoft refund policy?

Q.  Why buy SpeakToText from a reseller? 

Q.  Is there a list of CoolSoft resellers?

 

COMPATIBILITY:

Q.  Is STT supported in Windows 98 or ME?

Q.  Is SpeakToText supported in languages other than English?

Q.  Is STT compatible with MAC computers?

 

TROUBLESHOOTING:

Q.  I am dictating with SpeakToText.  The text I dictate shows up and disappears immediately.  What is causing this?

Q.  I am dictating with SpeakToText.  When I am not speaking, it is recognizing words that don’t make sense.  How can I fix this?

Q.  I am not getting good accuracy.  Is there anything I can do to improve accuracy?

Q.  Why is SpeakToText making mistakes?

 

 

 

HOW TO:

Q.  What is the difference between Start/Stop listening and Microphone on/off buttons?

A.  Start/Stop listening is like a mute button for the microphone.  When listening is stopped, SpeakToText can still hear commands like “click focus,” “start listening/wake up” and “what can I say/close me.”

 

When the microphone is turned off, SpeakToText cannot hear anything and the microphone must be turned on again by clicking on the microphone button or using the Microphone On/Off option in the Tools menu.

 

Q.  How do I dictate into SpeakToText?

A.  Start SpeakToText and it opens in basic mode with the microphone on.  Say “start listening” or “wake up” or click the “start listening” button.   Then start speaking, and your text will appear in SpeakToText.

 

 

Q.  How do I copy the text I dictated in SpeakToText into Microsoft Word?

A.  Say “copy to word” or click the “Copy to Word” button.  If Word is not running, it  will be launched and your text will be copied to a new document in Word.  If Word is running with a document open, the text will be copied into the document at the point where you have placed the cursor.

 

 

Q.  How do I save the text I just dictated?

A.  If you have dictated text into SpeakToText in basic mode, use the “click save” speech command, the save text button, or the save option under the file menu.  If you have copied text from SpeakToText to Microsoft Word, save the text in Word.  If you are working with SpeakToText in advanced mode, save the text in the program into which you have dictated the text.

 

 

Q.  How do I use SpeakToText with other programs?

A.  To use SpeakToText with other programs, you must enter Advanced Mode.  From Advanced Mode, SpeakToText works with most applications that accept text.  In other words, if you can type into a program, you should be able to use SpeakToText with that program as well.  Just place the cursor in the program you want to use, in the place where you would normally type and start speaking.  Your words should appear in the desired program.  To switch to another program or dialog (such as switching between chat program windows)  Just use the mouse and click where you want the text to appear. 

 

Q.  How do I use SpeakToText to chat hand’s free?

A.  Put SpeakToText in Advanced Mode, and put the cursor in the chat box where you want to speak your message.  Speak your message.  When you are done, say “send message” and SpeakToText will deliver a send command to the chat dialog.  If you are using MSN Messenger, AIM Instant Messenger, ICQ, Yahoo Messenger, or PalTalk, SpeakToText will do this automatically.  The send message command also works with Outlook, Outlook Express and Netscape compose email windows. 

 

If you are using a chat or email application other than above, you will need to go to the General tab under settings, and experiment with the available options in the dropdown under the section entitled “When I use the "Send Message" command, I want SpeakToText to:"

 

Note that some chat applications use the “enter” command to send a message.  The “new line” and “new paragraph” dictation commands send the “enter” character once or twice, and if you use these while you are dictating a message in a chat application that uses “enter” for a send command, the message will be sent before you are finished.  Likewise, if you prefer, you can use the “new line” and “new paragraph” commands to send chat messages in place of the “send message” speech command, although this is not recommended or supported.

 

Q.  Can I dictate directly into Microsoft Outlook or Outlook Express?

A.  Yes.  In Outlook, Outlook Express or most other email clients, you can dictate the email subject and body from SpeakToText in Advanced mode.  With Outlook and Outlook express, you can also use the send message feature to sent your message.  Speech recognition typically works better with sentences, and does not work well with names or email addresses. 

CoolSoft does have a program designed specifically for creating and sending email.  Check out Speak-to-Mail at http://www.coolsoftllc.com/main.asp?bm=13#13.  This program reads the contact list from your default email program when it starts.  It uses a template to compose your email, and lets you address the email as well as dictate the subject and body, and make attachments, all by speech.  When you send the email by speech command, Speak-to-Mail populates a message in your computer’s default email program (Outlook, Outlook Express and Netscape 7.x are supported), and sends the message.  Your message is saved in the sent folder of your email program, so all your messages stay in one place.  Speak-to-Mail also includes a natural language model that let's you set up an email with a single sentence.  For example, you can say “send an email to John Smith about Tuesday’s meeting,” and Speak-to-Mail will place the John Smith’s email address in the “TO” field, “Tuesday’s meeting” in the subject field and the cursor in the body of the email for you to dictate your text.

 

 

Q.  What is a user command and how do I create one?

A.  A user command is a powerful tool that lets you create your own speech commands to open files, play music or play videos (open), start programs (run), browse webpages (browse), or make your computer talk back to you (speak).  How to do this is explained in detail in the SpeakToText user manual, but creating a user command is simple.  Just go to the User Commands tab in the settings dialog, click on new, and fill in the fields.  After creating a user command, whenever you speak that command, SpeakToText will process that command.  

 

 

Q.  How do I make SpeakToText read text back to me?

A.  SpeakToText has four read speech commands. 

·         Say “read all” to read all the text in SpeakToText.

·         Say “read right” to read all the text in SpeakToText to the right of the cursor.

·         Say “read selection” to read the selected (highlighted text)

  • Say “read clipboard” to read the contents of the clipboard.  You must copy the text you want read to the clipboard before using this command.  The clipboard must contain text that you copied from another application, not graphics.

 

To stop reading, you need to use the “Stop Reading” button (second button from the right).  There is no “stop reading” speech command because while SpeakToText is reading, the microphone is turned off to prevent the text that is being read from being recognized as speech.

 

 

Q.  How do I learn all the speech commands I need for SpeakToText?

A.  This will take time, but eventually you will become proficient.  Meanwhile, SpeakToText is designed to make it easy for you to find and learn the speech commands you need to run the program and dictate text.

 

·         Say “what can I say” or use the “What Can I Say?” button to open a dialog of all the available speech commands for the program.  This dialog is dynamic, and shows only the speech commands that are available at the time.

·         Say “open manual” or “open quick start” to open the manual and quick start guide.  Both contain a list of speech and dictation commands.

·         Say “open dictation commands” to open a list of dictation commands.  Dictation commands are different from program commands.  Program commands affect the functionality of the SpeakToText.  Dictation commands apply only to the text being dictated.

·         All of the above can also be accessed from the help menu.

 

Each button in SpeakToText has a tool tip with the speech command for that button.  To display a tool tip, just place the mouse arrow over that button.

 

Finally, you can refer to the SpeakToText quick reference guide at:

http://www.coolsoftllc.com/downloads/SpeakToText2_0QuickReferenceGuide.pdf

You need Adobe Acrobat Reader to view this file.  The Adobe Acrobat Reader is available for free at:

http://www.adobe.com/products/acrobat/readstep2.html

 

When you are just beginning to use SpeakToText, it is a good idea to print a copy of the quick reference guide or the commands section of the manual and keep a copy handy for reference. 

 

 

Q.  How do I train the speech recognition engine?

A.  Training is a simple process that takes 20 to 30 minutes, but it’s a critical step to achieving good quality speech recognition.  How to train speech recognition is explained in detail in the user manual.  You can also watch the “How to Setup and Train Speech Recognition” video that comes installed with SpeakToText.  This video can be accessed from the Help menu.

 

 

Q.  I see that SpeakToText has an update feature.  Is this automatic?

A.  No, the update feature is not automatic.  You have to access the update feature manually to see if updates are available.  If you did not opt out of being contacted when you purchased your license for SpeakToText, then typically you will receive an email when an update becomes available.  If you opted out of being contacted by CoolSoft, you can opt-in by sending an email to support@CoolSoftLLC.com with your order number.  Our support staff will adjust your status in our database.  However, you should check the update feature periodically to see if there are any new updates available.  We recommend an interval of at least every three months.

 

 

Q.  How can I use the .wav to text feature?

A.  The .wav to text feature works for one voice at a time and is intended to be used with a wav file that was recorded in the voice of a person who has trained a SAPI speech recognition profile that is currently in use.  The person whose voice is on the recording must have trained this speech recognition profile to obtain a level of consistent accuracy.  For example, if the recording is only in your voice and you have trained a speech recognition profile for your voice, then you can use the .wav to text feature.  You cannot record someone’s speech and get accurate recognition unless the speaker has trained a speech recognition profile on your computer.

 

If there is more than one voice on the .wav file, then SpeakToText will only recognize the portions with your voice accurately.  You can solve this problem by having each person on the recording train a speech recognition profile, and running multiple passes with each person’s speech recognition profile selected. 

 

The .wav file must have been recorded on a personal recorder and transferred to your computer, or recorded directly on your computer.  For best results, use a higher quality .wav file (the minimum is 22,050 KHz), although the closer to CD quality the better the results will be.  Your .wav file can be stereo or mono.  We also recommend breaking the .wav file into five minute or less segments. 

 

To process your .wav file, just drag & drop the .wav file into the SpeakToText dictation window and wait for the .wav file to be processed.  You can also use the file menu or open sound file button to process your .wav file.    However, if another person's voice is on the recording and they don't have a profile trained, SpeakToText may not have good accuracy (if any at all).  STT is not equipped to process 2 voices in the same recording. 

 

 

Q.  SpeakToText disappeared.  How do I find it?

A.  Say “click focus” and if the microphone is turned on, SpeakToText will appear in the foreground.  The “click focus” command works only if the microphone is turned on.  If SpeakToText has been minimized or moved to the background by another application, you can look for it in the task bar at the bottom of your desktop.  Clicking on the SpeakToText tab should bring it to the foreground.  If you still can’t find SpeakToText, try starting it again.  If it is already running, then the second copy will inform you of this and close down.  If this happens and you still cannot find SpeakToText, then use the task manager to find it.  Press the Ctrl, Alt, Del keys at the same time to open the task manager.  Highlight SpeakToText and click “switch to” and that will bring SpeakToText to the foreground.

 

 

INSTALLATION AND REGISTRATION:

Q.  I have downloaded the trial version of SpeakToText and installed it.  I speak and it is not working.  What is wrong?

A.  The 30 day trial of SpeakToText 2.0 does not come with the speech recognition engine, and to use it you must have the Microsoft SAPI speech recognition engine installed on your computer. Speech recognition is a valuable tool, and the speech enabled version of SpeakToText is a large download that uses lots of bandwidth.  We have decided to reserve the speech enabled version for our paying customers. 

 

You can get the SAPI 5.1 speech engine when you buy the full version of SpeakToText or Speak-to-Mail from CoolSoft, or another third party speech recognition product that runs on the Microsoft SAPI speech recognition engine.  The SAPI speech engine also comes with Microsoft Office XP (SAPI 5.0) or Office 2003 (SAPI 6.1).  If you do not have the Microsoft speech engine on your computer, you can still try all of SpeakToText's features with your mouse and keyboard, and this should give you enough information about the program to determine whether or not you like its features.

 

 

Q.  I purchased SpeakToText from the CoolSoft website.  Now what do I do?

A.  First, if you have a spam filter, make sure you can receive email from the CoolSoftLLC.com domain. 

 

Print the confirmation page at the end of your purchase.  It contains information about your order and we will need some of that information to help you if you have a problem downloading the program or if you do not receive your confirmation email and registration key.  From the confirmation page, click on the download link.  It will take you to the download page, and auto-populate the email address and order number fields.  From here you can enter your download page and download the CoolSoft products from your order.  When the download is finished, double click on the file to start the install and follow the instructions.   

 

You will receive a confirmation email with your order number.  Save a copy of this email.  You will need the email address and order number from this email to obtain support in the future.  You will also receive an email with your user information and registration key.  Typically, you will receive these two emails almost immediately, but the email with your registration key could take up to 24 hours. 

 

Once installed, SpeakToText will function in trial mode for up to 30 days, so you can start using it right away even if you don’t receive the email with your registration key immediately.  We recommend registering SpeakToText as soon as you receive your registration key by email.  Just click on the “Register” button in the SpeakToText trial dialog and follow the instructions.

 

 

Q.  I purchased SpeakToText on CD from a reseller other than CoolSoft.  Now what do I do?

A.  Put the CD in your computer, and it should auto-run.  If not, then browse to the CD, and click on the STT.exe file.  The CD will start.   You can install SpeakToText from there.  It will function in trial mode for 30 days.

 

Next, if you have a spam filter, make sure you can receive email from the CoolSoftLLC.com domain. 

 

On the back of your CD sleeve, there is a sticker with a CD key and a link – typically http://www.CoolSoftLLC.com/register.  Go to this link, enter your CD key exactly as it appears on the back of the CD sleeve, and enter the rest of your user information.  Completing and submitting this page will enter you in our system, and you will end up at a confirmation page.  Print this confirmation page.  It contains information we will need to help you if you have a problem receiving your confirmation email and registration key. 

 

You will receive a confirmation email with your order number.  Save a copy of this email.  You will need the email address and order number from this email to obtain support in the future.  You will also receive an email with your user information and registration key.  Typically, you will receive these two emails almost immediately, but the email with your registration key could take up to 24 hours.

 

Once installed, SpeakToText will function in trial mode for up to 30 days, so you can start using it right away even if you don’t receive the email with your registration key immediately.  We recommend registering SpeakToText as soon as you receive your registration key by email.  Just click on the “Register” button in the SpeakToText trial dialog and follow the instructions.

 

 

Q.  I am installing SpeakToText, or the trial version of SpeakToText, and I am getting an error message, “The file “C:\Program Files\SpeakToText v2 (trial)\UNWISE.EXE” could not be opened.  Please check that your disk is not full and that you have access to the destination directory.  The system cannot find the path specified.  What just happened, and how can I install SpeakToText?

A.  CoolSoft uses the Wise Installer to install and uninstall SpeakToText.  If you have previously installed a copy of SpeakToText, it must be uninstalled before installing it over again, and this is a documented issue with the Wise uninstaller.  It means a copy of SpeakToText or the SpeakToText trial version was previously installed on your system, and the UNWISE.EXE file was called after being uninstalled.  If you run the install again it should complete normally the second time.

 

 

Q.  When I try to register STT, I get a message that says "failed registration key."  Why can’t I register SpeakToText?

A.   

Note that the registration key and user information (first name, last name and email address) are case sensitive, and must appear in the registration dialog exactly as they appear in the confirmation email with your registration key. 

 

 

Q.  What is the difference between the registration key and the CD Key?

A.  To run past the 30 day trial period, SpeakToText must be registered.  All licensed copies of SpeakToText require a registration key.  If you buy SpeakToText directly from CoolSoft as a download, you will get a registration key automatically from our key generator.  This registration key is generated on the fly, and is personalized based on your first name, last name and email address. 

 

If you purchase SpeakToText on CD from a reseller, there is a CD Key on the back of the CD sleeve.  You need this CD key to register with our website so our key generator can send you your personalized SpeakToText registration key as it does with our own customers.   This process also registers you in our support database, so you are eligible to submit a request for support on our support page.

 

 

Q.  I am using the CD key on the back of my CD sleeve, but I am getting an invalid key message.  Why doesn’t my key work?

A.  The CD Key is case sensitive, and must be entered exactly as it appears on the back of the CD sleeve.  A CD key can only be used once.  If you or somebody else has already used the CD key, it will not work again.  Here are some possible reasons why you CD key doesn’t work, and what to do about it.

·         Chances are that you probably entered a character wrong or missed a character.  This is the most common problem we experience with CD keys.   Try again, and double check and make sure you are careful not to transpose the letters O and L with the numbers zero and one.  

·         It’s possible that you didn’t fill in all the required fields on the page. 

·         If you are trying it for a second time, it is possible that it worked on the first try and you didn’t wait long enough for your confirmation page.  Check your email for the confirmation email and registration key email.  Make sure your spam filter allows email from the CoolSoftLLC.com domain.  If  you receive one or both of these emails, it means the activation of your CD key was successful, and there is no problem.

 

If the problem persists, please send an email to support@CoolSoftLLC.com and our support staff will be glad to assist you.  Please be sure to include your CD key, and the point of purchase information in your email, as well as the text of any error message you are getting.

 

 

PURCHASING:

Q.  Does CoolSoft accept Checks or Money Orders?

A.  Yes, we do.  However, you will not be able to download SpeakToText until the money order or check has been received and clears.  We do accept all major credit cards as forms of payment.

 

 

Q.  Does CoolSoft ship overseas?  If so, at what cost?

A. SpeakToText is available by download from anywhere where you can download programs in the USA.  We ship CD’s and Microphones anywhere in the USA and Canada.  If you order a CD or Microphone from a country other than the USA or Canada, extra shipping charges will apply.  We will contact you with the amount of these extra charges and if you agree we need to process the charges separately over the phone or by email.  You will need to give us your credit card information again, as we do not save it when your purchase is complete.  If you do not approve of the additional shipping charges to your location, we will refund only the portion of your order that we did not ship.  There is no refund on downloads.

 

 

Q.  What is the CoolSoft refund policy?

A.  If you purchased SpeakToText download or CD, or a microphone from CoolSoft, the following refund policy applies. 

 

  • Absolutely no refunds on downloadable products.
  • Any defective CD will be exchanged.
  • All other products, 30 day money back or exchange at customer's option.

 

If you purchased SpeakToText on CD from a third party reseller, you must contact the point of purchase and follow their refund policy.  CoolSoft cannot issue any refunds for products not purchased from CoolSoft.

 

 

Q.  Why buy SpeakToText from a reseller? 

A.  There many good reasons to buy from our resellers.

·         If you don’t need the download right away, or if you want a CD instead of the download.  CoolSoft does not sell the CD separate from the download and the CD is $15.00 extra when you get the download.

·         Price – Our resellers buy CD’s from our distributor at wholesale prices.  Sometimes they may offer better prices than CoolSoft, or may bundle SpeakToText with a microphone for a better overall value.  It’s also a good to purchase a microphone, and receive both SpeakToText and the microphone at the same time.  You will have better results if you start using SpeakToText with a good quality noise cancellation headset microphone.  So, unless you need the download right away, buying from a CoolSoft reseller may be a good option for you.

·         You get the same quality product and support as if you purchased directly from CoolSoft.

 

 

Q.  Is there a list of CoolSoft resellers?

A.  No.  Resellers of SpeakToText vary from time to time.  There are always new resellers and some resellers only offer CoolSoft products at certain times.  Sometimes, copies are available on eBay.com or Amazon.com.  To find a CoolSoft reseller, search for SpeakToText or Speak to Text.  If you don’t find a reseller, you can always purchase SpeakToText from CoolSoft. 

 

 

COMPATIBILITY:

Q.  Is STT supported in Windows 98 or ME?

A.  No.  SpeakToText was designed and developed to function only in only in Windows 2000 and Windows XP (Home and Professional).  Our development team looked at supporting Windows 98se/Me and determined that because of the way these older operating systems interact with SAPI, it would require changes to the speech layer of SpeakToText that could possibly degrade performance in Windows 2000 and XP.  We decided not to take this risk because very few people still run Windows 98se/Me, and to our knowledge Microsoft has discontinued support for the former. 

 

 

Q.  Can I use SpeakToText if I have an accent?

A. This varies from person to person.  You need to train your speech recognition profile and try.  Most users with an accent are able to achieve a reasonably high level of accuracy.  We have many users from the UK, Australia and New Zealand who are quite pleased with the results, and several members of our development team have a French accent and are also successful with the program.

 

 

Q.  Is SpeakToText supported in languages other than English?

A.  No.  SpeakToText works only in English.  Unfortunately, Microsoft only offers the SAPI 5.1 speech engine in English and Chinese at this time.  If Microsoft releases speech engines in other languages, then we may consider developing a multi-language edition of SpeakToText.

 

 

Q.  Is there a 'list' of applications with which SpeakToText works? 

A.    We don't have a list, but we have yet to find a program that you can type in that does not also work with SpeakToText.

 

Q.  Is STT compatible with MAC computers?

A.  No.  SpeakToText works only on Microsoft operating systems, Windows 2000 and XP, with Microsoft SAPI 5.0, SAPI 5.1 and SAPI 6.1. 

 

 

TROUBLESHOOTING:

Q.  I am dictating with SpeakToText.  The text I dictate shows up and disappears immediately.  What is causing this?

A.  You have background noise filtering enabled, and the filtering is too strict.  You can adjust Background Noise Filtering from the Formatting Tab of the Setting Dialog box.   If the speech engine displays and rejects words that you have spoken (you see the words appear and disappear from SpeakToText) then move the slider to the left (looser filtering).

 

 

Q.  I am dictating with SpeakToText.  When I am not speaking, it is recognizing words that don’t make sense.  How can I fix this?

A.  Because of the environment you are in, SpeakToText is detecting and recognizing background noise as speech.  You need to enable Background Noise Filtering from the Formatting Tab of the Setting Dialog box.   Start with the slider in the middle and experiment to find the optimal level.  If filtering is too loose (slider to the left) you will continue to have unwanted input from background noise.  If the filtering is too strict (slider to the right) SpeakToText may reject words you have actually spoken.  If this is happening, the words appear then disappear immediately. 

 

 

Q.  I am not getting good accuracy.  Is there anything I can do to improve accuracy?

A.  There are many factors that affect accuracy.  Try these steps to improve your accuracy.

1.                  Make sure you have trained your own speech recognition profile at least three times, and that you have that recognition profile selected as the default.  You can do this by accessing the speech properties dialog from SpeakToText.  Before training, you should position the microphone boom about an inch from your mouth, and configure the microphone.   Configuring the microphone is also done from the speech properties dialog.  

2.                  Each time you use SpeakToText, you should make sure the microphone position is consistent to the position you used in training.  You should also configure your microphone at the start of each session.

3.                  Use a high quality noise cancellation headset microphone that has been designed for speech recognition.  The Andrea NC-91 available from CoolSoft http://www.coolsoftllc.com/main.asp?bm=10#10 is a good choice.  Desktop microphones do not typically provide good results as the microphone position to your mouth will typically be inconsistent.

4.                  Make sure you are in a quiet environment without too much background noise.  Background noise can have an adverse affect on your results.

5.                  Make sure your system meets the minimum system requirements for SpeakToText as indicated in the user manual.

6.                  Speak clearly and at a normal pace, and pronounce your words clearly.  Do not speak too quickly, and do not pause between words.  The speech recognition engine works best when speaking sentences normally.  For a good example, listen to the news on radio or TV and try to imitate the newscaster’s speaking style.  

 

 

Q.  Why is SpeakToText making mistakes?

A.  Try the steps for improving accuracy in the question above.  Accuracy will vary from user to user because of differences in the speaker, training, computer resources, the environment and many other factors that affect accuracy.  In fact, its normal for two computers in the same room with the same speaker following all the required steps for accuracy to produce different results from the same reading.  Just as you make mistakes when you type, the speech engine makes mistakes in recognition.  This is normal functionality.

 

Once you have maximized accuracy for your individual factors, using speech recognition effectively requires modifying your habits.  When most people type, they still make mistakes and they typically fix their mistakes and format their text as they type.  With speech recognition, it is best to finish your dictation and then go back and correct any recognition errors and add formatting.  Once they get used to this, typically most people find their productivity has improved. 

 

In one study, it was determined that the average person types 35 words per minute with a 5% error rate.  The average person using speech recognition can achieve 50 words per minute at the same 5% error rate.  Although these are only averages, for the average person using speech recognition this represents a 43% increase in productivity.

 

This FAQ last updated on June 2, 2006.

 
(c) Copyright 2002 - 2009 - CoolSoft, LLC -All Rights Reserved
Last Updated on 10/18/2019