13120 I've been delighted, blown away even, with MacSpeech Dictate 1.3's performance in use over the past several months. I was a big fan of its predecessor (but not ancestor) MacSpeech iListen, but Dictate is so much better in so many ways that there's nothing at all about the older program I miss. iListen was pretty good dictation software. Dictate is great dictation software.

Given the degree of satisfaction Dictate 1.3 has provided, MacSpeech's announcement that they had found a way to squeeze out up to 20% better voice recognition accuracy out of version 1.5, which was released a little over a month ago, that it seemed sounded too good to be true, but while it's difficult for me to gauge the degree of improvement with any scientific rigor, I can vouch for the fact that Dictate is even more impressive than it had been, and the bar was already set very high.

Installing the Dictate 1.5 update, which costs $54.95 for registered users of older Dictate versions, and has a full price of $199.00 with a bundled headset microphone, proved simple and easy.

I inserted the Dictate 1.5 application disc and dropped the program into my Applications Folder. Next I inserted the Data Disk and started the program, which copied the requisite updated files from the disk automatically. Final steps were to update the program registration with this version's new number key, and to read a training story in order to update my voice profile, optimizing it for version 1.5's improved speech engine.

As with the previous versions of Dictate, it's necessary to have a live Internet connection in order to complete the registration process - a point that is mentioned on the program's packaging, but easy to miss in the fine print. This should only be an issue for a very small minority of potential users, but it would be nice if there were some sort of workaround available. Once registered, Dictate works fine when you're not online.

Reading the training story took me about seven minutes, the program took another several minutes to process the new voice data, and we were good to go.

And go it does. Even on my 2.0 GHz MacBook with 2 GB of RAM, a relatively modest spec. power-wise these days, Dictate just flies, transcribing spoken input in near real-time. I've been intending to upgrade my RAM to 4 GB, but the little MacBook has performed so well with the standard 2 GB that I haven't had much incentive to stop procrastinating.

In addition to the improved speech engine, MacSpeech CEO Andrew Taylor says that Dictate 1.5 also includes some "under the hood" refinements such as a new vocabulary editor, improvements in compatibility with certain applications (including Microsoft Word), and some minor modifications to the program's interface and documentation.

The vocabulary editor makes it easier to train dictate to recognize esoteric and unusual words easily - even abbreviations such as those used in tech writing.

When using Dictate 1.5 with Microsoft Word, automatic execution of the "Cache Document" command is now supported when Dictate is launched, and also allows Dictate to read a document's contents, navigate the document, and enter or edit text.

There are an additional 100 websites supported by Dictate 1.5's "Jump To" command, and the new Save Profile menu item for capturing phrase training information prior to quitting the program, as well as an upgraded updated user manual.

This is all gilding willingly, as it were, but Dictate's main virtue remains its speed and uncanny accuracy even with a minimum of training.

At the end of May, MacSpeech also announced a Dictate 1.51 maintenance release that is downloadable free of charge, and incorporates several fixes for reported bugs, and some 24 hours after this review was originally posted, the MacSpeech Dictate 1.5.2 update was posted - a maintenance release that fixes reported issues to how Dictate works with existing applications. Most notable with the upgrade to Safari 4, versioned commands now appear for the newer versions of the same application.

The MacSpeech Dictate 1.5.2 update is available free of charge to registered MacSpeech Dictate 1.5 customers using the "Check for Updates" feature.

See the release notes in the Appendix to this review.

Dictate has four modes of speech recognition: Dictation Mode, Command Mode, Spelling Mode, and Sleep Mode all switchable using your voice alone. Dictation Mode is self-explanatory but while you're in it you can also issue commands, In Command Mode the program recognizes and executes only commands - thus it can be looked on as "Command/Only Mode."

Spelling mode allows users to spell words, names, or acronyms by using either natural language or the International Radio Alphabet. Sleep Mode simply lets you toggle the microphone and application on and off.

Phrase Training helps you increase accuracy by letting you train your voice profile as you proceed, helping refine Dictate's ability to recognize what you intended to say. With phrase training you are not stuck with the level of accuracy from your initial voice training until you take the time to read more training stories, and you don't have to specifically train it from a document or selection. Instead you can train a phrase immediately when you see it was misrecognized. The next time you dictate that phrase or word in the phrase, Dictate is more likely to recognize it accurately.

Spelling mode and phrase training work much the same way. You pause between navigation or spelling/correction instructions and general dictation in order to allow the program to distinguish between the two.

The Plantronics mic headset that comes bundled with Dictate haas proved excellent, equal to or better than any mics I ever used with iListen. it's light, adjustable, attractive looking, and it has a conveniently long cord. The mic does require a USB dongle, which does use up another precious USB port.

For just about any user, Dictate's voice transcription capability will be a dream come true, and it's so accurate that not a whole lot of correction is necessary anyway.

Dictate's Notepad application, according to MacSpeech, will always provide the most optimized results, particularly when executing advanced features, complicated navigation, and massive amounts of editing with your voice, however I've found performance more than satisfactory and my favorite application for composition and editing, Tex Edit Plus.

MacSpeech vigorously discourages using a hybrid mode of partly dictation and partly keyboarding and/or mousing, maintaining that alternation between speech recognition and keyboard/mouse is "a fantastic way to confuse the software." I've been trying to break myself of the habit, and I perceive that Dictate's very lively responsiveness is going to help greatly with that effort, but thus far I still tend to get impatient and resort to the keyboard and mouse selection for editing and correction.

Dictate will work with any document, including ones that were not created with Dictate. The program still takes a relatively long time to start up, which I suppose is inevitable when it has to load a voice profile and dictionaries before you can get underway. Adding more RAM, which I plan to do in the not too distant future, or a 7200 rpm hard drive, which I don't, would probably speed things up some more.

MacSpeech Dictate has the following system requirements:

• An Intel based Macintosh computer
• Mac OS X 10.4.11 "Tiger" or later
• Mac OS X 10.5.2 "Leopard" or later
• 512MB of RAM (1GB or more for better performance)
• 2.0GB of available hard drive space

MacSpeech Dictate sells for $199 (including the Plantronics headset), and is available from the MacSpeech store at:
http://www.macspeech.com

------–

Appendix

MacSpeech Dictate 1.5.2 Release Notes
Fixed:
• Versioned commands now appear for newer versions of the same application
• This was most notable with the upgrade to Safari 4.

MacSpeech Dictate 1.5.1 Release Notes

Changes:
• Older licenses are now identified as "Older"
• Changed the text in our License dialogs for older licenses to reduce confusion for customers
Fixed:
• UK Locale Format issue
Localized UK user would crash when profile was created.
• Creation of a profile when focus was last in TextEdit caused crash
• Crash after altering some kinds of documents (e.g. Excel)
• Editing words with hyphens causing crash
• Status window and Available Commands window overlap
On first launch the Status window and Available Commands windows would overlap.

MacSpeech Dictate 1.5 Release Notes

New Features:
Better, Stronger, Faster speech recognition
Vocabulary Editor
New Commands
• Web 100 Commands
• Activate / Launch / Quit / Kill / Exit for applications
• Improved organization of Dictation group commands
Microsoft Word and TextEdit text control enhancements
Save Profile menu item
Dictation target info in Status window

Additions:
• New revision of the speech recognition engine.
Increased speed and accuracy.
• Vocabulary Editor.
Provides the capability to Add/Delete, Manage, and Train words for a profile's vocabulary.
Customization of word behaviors (e.g. spacing before and after, capitalization within titles, and pronunciation).
• New Dictation Group in the Available Commands window.
Within the Available Commands window, there is now a group called Dictation with subgroups entitled Capitalization Commands, Text and Editing and Movement, Punctuation and Special Symbols.
• Web 100 subgroup with global "Jump To" commands for the top 100 websites.
Added a set of commands for the Top 100 websites based on information found at Alexa.com.
• Extended Application Commands
Added Activate / Launch in addition to Open, as well as Quit/ Kill / Exit to the Application Commands subgroup. Only Activate and Quit are shown but the other terms will function.
• Microsoft Word text control enhancements.
Open a Microsoft Word document and Dictate will automatically run a "Cache Document". The document is then ready for dictation and voice navigation.
• File/Save Profile menu item for those that wish to save before Quit.
Added the ability to save Profile before quitting Dictate.
• Help buttons
Help buttons are now displayed on interfaces for Profiles, Microphone Setup, Voice Training, Vocabulary Training, Vocabulary Editor, Commands window, Preferences.
• Speech Mode button to the Status window
Graphical display of the current speech mode and provides alternative method to switch modes.
Changes:
• MacSpeech Dictate 1.5 profile names now end in .dictateprofile
Profiles previous to Dictate 1.5 ended in .dictationprofile

Known issues with MacSpeech Dictate 1.5

• Vocabulary Editor: Advanced word options are not consistently exported. Advanced word options (e.g. spacing, capitalization, and title regulation) may not persist when words are exported. These settings will need to be re-added after import.

• Dictation into some third party applications may still yield some unpredictable results. This is due to the many different ways that applications handle text. MacSpeech will continue to address these issues over time. To report a problem with a third party application, please send an email to

• There are likely to be dictation and editing issues when using a list or outline mode with various word processors. Typically in these modes, word processors insert additional tabs, spaces or other characters into the document. MacSpeech Dictate may not be aware of this added text, so attempting to alter the dictated text can be problematic.

• Under certain circumstances, MacSpeech Dictate can become out of sync with the text on the screen. This happens most frequently in third party applications, but can also happen in MacSpeech Dictate's built-in Note Pad. To avoid this issue edit by voice or by using the keyboard and mouse, but not both. For more details on selecting and editing text by voice, please see MacSpeech Dictate Help under the Help menu, or the MacSpeech Dictate User Manual.

• Use of the Command "Scratch That" in the Recognition window will not remove the entire utterance. It will remove a single word at a time. You may also use "Scratch Word" or "Forget Word" in the Recognition window.

• After re-formatting general dictation, if you delete a word by voice a leading space may remain. Use "Scratch Word" to delete the space.

• Importing commands may cause duplicates of commands to be displayed. Restarting MacSpeech Dictate will resolve this issue.

• "Cache Document" Command and "Unknown elements detected in this document" Warning Message:

Under some circumstances when executing the "Cache Document" command, MacSpeech Dictate may encounter a document that it can't read and for which it can't create a correctly formatted cache . Under those circumstances, due to the erroneous information in the cache, MacSpeech Dictate may not move the cursor to the correct place in the document. When such a navigation error occurs is dependent upon where the uninterpretable items in the document are encountered. These uninterpretable items are: pictures, hyperlinks, multi-byte characters and other "non-text" items that can be placed in a document.

If MacSpeech Dictate encounters such items while executing the "Cache Document" command, it will present you with a dialog box warning you of this problem. At this point MacSpeech Dictate will suggest that if you wish to, you can issue the "Fix Document Elements" command to correct the problem. WARNING: DATA LOSS MAY OCCUR IF YOU ISSUE THIS COMMAND! You should close the document and make a copy of it and then attempt this command on the copy of the document.

The "Fix Document Elements" command collects the document contents, removes the uninterpretable items, does a "select all" and "paste". This removes those items from the document. If all that is encountered are multi-byte characters such as 'é', then your data loss may only be any formatting, such as "bold" or "underline" or font changes. If there are pictures or hyperlinks or other non-text items in your document, then they will be lost.

After issuing the "Fix Document Elements" command, then you need to issue another "Cache Document" command and see if MacSpeech Dictate is able to correct cache the document. If you do not receive another warning dialog, then MacSpeech Dictate should be able to correctly navigate in this document.

In the event that you wish to continue without issuing the "Fix Document Elements" command, then you may not be able to navigate correctly within the document by voice. You should only add to the beginning of the document or to the end of the document.

To reiterate:
- During the "Cache Document" process you may receive a warning message regarding unknown elements in the document.
- Close the document, make a copy and open the copy. Issue the "Fix Document Elements" command.
- Issue the "Cache Document" command.
- Examine the results.

OR

- Continue on, but do not expect to reliably navigate within the document. You can still add to the beginning or to the end of the document.


or additional information, including a knowledge base of technical information and answers to frequently asked questions see the Support Center on the MacSpeech.com web site.



Charles W. Moore

Tags: Hot Topics ï Reviews ï Writing/Publishing ï

Login † or † Register † †