A Dragon That Takes Dictation and Controls a Mac by Voice

I am writing these words—without touching a key. I’m dictating them into a word processor on a laptop. To do this, I’m using a very versatile program that not only allows me to dictate text but to correct it, delete it, and format it, all with my voice.

This software enables me to perform other tasks on the computer by just talking. I can launch and close applications. I can search the Web and jump directly to Web pages using only my voice. I can address, compose and send emails. And I can even dictate and post status messages to Facebook and Twitter.

The product that’s letting me do all this is the latest in the software line called Dragon. In particular, I’m using the newest Dragon dictation software for the Mac, called Dragon Dictate 3. Dragon’s maker, Nuance Communications Inc., has for years focused on the Windows platform. In fact, it released a new version for Windows, called Dragon NaturallySpeaking 12, in July. But with Dragon Dictate 3, which launched last month, Nuance has finally brought its Mac version nearly up to speed with the Windows version.

image

Dragon Dictate uses a small window to show its status and often displays two larger windows, one with available commands and one with alternate spellings

Despite some feature differences, the two versions use the same improved voice-recognition engine, so the company said my experience on the Mac would be a good indicator of its Windows product’s performance. I chose to test the Mac version because it can finally do some things formerly limited to Windows, such as selecting a single word by voice for correction or deletion or formatting. Based on my tests, I can say that Dragon Dictate 3, and by extension Dragon NaturallySpeaking, are quite accurate.

However, this is old-school software. By that I mean it’s relatively costly, at $200, and requires time to learn how to use. It could take weeks or months to remember and master the specific wording of each of the many commands. For example, you need to say “File Open” as it won’t respond to “Open File.” So you’d need a good reason to make that investment in money and time.

The software would be a good choice for people who are disabled and can’t type, or even those prone to mild repetitive stress injuries.

So why choose a $200 program? Especially since Apple recently added free built-in dictation to the Mac. First, Apple’s dictation doesn’t allow you to command the computer, or edit by voice. Macs do have a feature that make a limited number of spoken computer commands available. But they don’t match Dragon’s voice command set. You also can’t use Apple’s dictation feature unless you’re connected to the Internet, while Dragon works right on the computer.

The Dragon software has many, many functions (the user manual is 207 pages), but you can boil it down to two: dictation and computer control. The latter means controlling menus and commands either common to the whole operating system or specific to a program you’re using. You can say “File New” to open a new document or “Jump to Wall Street Journal” to open the Web browser and go to the Journal’s website.

One of the deficits still lingering in the Mac version is it has fewer sets of application-specific commands than its Windows counterpart. Also, the Windows version allows you to navigate from link to link on a Web page; while the Mac version doesn’t.

Dragon comes with an over-the-head microphone and earpiece that connects instantly via USB, but it also works with many types of microphones. The included gear offers the greatest accuracy. But I was able to get decent results using a Dragon wireless-mic app on my iPhone and even using the built-in mic on the Mac laptop, either directly or via the tiny mic included in Apple’s standard earbuds. You can also use a Bluetooth microphone.

This new Mac version finally is able to transcribe audio files recorded on a digital voice recorder or an iPhone voice-recording app. It worked pretty well, but wasn’t as accurate as real-time dictation.

For best results, you have to train Dragon Dictate before using it, by reading one of several canned texts. This takes five to 10 minutes or so. You can teach the program about uncommon words you use by letting the program analyze documents you wrote. And you can add words or pronunciations manually.

The program relies on context to decide what you meant to say, even if the words sound the same. In a test exercise suggested by Nuance, Dragon Dictate flawlessly handled the phrase “you were right to write me right now.”

I found it could handle many—but not all—proper names, product names, long words, addresses and phone numbers. It recognized the sentence “I’m dictating this on a MacBook Air, but I could be using a Lenovo Ultrabook” perfectly, with correct capitalization. It automatically formats phone numbers and addresses. Other sentences it handled with ease included “Barack Obama is running against Mitt Romney” and “I prefer chrysanthemums to hydrangeas.” But it stumbled on some names, like “Kara,” which it interpreted as “Camera” or “terra.”

And it made other errors, some of them simple, like misinterpreting “an” for “in,” even if it got that right most of the time. Fortunately, if you do see a mistake, you can just say, “Scratch That,” and try again, or you can choose an alternate spelling from a window that pops up.

And there are other downsides. For best results, the company suggests you use Dragon in a fairly quiet place, speak in complete sentences and phrases, and think ahead to what you want to say. Meeting those conditions isn’t easy.

Also, Dragon can clutter up your screen. It has a small window showing its status, and often displays two larger windows, one showing available commands and one showing alternate spellings.

Finally, on the Mac version, one of its best features—the ability to learn the names of your contacts—is crippled by an arbitrary limit of 300 names. The feature won’t work at all if your contact list is larger. The company says it plans to fix this.

Overall, however, Dragon Dictate is a step forward for Mac users who need, or prefer, to use voice to write and control a computer.

Tags:

About Nlyten

I have always been a tech enthusiast, to the point where i have become an addict. Tech to me is crack; Always trying to get my fix every chance i get ! I have always loved sharing anything that fascinates me which again 90% of the time is about tech related content. I used to share content on Google Reader Shared pages (http://www.google.com/reader/shared/surdie) but after their not so brilliant idea to shut it down i felt i needed a new platform where i could share and distribute content and thats how Nlyten.com came about. So keep reading and get Nlyten ed !

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: