How To Talk To Your Dragon…Voice Recognition Comes of Age

Dragon Naturally Speaking 11 Logo

by Dan Alexander

Not Science Fiction anymore…

Operating your computer by voice is now a mainstream reality with the arrival of Dragon NaturallySpeaking 11 from Nuance.  And this happens straight out-of-the-box without any time-consuming or elaborate set up.

On the average people speak three times faster than they can type. You can easily close that gap using Dragon as a dictation tool. This program also allows you to launch applications, open files, search the web as well as manage and write e-mails. You can even work an Excel spreadsheet hands-free while you sort through your paperwork.

The implications for the disabled and impaired are far reaching.  Voice recognition can enable those who are disadvantaged with more accessible options to everyday computer use.

Of course, to attain the full capabilities of Dragon NaturallySpeaking 11 you will need to familiarize yourself with the array of commands that allow this product to be so versatile. But to the casual user who just wants a break from pecking at the keyboard this program is user-friendly with minimal setup.

In fact, this entire article was dictated with Dragon voice recognition with approximately 10 min. of setup time.

Dragon NaturallySpeaking 11 Dictation



Long Time Coming

Dragon NaturallySpeaking 11 voice recognition software is the culmination of nearly 30 years of research and development. That’s a long time for any piece of software in this day and age.

Voice recognition research actually began almost 75 years ago at AT&T’s Bell Labs.  In later years university and government research joined the race to what was then called automated speech recognition and transcription or ASR. You probably would not be surprised to find out that the gadget spooks at DARPA have their fingerprints all over this technology as well.

But it hasn’t been an easy road getting to this point.

Spock with Early iPad

High Expectations

We have lived with science-fiction characters speaking offhandedly to their computers with perfect compliance for over 40 years.  To wit: “Open the pod bay doors, Hal.”  Well, maybe not so perfect.

The miniature communicators/computers (smartphones) used in sci-fi are commonplace today. And we have those high-powered tablet devices (iPads, etc.) that the science officers employed to their advantage. So, why don’t we enjoy verbal control of our computers?   We could rid ourselves of bothersome things such as typing or pointing and clicking?

Dragon NS 11 and Working with Applications



“I’m giving her all she’s got, Captain!”

Resources and Dragon NaturallySpeaking 11

Creating voice recognition technology to enter content, open and manipulate programs, fill in data fields or surf the web isn’t just writing a software application. It’s more on the level of getting your cursor to move accurately with telekinetic commands.

Let’s take the English language as an example. Think of the differences in spelling and sound when English is used by people from Great Britain, India, Southeast Asia or the United States. Let’s drill down to the United States and think of how English differs between people that live in the Deep South, New England or the Midwest.

Also consider non-natives in America that speak English with Hispanic, Asian, Indian or any number of accents. Now consider the difference between a 10-year-old girl from California and 78-year-old male from Maine.  Now factor in people who speak fast or slow; loud or soft.

We are talking about an enormous amount of variance for the words spoken in a simple sentence such as, “The quick brown fox jumps over the lazy dog”.  In turn, we have to consider an enormous database from which to draw these distinctions and variations.

Before 2000, 2.5 GB for a program install was a sizable chunk of storage capacity for hard drives  . Today most computers sold will have a minimum of 500 GB trending up to 1+ TB of disk space.

Processor speed and memory have slightly outpaced Dragon’s voice recognition technology. To function quickly and seamlessly NaturallySpeaking 11 needs some horsepower to drive it. Luckily, unless your machine is old and clunky, you should have no problems running this program.

Nuance states an absolute minimum of 1 GHz Intel Pentium CPU, 512 KB of cache and 1 GB of RAM. Again, these are absolute minimums. You can find recommended system requirements here.

Once more with feeling

It’s been about eight years since I last took NaturallySpeaking for a test drive. It was pretty amazing back then, but the accuracy and fluidity of the technology wasn’t quite there. There were more than a few occasions when the output would elicit a chuckle after it finally displayed.

Recently I came upon Dragon Dictation for iPhone app and decided to try it out. If you have an iPhone I recommend you download this free app.  It will give you an idea of the accuracy of Dragon’s voice recognition technology. Be sure to take a peek at it’s sister app, Dragon Search.

Dragon Dictation for iPhone allows you to text your dictation, e-mail it or copy it to another location. You can also link Dragon dictation to your Facebook or twitter accounts. Dragon dictation is available for the BlackBerry as an e-mail application.

Nuance, the makers of Dragon Naturally speaking 11, produce products for both Mac and PC. They have medical and legal editions that are specialized for those professions and they also offer three other editions for varying degrees of use. The software begins at $99 and includes a Nuance approved headset.

To see a comparison of all editions click here.

So, there is a new interface in town that will only further the ease of interaction between human and machine.

And they call him Dragon.

Related posts:

  1. Who REALLY Owns The 100 Most Popular Sites on Digg?

Comments on this entry are closed.

Previous post: