Dragon Naturally Speaking 9 – Voice Recognition Accuracy at 99% Out of the Box

A recent article in PC World(Australia) claims that Nuance claims breakthrough on speech recognition. It looks like Nuance is trying to continue their lead in the voice recognition market in their latest version of Dragon Naturally Speaking 9. The article says it will ship to Australia in August, but it is available immediately through Nuance’s global network of reseller partners, software retailers and professional sales organizations according to their recent press release.

Here’s the good information from the article.

Nuance Communications says the latest version of its speech-recognition software can achieve — with some speakers — 99 percent accuracy out of the box, without a “training” session to familiarize the software with how a particular person talks.

The accuracy rate, or what percentage of words the software spells correctly by itself, varies depending on sound quality and how a person talks, Revis said. But Nuance has improved it by 80 percent since NaturallySpeaking 8 was introduced in 2004, according to the company.

Version 8 could reach 99 percent, but only after the user read a prepared script, Revis said. Now users can get that level of accuracy right after installing the software and starting it up, though a script is still available if a user isn’t satisfied with the results on the first try. In any case, the software can continue learning on its own just through normal use, Revis added.

The out-of-the-box 99 percent accuracy figure is for the American English version, but the new Dragon releases for other languages get similar boosts in accuracy, Revis said. The software is available for Dutch, French, German, Italian, Japanese and Spanish, as well as for Australian, Asian, Indian and U.K. English. Also with NaturallySpeaking 9, Nuance is certifying two Bluetooth wireless headsets for use with the software.

Nice to expand the product internationally. I can only imagine what new technology Nuance has come up with to be able to recognize out of the box this many languages. It would make sense if they only had listed latin languages. Howevever, Japanese, Asian(what language is that) and Indian is definitely not anything like latin languages.

A few other interesting notes:

-Dragon Naturally Speaking 9 can be used anywhere on the network including at thin clients
-New support for Nuance-approved Bluetooth headsets
-The Upgrade cost is usually a know brainer if you use Dragon Naturally Speaking 8
-The RAM and CPU requirements are of course higher
-Still requires Windows XP

Call me a skeptic, but my feeling is that despite the increase in accuracy available out of the box with Dragon Naturally Speaking 9, you are still going to want to do the “training” or “enrollment.”.

About the author

John Lynn

John Lynn

John Lynn is the Founder of HealthcareScene.com, a network of leading Healthcare IT resources. The flagship blog, Healthcare IT Today, contains over 13,000 articles with over half of the articles written by John. These EMR and Healthcare IT related articles have been viewed over 20 million times.

John manages Healthcare IT Central, the leading career Health IT job board. He also organizes the first of its kind conference and community focused on healthcare marketing, Healthcare and IT Marketing Conference, and a healthcare IT conference, EXPO.health, focused on practical healthcare IT innovation. John is an advisor to multiple healthcare IT companies. John is highly involved in social media, and in addition to his blogs can be found on Twitter: @techguy.


  • I think Dragon, and voice recognition in general, is great if it can help bridge the gap between paper and electronic by avoiding transcription, but…

    An anecdote told by a physician highlights my concerns. One day, he added a note that he thought said “Patient is fine, just continue with the meds”. Dragon transcribed “Patient is fine, discontinue the meds”. He was finished with Dragon. His thought was that if he had to review everything and make corrections, he could type it in just as fast or faster.

    99% is good. A human life needs 100%. Dragon, and voice transcription as a technology, is not there yet.

  • How to install on the same computer 2 versions of Naturally Speaking V9 Preferred:
    – the first one in French recognition
    -the other one in English (american english and other dialects if possible).
    Currently I can’t install these two versions on the same computer: intalling one of them erases the other.
    I succeded one time such an installation but can’t do it again.

  • I would like info on how a deaf person could use the equipment. Since most people do not sign, maybe your equipment could signiticantly help communication by having the non-deaf person speaking into the microphone and the deaf person could read the message on a screen. The deaf person would type his message back.Since this would be more normal for the non-deaf/non-signer and much faster itwould encourage many more communication partners for people with hearing loss, including total deafness and for whom hearing aids are not of help. It would be a wonderful tool to help them in society, instead of being so isolated.

    My question is: how many individual voices can the equipment handle? If not many, then at least the deaf individual could communicate with a few if the people most likely to be living with them, like a spouse. Or if he is school age, a teacher or tutor, and his family.

    If it worked well, the deaf person would be much more employable.

    Is anyone using the Dragon Naturally Speaking 9 in that way? If so, could you put them in touch with me?

    Preferably, the equipment could take any human voice ( a clerk at a store, a doctor or nurse during an office visit, etc.) The deaf person could carry a small microphone around with him and give it to the person with whom he wants to communicate. He’d have to have his lap top with him, of course. But as you develop your equipment, it could get smaller and smaller. Like an I Phone, with a pencil-sized micro phone attached. Or maybe just a pull out antenna kind of thing.

    I’m eager to hear from you.


    Looking forward to your answer

  • Ruth,
    I honestly haven’t heard of anyone using it that way, but I don’t see why it couldn’t. Just use a nice little tablet and carry the tablet around with you and have a microphone on it. Then, you could just hold it up to a person’s voice and have them speak on it.

    The main issue is probably with having multiple people speak into it. If you are the one speaking, then it will work fine for someone else to read something you’ve said. However, the software is designed to learn about a person’s unique voice characteristics and tendencies. By using multiple people, that would be a problem.

    I hope that nuance happens upon this post and considers using their technology to support the deaf community. It could really be a pretty amazing product since they claim that it is 99% out of the box. They would just need to design it to support multiple voices. The technology is getting there, but it’s about Nuance supporting that use of the technology.

    Best of luck and if you try it please let me know how it goes.

  • I have Dragon Naturally Speaking Preferred. Do EMH programs require the Professional or Medical version?

  • Al Brent,
    I assume that you meant EMR programs and not EMH programs.

    I’m sure that preferred will work with almost all EMR programs. The fact of the matter is that Dragon is really just typing the letters similar to the way a keyboard might enter the letters. Just speaking is faster than typing for most people.

    The one caveat is that if an EMR vendor has included deep integration with dragon, then you might need to get the professional or medical version. That would depend on the EMR. The point being that one EMR that says they support voice recognition and another one that says they support voice recognition may not be apples and apples. There are different levels of dragon naturally speaking integration that are available.

  • I currently use dragon medical with vocab upgrade of two specialties within my EMR. I use Practice Studio and quickly learned there were incompatibility issues with Dragon and the EMR. There is a green light portion of dragon ….meaning dragon works well and all the bells and whistles of dragon program will run. If that green light is not on…..as in Practice Studio…..then the accuracy is significantly less. Dragon is very cool and does take a lot of time to coach to your voice. I am probably more frustrated with Practice Studio EMR than dragon but as others have stated……proofreading your Dragon dictation is a must for “breathe” vs “breed” errors.

Click here to post a comment