The geek in me has always enjoyed technology that made my life easier. Especially now in law school, time is of the essence. After reading a few cases the night before class, I will usually type out some basic info about the reading, some holdings and takeaway rules. I figure that skeleton briefing three classes per night takes about an hour, and that just provides basic prep for class.
Enter Dragon Naturally Speaking, voice recognition software that comes almost entirely pre-trained, adapts to the way you talk, and includes a dorky microphone that will allow me to get 30 min of extra sleep per night. I had been really excited to try an earlier version of Dragon (8/9?), but I gave up during training as I kept getting words wrong.
You hafta check out this review. David Pogue, the tech editor for the NY Times, has an amazing video demonstration of Dragon 10 via CNBC, conducted entirely by voice with his hands tied behind his back. (NYT text review). And I know I sound like an infomercial, but it takes very little training to get started (thanks in part to scanning emails and word docs for data), probably around 20 minutes. Dragon offers advanced training for accuracy, and can be used to control most Windows programs, as well as Firefox including searching for video/photo, or navigating and clicking on links by voice. There’s also Legal and Medical editions but they’re pricey ($1k), whereas Dragon Preferred 10 retails for ~ $180 – 200 and can support bluetooth.
A cnet review has the following:
[Dragon Preferred] supports commands in Microsoft Word, Corel WordPerfect, Microsoft Outlook Express, Internet Explorer, and AOL. Using Dragon with the Google Docs online word processor was trickier in our tests than with Microsoft Word 2003 or 2007. People with disabilities can mostly drop the mouse and the keyboard, asking Dragon to do the work for them. But if you have the choice, we still prefer manual controls to the tedious attempts at using Dragon to cut and paste chunks of text within a long document.
The more you use Dragon, correcting its errors and adding your own lingo to its vocabulary, the better it gets. It already recognizes everything from “a cappella” to “smiley-face” to “ZZ Top.” Abbreviations and tech slang work, too; speaking “dot ASP” spells “.ASP” and saying “smiley-face” will spell out this character: :-) Dragon NaturallySpeaking 10 has the intelligence to detect words within context. For example, it knows to type “eating a carrot” instead of “eating a karat.”
You don’t need to speak like a robot into the mic, although enunciating helps. If you tend to mumble, then act as if you’re reading a book to a child or a teleprompter for a newscast when using Dragon. Out of the box, the application does very well with long, polysyllabic words. But we’ve found it difficult for Dragon versions 8, 9, and 10 to distinguish between short words with similar vowel sounds, such as “a,” “the,” and “of.”
The big downside is words or names that it has never heard of before, and memorizing all the voice commands. But for what I’m using it for — taking notes that can have a couple of spelling errors in them — it should save me some time. Carpal Tunnel be damned!
Incoming search terms:
- voice recognition software
- voice to text software
- dragon voice to text review
- dragon infomercial
- dragon naturally speaking
- dragon voice recognition software
- dragon voice recognition
- speak to text softwear
- dragon naturally speaking voice to print software
- dragon word processor demonstration
No related posts.
{ 8 comments… read them below or add one }
Oh voice recognition — you’ve come a long way:
http://uk.youtube.com/watch?v=2Y_Jp6PxsSQ
Hey, I don’t have Microsoft Word, and I’m dying to know how DNS 10 works with either Google docs or Openoffice. All I can find is what is in the Cnet review – that it was “trickier” than other doc apps, but what does that really mean? “Trickier” as in inconvenient, or as in, don’t bother?
If you have any experience with either of these, if you wouldn’t mind saying a word or two about it, that would be so helpful.
Peace,
Tom
@Tom, interesting question. My guess is that by “trickier” they mean that you can’t select text by voice and/or apply commands by voice (i.e. “cap that”). If Andy can’t check though, I’ll check for you in a little bit.
Yeah I don’t have either application but Josh’s response sounds right. The text-to-speech works everywhere (notepad, text fields in firefox, ect) but it’d be editing commands and/or controlling windows (save/open new doc) that might be affected.
Thanks for the insight, guys. Here’s a question that is perhaps more within your speculative realm:
If you were in my shoes (access to notepad, or some other ‘questionable’ application that may not work in the fully-featured way), would it still be a worthwhile investment? I’m a theology student – lots of papers and sermons and what-not.
I appreciate your POVs before I make this financial commitment! :)
Peace,
Tom
@Tom, it really depends what you’re using it for. I’ve been taking notes and organizing them by hand because I don’t have the time to memorize the voice edit commands, so it’s just a replacement for typing. But I’d recommend contacting nuance and asking about their return policy and some questions about your non-msft apps.
GL,
a/
@andyfromcornell, word. Thanks for your help.
Peace,
Tom
I find that one of the big downsides is that for all the words you want to use that aren’t in its dictionary, you have to add just like a spellcheck. Besides having to correct yourself after every mistake. But that’s alright. Very informative post. They just came out with Dragon 11 and Dragon Medical 10.1.