(This is the third installment of “Look Ma No Hands- Voice Recognition Technology Today”. Once again there is a video at the end of the post showing the process of writing it prior to the physical clean-up that took place after. The video is truly boring but it does give a good sense of the speed, or lack there-of on a netbook.)

Yesterday I did a quick post on setting up Dragon NaturallySpeaking Version 10. It’s a relatively easy process but it does take some time. It was done on a MacBook Pro with a 2.4 or 2.5 GHz processor and 4 GB of RAM. While that wasn’t able to use all of the resources since it was a virtual machine and the computer was also running Mac OSx it still was quite speedy. Better still, because the microphone is as good as it is I didn’t need to use a headset.

This is a quick post simply to show how well you can use Dragon NaturallySpeaking Version 10 on a netbook. I’m actually recording this post using a Toshiba NB205. Even better, once again I’m not using a headset. I’m sitting in front of the computer and speaking. And while I am speaking slower than I normally would and I’m articulating more clearly than I normally do it isn’t totally unnatural for me to speak this way and as you can see it’s capturing all the words quickly and with almost 100% accuracy.

Continue Reading

In this, the second edition of “Look Ma No Hands, Voice Recognition Technology Today” we’ll take a look at getting started with Dragon NaturallySpeaking Version 10 Professional.

Dragon NaturallySpeaking 10 Standard has an MSRP of $99.
Dragon NaturallySpeaking 10 Preferred has an MSRP of $199.

Back in August Larry and I both got a chance to check out an early version of Shape Services coming update to their IM+ Instant Messaging application. Well the update is out and is sports some nice improvements! Version 3.4…

- Brings Animated Emoticons.
- Allows you to copy text from chat bubbles.
- Brings new email notifications via Push – for Hotmail, Yahoo and GMail.

AND… it introduces Speech recognition as an in-app purchase feature.

Continue Reading

photo credit: Ozyman

Technology can be a lot of fun. It can keep us connected with one another, it can transform our work habits, and it can help us become more organized and efficient.

It can do all these things, but of all the great uses of technology the one that stands as most important is this: that it can help us solve the problems life throws at us.

Dan is a good example of some of the ways technology can help overcome otherwise challenging issues. Back in January he wrote a post entitled A Crash Course in One-Handed Blogging. Dan had just undergone surgery to fuse one of his wrists after it had been destroyed by rheumatoid arthritis. With one arm tightly wrapped and constantly elevated, voice recognition software allowed him to keep blogging and writing for his rabbinic works (and for Gear Diary), even as he recovered from a surgery that was far more extensive than he expected.

Dan continues to do a great deal of his writing using various voice recognition solutions, and he will be introducing a new series about the subject here on Gear Diary later this week. But this story isn’t about VR, it is about another technology that has helped Dan deal with some of the after-effects of RA and his surgery; it’s about a technology that I have long advocated.

With his permission, here is the story…

Continue Reading

If you have ever wanted to give voice-to-text software a shot but found it a bit too pricey here’s your chance to jump in.

Buy.com has Dragon NaturallySpeaking V10 Standard for just $31.99.

That’s a huge savings on the best voice recognition software ever.

SORRY… looks like it is gone…

Vlingo provides voice-control for BlackBerry devices, iPhones, and is currently beta-testing a version for Nokia smartphones (N95, N95 8GB, E51 and E71). I have been using free versions of Vlingo on both a BlackBerry and an iPhone, and have been very impressed. When Vlingo released version 3.0 for the BlackBerry, I immediately wanted to put the update through it’s paces.

Continue Reading

This post began as a quick post about the update to MacSpeech Dictate — software that brings accurate voice recognition technology to the Mac. Dictate was first released a little over a year ago and was the first time that voice recognition on the Mac worked well. It was built on the Dragon NaturallySpeaking Version 9 engine and, while falling short of that application and its accuracy, worked fairly well. Version 1.5 of the software just began shipping last week. At first I was a little miffed because it was a rather expensive upgrade but then the customer service rep explained that it’s actually entirely new software. This version is built on Dragon’s Version 10 and brings world-class voice-recognition to the Mac. How good is it? I just dictated this first paragraph without having to make one correction.

There was, however, a bit of a problem.

The version that initially shipped allowed me to use the internal mic on my MacBook for voice recognition. No, it wasn’t the most accurate thing in the world. The recognition did decrease significantly when I didn’t use a noise canceling headset. But the convenience of being able to sit in front of my MacBook and simply speak without having to worry about headsets and a variety of cords running from my MacBook, made it a worthwhile trade-off. Unfortunately, this new updated software doesn’t allow the internal mic to be used at all. It makes sense since someone using the internal mic might encounter a significant reduction in the quality of the recognition and that reflects badly on the manufacturer but it was nothing short of a major annoyance for me. I don’t want to have to carry a headset. I don’t want to have to deal with it. And I’m willing to give up a little bit of accuracy for the convenience. And I want that to be my choice not to companies.

And that’s really where the gist of this post comes in –

It took me about seven minutes of googling to find people who had encountered the same issue, wanted the same outcome, and had posted workarounds. And it actually turns out that it’s a simple workaround.

There’s a little piece of free software from RogueAmoeba called LineIn that allows the internal mic to be fed into the virtual audio device application Soundflower (corrected) which is then recognized by the software as a microphone. Simple and quick. Best of all it works! In fact, I’ve gotten to this point in the post without a headset and have not had one inaccuracy in the transcription.

There is something wonderful about the online community. Strangers are often willing to share their knowledge and support with others. Developers create all kinds of software and give it away for pennies or even free. And people who have an issue and find a way to resolve it are often more than happy to post the solution so that others don’t have to spend time searching for it.

By the way, if you do get the new version of MacSpeech Dictate the solution/workaround for using the internal mic can be found here.

A note to the good folks at MacSpeech –

Your software is fantastic. It finally stands up to the accuracy that I lost when I made the move from Windows to Mac. I am absolutely blown away by the accuracy of this new version and the fact that it is far faster than the previous release. Between this new software and the fact that I’m using a superfast notebook, I’m able to speak as quickly as I want and have the transcription pretty much keep up with what I want to say.   This new version of the software  is worth every penny, whether buying it outright or paying for the upgrade, even if it is rather expensive as software goes these days. But please recognize this — it ought to be the customer’s choice whether or not they want to trade off convenience for accuracy. And more importantly – the accuracy doesn’t suffer one bit from using the internal mic because I’m now at the end of this post and there’s not been one time I had any problem whatsoever with the software recognizing what I wanted to say.

I use speech recognition quite a bit.  I need to. I do a lot of writing and never learned how to touch type. Add to that the fact that I destroyed my wrist a number of years ago and — voice recognition is a godsend.

Unfortunately, just a few years ago even the best voice recognition was slow and inaccurate. Over the past few years however, the speed has improved significantly and the accuracy gets better and better. Yes, it isn’t perfect, and if I don’t take the time to carefully go through any mistakes that might have crept in… well you might say I have embarrassed myself more than once. But overall, the speed and accuracy are quite good. For Windows users, Dragon NaturallySpeaking is excellent. I have version 10 on my netbook, and it runs quite well. Mac users haven’t been quite as lucky. For years the applications that were available lagged behind. Fortunately last year MacSpeech Dictate was released. Built on the Dragon NaturallySpeaking version 9 engine, it was a huge leap forward. Not as good as Dragon, but quite good.

Earlier today I received notice that MacSpeech Dictate version 1.5 has just been released. It was a case of good news… bad news. Continue Reading

Do you like talking to your BlackBerry?  Do you dream of being able to use your voice to write emails,  memos, tasks, and calendar entries?  Well, MyCaption.com, Inc now has their voice-to-data MyCaption product available on BlackBerry® App World™.

Too busy to type?  This application may just be what you’ve been looking for.  The application is geared towards business users who need to convert their voice to searchable, actionable data.

Continue Reading

I’ve been using voice recognition software for a number of years. Back in my Windows days I used Dragon NaturallySpeaking Speaking. I started with Version 8 and the accuracy improved through my use of Version 9.5. (Version 10 has just been released and it promises to be even better!) When I made the move to Mac a little bit over a year ago I went searching for voice recognition software. I found and purchased iListen from MacSpeech. It worked well enough but was a far cry from the power and accuracy of Dragon. A few months ago MacSpeech updated iListen, (or more actually reinvented the software) and introduced MacSpeech Dictate. This was a giant leap forward, particularly because the new software employs the same voice recognition engine as Dragon. It works quite well.

There are two main components that impact the accuracy of creating text to voice.
The first is the quality of your speech. If you speak in a smooth, consistent manner, using, as the software suggests “a newscaster’s voice” you will have much better results than if you speak the way you would to a friend.

The second, and perhaps more important, is the quality of the microphone. A better microphone will cancel background noise and significantly improve your results.The best option I found was my Fujitsu Stylistic Tablet PC. It had array mics built in which filtered out noise quite well. As a result I was able to sit with the tablet on my lap and simply speak in order to write with tremendous accuracy. When I was using a desktop or laptop, however, the lack of a dual array mic meant that I needed to use a headset. My headset of choice was a noise canceling Andrea USB headset. For about $90 it worked incredibly well. It was, however, cumbersome, slightly uncomfortable, more than slightly inconvenient, and it made me look like a phone operator. But it worked.

A few months ago I noticed that a new wireless microphone had been released by Revolabs. The ex-tag microphone, while rather pricey at $250, promised excellent accuracy while cutting the cord. Mine came today and it is a small wonder. Continue Reading