David Pogue says Windows Speech Recognition is "beautifully designed"


David Pogue had something nice to say about Windows Speech Recognition (WSR) in this morning’s NY Times…



“And besides, not all of the new goodies fell from the Apple tree. [] Windows Speech Recognition isn’t as accurate as, say, Dragon NaturallySpeaking, but it’s beautifully designed and much better than previous Microsoft attempts.”


You gotta love “beautifully designed”… Thanks David!


It’s also interesting that he didn’t find it to be as accurate as Dragon Naturally Speaking (DNS). This varies from person to person. There are many individuals that have used both products and for some, DNS is more accurate; for others WSR is more accurate. There have been some interesting debates about this on the Microsoft Speech Forum on Yahoo. If you like speech recognition, that’s a forum you ought to check out from time to time.

Comments (1)

  1. pmad says:

    I’ll check out the forum.

    One way to compare accuracy in a scientific way would be for each user to dictate the "Rainbow Passage" with Dragon and with WSR using the same microphone with USB sound adaptor (if you use one of those) as input and measure percentage accuracy in each programme. For statistical soundness, repeat the dicatation say 3 times in each programme and measure the average. Here’s the procedure I use:-

    http://speechempoweredcomputing.co.uk/Newsletter/?p=69

    Peter