Since I got my new headphones with a microphone, I decided I would try out the Vista Speech Recognition. Unless I can't avoid it, this whole post is dictated.
The problem, though, is that it doesn't work naturally everywhere, and I can't just use it free form in whatever I want, I instead get this Correction panel come up every time to confirm what I said.
As explained by a Microsoft blogger, it uses specific APIs for the speech box, and everything not using it can only be enabled by an option called "Dictate Anywhere", which gives the option to correct what was said, because the correction commands cannot be used in these forms. This, unfortunately, means that I cannot use it normally in either Firefox, and OpenOffice. Making it somewhat less than useful for my usual functions.
I have stopped dictating at this point and am typing the rest, because it is faster. I haven't trained the listener yet, so it won't have all my speech nuances down yet, but it is also having problems with my Utahn accents and the way I (don't) pronounce my 't's, and as I was writing this entry it was having a problem recognizing and entering "Dictate Anywhere", "Firefox", and, strangely enough, "Vista", and I had to type those both in by hand.
And further aggravating is I had do disable my Sound Blaster Audigy and go to my onboard audio to be able to use the microphone, because on my Audigy the line in and microphone jack are the same, and it didn't seem to know how to put the jack in a boosting mode so I had enough input level to be able to use the microphone.
Date posted: 01 January, 2008 Tags: programming software voice_recognition
« Blogotubes | The Zompire »