Speech recog. though it has been around for years is still in it's infancy as a tech. 
The main problem is the microphone. The mics. used on PCs are so far beyond garbage, 
it's amazing they work at all. To do it right one has to have a mic. that has good 
omni directional capabilities and either auto leveling or is in itself able to capture 
enough quality sound that the movement of the person speaking has little or no effect 
on it. What I am saying here is that 99% of the time when someone sets up a speech 
recog. system the are kicked back and relaxed talking in a mono-tone voice trying 
to make themselves punctuate and have a less than normal inflection in their voice. 
Then when they try to use it they speak normally or even in a forced voice. Think 
about how accurate commands have to be typed into the machine. Computers have little 
to no forgiveness. Computers do not intemperate or guess. Look at how lousy spell 
check is at guessing. 

You might try to use the mic, not intending for it to answer you, but just to get 
used to having it on, for a week or two. Then try to retrain your Speech Recog. again. 
An remember when you train it, to be sitting up in the same manner you would be when 
using it. If you normally slur certain words, do so. It is easier to speak normally 
than in a clumsy computer voice.

• Reply or follow-up to this message

• Search this forum
• Start a new thread on a different subject

• Report abuse of the forum