J
John Doe
Speech recognition, Naturally Speaking vs Windows Speech
Recognition
You cannot tell how speech recognition works unless you try using
it for at least weeks. You have to get over the initial impression,
usually either "Wow!" or "Shit!".
I've been using Microsoft's built in utility for many months now. I
can do it, my speech is clear enough, but seems I have to talk
louder and put more effort into it. At the same time, it picks up
background sounds/noises, that is very irritating. Naturally
Speaking is excellent at filtering out background noises. Its user
dictionary sucks. And then there are all sorts of little things to
complain about with Microsoft's version. Compared to Naturally
Speaking, Windows Speech Recognition is practically featureless.
There are hardly any configuration options. There's no keyboard
shortcut to toggle the microphone on/off. That (efficiently turning
off the microphone) would be very useful since it picks up
background noises. You cannot force dictation (instead of allowing
a command) by pressing and holding a keyboard key (I use the
control key, makes it a breeze). You must say "insert" prior to the
dictation. That is cumbersome. It won't automatically stop
listening after a certain number of minutes (not a big deal, but
definitely useful). And many more.
Another thing is that the newer version of Naturallyspeaking
includes text-to-speech. It says it's not intended for web pages,
but I think that just means it's not fully functional on a webpage
(it has some fancy features that might not work). All I need is for
it to read selected text, like my current faulty TTS program does.
Will see.
I cannot say the following with enough emphasis...
Microsoft includes a screen reader in windows 8, but MICROSOFT IS
TOO FREAKING LAZY TO ALSO MAKE IT READ SELECTED TEXT so that it
functions like a simple text-to-speech program. Lots of people can
(and some do) benefit by text to speech, like people with poor
eyesight or people who tremendously enjoy reading without keeping
their eyes glued to the screen. And all Microsoft would have to do
is a fairly simple programming task of allowing its built in screen
reader to simply read selected text. The voices are there. The
ability to read text is there. The only thing that's missing is the
ability to read selected text on demand. Makes me wanna tear at my
garments. Some people who don't use speech might disagree, but if
they were able to think, they would realize that speech input and
output is fundamentally no different than keyboard input and video
output. If anything should be built into the operating system,
speech is it. And yet Microsoft has always been horribly lagging in
that regard. It is very nearly criminal IMO. Of all the complaints
I have against Microsoft, that is easily the most serious.
Recognition
You cannot tell how speech recognition works unless you try using
it for at least weeks. You have to get over the initial impression,
usually either "Wow!" or "Shit!".
I've been using Microsoft's built in utility for many months now. I
can do it, my speech is clear enough, but seems I have to talk
louder and put more effort into it. At the same time, it picks up
background sounds/noises, that is very irritating. Naturally
Speaking is excellent at filtering out background noises. Its user
dictionary sucks. And then there are all sorts of little things to
complain about with Microsoft's version. Compared to Naturally
Speaking, Windows Speech Recognition is practically featureless.
There are hardly any configuration options. There's no keyboard
shortcut to toggle the microphone on/off. That (efficiently turning
off the microphone) would be very useful since it picks up
background noises. You cannot force dictation (instead of allowing
a command) by pressing and holding a keyboard key (I use the
control key, makes it a breeze). You must say "insert" prior to the
dictation. That is cumbersome. It won't automatically stop
listening after a certain number of minutes (not a big deal, but
definitely useful). And many more.
Another thing is that the newer version of Naturallyspeaking
includes text-to-speech. It says it's not intended for web pages,
but I think that just means it's not fully functional on a webpage
(it has some fancy features that might not work). All I need is for
it to read selected text, like my current faulty TTS program does.
Will see.
I cannot say the following with enough emphasis...
Microsoft includes a screen reader in windows 8, but MICROSOFT IS
TOO FREAKING LAZY TO ALSO MAKE IT READ SELECTED TEXT so that it
functions like a simple text-to-speech program. Lots of people can
(and some do) benefit by text to speech, like people with poor
eyesight or people who tremendously enjoy reading without keeping
their eyes glued to the screen. And all Microsoft would have to do
is a fairly simple programming task of allowing its built in screen
reader to simply read selected text. The voices are there. The
ability to read text is there. The only thing that's missing is the
ability to read selected text on demand. Makes me wanna tear at my
garments. Some people who don't use speech might disagree, but if
they were able to think, they would realize that speech input and
output is fundamentally no different than keyboard input and video
output. If anything should be built into the operating system,
speech is it. And yet Microsoft has always been horribly lagging in
that regard. It is very nearly criminal IMO. Of all the complaints
I have against Microsoft, that is easily the most serious.