launchy / listary

[marcspc]

I want to use voiceattack for launching ANY program on my computer.

Right now we have launchy / listary programs, on a key combination (windows + < on my case) it shows a bar on top where you can input whatever and it will display all results from your hdd:



I would like to be able to implement voiceattack here, I was thinking on something like
<computer, search>
send windows+<
start dictation

but I don't really understand what's the best way to stop dictation, can I make it stop after some seconds of silence?
and how do I send the dictation variable to the search input? can't just dictate on real time?

[Pfeil]

Basic dictation input can be accomplished using the "Start Dictation Mode" action, and the "{DICTATION}" token:

Code: [Select]

Start Loop While : [{STATE_KEYSTATE:ENTER}] Equals '0'
Start Dictation Mode (Clearing Dictation Buffer)
Start Loop While : [{EXP:{DICTATION} + {STATE_KEYSTATE:ENTER}}] Equals '0'
End Loop
Stop Dictation Mode
Quick Input, '{DICTATION}'
End Loop
This will type out what you speak, after it's recognized.

I do suggest you start out by getting a feel for the accuracy of dictation as interpreted by the Microsoft Speech Recognition engine, as the performance may not be satisfactory in this role.
It is most competent recognizing given phrases, as used when speaking VoiceAttack commands.

can I make it stop after some seconds of silence?

You could use the "{STATE_SPEECHACTIVE}" token to determine whether the speech engine is currently detecting speech.
An example of such an implementation can be found here.

how do I send the dictation variable to the search input?

In the above example, quick input is used to type the recognized text, one character after the other; A faster method is to paste everything at once:

Code: [Select]

Set Windows clipboard to '{DICTATION}'
Press Left Ctrl+V keys and hold for 0,06 seconds and release


can't just dictate on real time?

If by "real time" you mean a true "speak as you type", where individual words appears as you say them, that is beyond the capabilities of the Microsoft Speech Recognition engine. Even Dragon Naturally Speaking, which is arguably more advanced(and quite expensive), recognizes segments of sentences rather than individual words.

[marcspc]

I do suggest you start out by getting a feel for the accuracy of dictation as interpreted by the Microsoft Speech Recognition engine, as the performance may not be satisfactory in this role.
It is most competent recognizing given phrases, as used when speaking VoiceAttack commands.

maybe I should just create a command for every most used program, thanks for the explanation