[tabi] for WindowEyes users: release of a script for Windows Speech Recognition

  • From: "Chip Orange" <Corange@xxxxxxxxxxxxxxx>
  • To: <tabi@xxxxxxxxxxxxx>
  • Date: Tue, 22 Jun 2010 15:20:00 -0400

Hi all,

Below is an announcement of a WindowEyes script, which I recently
published on the GW mailing lists:



I have just released a script for the free speech recognition which is

into Windows Vista and Windows 7. This capability allows you to both

dictate text into applications, and to control them, select menu

click buttons, switch windows, and otherwise operate Windows completely

hands free. There is also an optional component, which you usually have

download from the MS site, called WSR macros, which allows you to define

your own verbal commands and their actions (this is included in my WSR


If you are interested in this, be aware you will need a good quality

microphone (say one which is part of a wired headset, no bluetooth and

built-in microphones), and you will need to spend at least 30 minutes

setting it up and training it.

The help text for the WSR script is below. 




Speaks the dialogs of Windows Speech Recognition in Windows Vista and

Windows 7.

If the optional WSR macros module (from Microsoft) is installed, then

additional verbal commands for controlling WindowEyes are available,

allow you to use WE in a completely hands-free manner.

These additional verbal commands are defined in the file

which is in the speech macros subdirectory of the documents folder. They


"read line", causes Window Eyes to read the current line.

"read rest", causes Window Eyes to read from the cursor to the end.

"read title", causes Window Eyes to read the current window title.

"read window", causes Window Eyes to read the active window.

"read status line", causes Window Eyes to read the status line.

"read highlight", causes Window Eyes to read the highlighted block.

"system tray", opens up the window eyes dialog which displays the

of the system tray.

"say time", causes Window Eyes to speak the current time.

"hush" causes Window Eyes to silence speech (equivalent to the control

In order for this macro of additional verbal commands to function, after

WSRMacros component is installed, it will need to be started, and either

it's security level will need to be set to "low", or, you will need to

create a "signing certificate" and "sign" the WE-commands.WSRMac macro


You can do either of these by clicking the WSR macros icon in the system

tray, and choosing the "security" option, when it is running. It is a

deal less trouble to set the security level to low so that unsigned

will run, but this is of course your choice, as it may present a


This script includes a copy of the WSRMacros optional component (file

WSRMacros.msi in your WE profile directory), and should offer you the

of automatically installing it the first time the script is run. If you

not install the WSRMacros at that time, you will not be asked again, but

there is a choice for performing the installation in the WindowEyes

panel, in the script menu for the WSR script. 

However, it cannot automate the process of changing the security level

signing the macro file for you; and you will need to do this before the

verbal commands for WE will function. It is highly recommended that you

install and configure the WSR macros component, as it adds the ability

you to define your own verbal commands, which greatly increase the

usefulness of WSR in general.

The Details of What This Script Does:

When WSR does not know for certain which of several words or phrases you

have said, it will open a window near the bottom of the screen to ask

along with playing a distinctive sound, but this window will not become

active window. Because it's not the active window, WE will not speak

contents. It also will automatically disappear if you try to make it the

active window by moving the mouse, or by trying to alt-tab to it. The

way you can read this window is to use the WE cursor; therefore, one of

main purposes of this script is to read this window automatically.

This window is called the WSR disambiguation dialog. It's purpose is to

display a list of numbered choices to you when it's trying to determine

you meant. You should speak the number which is next to the correct

and then say the word "ok". If there are more choices than will fit on

window (it will display at most 9 choices at a time), then you can say

word "next" to go to the next window of choices. If you don't want any

the offered choices, you can say "cancel", or you can say "spell it" to

put into the spelling panel, where you can spell out the word or phrase

you meant. Doing this is important for allowing WSR to learn how you say

various words. 

Note: you cannot use any of the additional verbal commands for WE while

of these WSR dialogs is active, nor can you use any other WSR commands

except the ones defined for the dialog. You can however, use the ReRead

hotkey defined for this script, if you want to hear the dialog of
choices or

the spelling panel spoken again.

In addition to the hotkey for re-reading a WSR dialog, this script also

a convenience hotkey for starting both the WSR macros module and


Note: There is an undocumented WSR key command which allows you to

toggle the WSR listening state between on and off. It is the combination

the Windows key and the control key pressed together. When WSR stops

listening you will hear a higher tone followed by a lower one. When it

starts listening they will be reversed, and you will hear a lower tone

followed by a higher one.


Other related posts:

  • » [tabi] for WindowEyes users: release of a script for Windows Speech Recognition - Chip Orange