blindphoneman
commented
3 years ago Hi again:
Just wanted to make one more clarification.
When I mention using control codes, like ctrl+a to begin voice adjustment or ctrl+x to silence speech (clearing the buffers), these are not typed by the user, but, sent to the serial port (or isa bus for internal speech cards) by the screen-reader program itself.
I should mention that some speech devices may also use hardware handshaking, but, not sure which do. If there's interest, I'll find out.
Jookia
joncampbell123
Hi:
I am a blind computer user who started using DOS back in the '80's.
Due to projects like this one, interest in DOS by the blind is being revived.
Currently, we can use any of the emulators with a null-modem emulator (com0com), which then communicates with an emulator to simulate an external serial speech synthesizer.
This can be tricky to set up, though, and it occurs to me that, perhaps a new use could be made of DOSBox-X's serial port system.
Instead of using 3 different programs, could some kind of module be incorporated into DOSBox-X that could be addressed something like this:
serial1 = speech
When DOSBox-X sees this, it could redirect output to a serial port to the SAPI speech routines in Windows. I believe Linux also offers this, eand MAC-OS has speech as well, so it could work across platforms.
The ESpeak synthesizer could also be used, as it is lightweight, and I think is open-source.
Using something like this, a blind person could load his favorite screen-reader, which runs as a tsr intercepting DOS output and putting it out the serial port, with some control codes, when the program wants to change the pitch or speed of the voice.
Though the screen-readers are not free, as such, most of us still have our old files around,.
As for control codes, the minimum might be to use CTRL+X to silence the speech. But, it is possible to have some codes like CTRL+A, a number, and a letter to change, say, the speed. For example:
^A5S
to set the speed to 5. Some synthesizers used numbers from 1 to 9, but, others may have had finer adjustments. I believe, there was also CTRL+a, a number from 1 to 9, followed by a letter P to change the pitch. This is used a lot to identify capital letters when screen is being reviewed character-by-character. The pitch can go up to signify a capital letter, then back to normal for a lower-case letter.
The reason I see a need for this is that the setup of the current arrangement involves configuring, not only DOSBox-X, but, the COM0COM package, which may require turning off or modifying SecureBoot settings in UEFI firmware, something that a blind person may find difficult, if not impossible without sighted assistance.
This way, the blind person would only need to set up the DOSBox-X config file, then run DOSBox-X and the SAPI speech would take over. No need to run 2 other programs and no need to get sighted assistance for fiddling with UEFI settings. Again, the minimum of a CTRL+X would be needed to silence speech, but, it would be nice to have the others, at least, the ones to change pitch and speed. Don't know about a buffer-size limit, but, some speech synthesizers do send X-Off and X-On when their buffers are too full.
It is also possible to use some kind of 2-way communication, called "indexing" for the speech synthesizer to let the screen-reader know what it is speaking, This makes it possible for the screen-reader to be automatically reading text, and, when the alt or ctrl key is pressed, the screen-reader knows where the person has left off, but, that's probably in the realm of wishful thinking.
I would be willing to help test this and find out what codes are needed. Think I can dig up some manuals or something to get specifics on the codes.
Thanks for your attention.