You can call us for help with Video Voice any time between 8am and 9pm Pacific time. Our phone number is 415-644-1855.
Please have your telephone, your computer and your synthesizer next to each other. It's much easier for us to talk you through a problem if we can hear the synthesizer over the telephone.
If your synthesizer doesn't talk at all, please have a sighted person (or an Opticon) available to read your screen. Video Voice displays helpful error messages on the screen that aid in discovering the source of a problem.
If you are using an external synthesizer, know what port your synthesizer is connected to. Your choices are: COM1, COM2, LPT1, LPT2, and LPT3. Questions and Answers
Before calling us, please read over this list of common problems. These are real questions which customers call to ask us every day.
Different types of Votrax
Q: I have a Votrax and it says VOTRAX in my configuration file, why doesn't Video Voice work right?
A: Check which Votrax you have. There are three of them:
Votrax Type N' Talk
Votrax Personal Speech System
Votrax Personal Speech System model B.
The lines in the configuration file for these synthesizers should look like this:
Votrax Type N' Talk:
FirstSyn = TNT
Votrax Personal Speech System:
Firstsyn = Votrax
Votrax Personal Speech System B:
Firstsyn = Votrax-B
Synthesizer talks at wrong speed
Q: How do I make my synthesizer talk at the right speed when I load Video Voice?
A: Add a line to your configuration file which looks like this:
SpeechRate = 4
You can choose any SpeechRate between "1" and 9. "1" is the slowest and "9" is the fastest.
If your synthesizer does not support the chosen rate, it will pick the next closest rate.
Can't find files on hard disk
Q: Before I had Video Voice, everything worked fine. Why can't I find all the files on my hard disk now?
A: Unless we installed programs on your hard disk for you, we cannot be responsible for its condition.
If you follow the hard disk installation instructions for Video Voice in the Installation chapter of the Video Voice manual, nothing on your hard disk will be deleted.
If someone is setting up Video Voice on your hard disk for you, urge them to follow the instructions in the Installation chapter. We cannot help you if we don't know where your files are.
Also keep that person's phone number handy and make sure he documents (writes down) how he set up your machine. If you call us for support, we can help you better if your "expert" is available to explain what he did.
BASIC with Video Voice
Q: Why doesn't Video Voice work with BASIC?
A: It does. You need to reinitialize the synthesizer once you enter the BASIC language. The genuine IBM BASIC "kicks" Video Voice off the serial port. The "GWBASIC" that comes with many compatibles behaves better.
Use the Video Voice REINITIALIZE SYNTHESIZER (PREFIX ALT-BACKSLASH) command to regain control of your synthesizer.
SynPhonix (or Votalker) doesn't work right
Q: How do I make Video Voice work with my Synphonix (or Votalker)?
A: When you use Video Voice with either a Synphonix or a Votalker synthesizer, you need to load two programs first. They are SONIX and TTS.
You don't need to run the program called PORTTALK.
You don't need to specify a "Firstport =" or a "Secondport =" in the configuration file (GRCONFIG.YAK), as Video Voice will ignore them.
Video Voice with Sidekick
Q: Does Video Voice work with Borland's Sidekick?
A: Sidekick conflicts with many other programs, not just Video Voice.
We suggest that you give WordPerfect Library a try. It has some neat features that put Sidekick to shame, and it works very well with speech.
Too much punctuation spoken
Q: Why does Video Voice speak punctuation even if I have Video Voice's SELECT PUNCTUATION LEVEL (PREFIX SHIFT-4 level-number ENTER) set to NO PUNCTUATION (choice 3)?
A: Even when SELECT PUNCTUATION LEVEL (PREFIX SHIFT-4 level-number ENTER) is set to NO PUNCTUATION, some punctuation creeps through on the Echo PC, the Echo Plus, and the BOSS.
This happens because Video Voice sends some punctuation through to the synthesizer to improve emphasis. With less punctuation spoken it is hard to tell if a sentence has ended. Unfortunately, these synthesizers sometimes misbehave and speak the punctuation rather than the emphasis.
FIND BOX doesn't find multiple boxes
Q: Will the Video Voice FIND BOX (PREFIX f b) command find more than one box on the screen?
A: The Video Voice FIND BOX (PREFIX f b) command gets confused if there are multiple boxes on screen. FIND BOX (PREFIX f b) is unreliable if there are multiple boxes on the screen.
Try the Video Voice READ BOX TEXT-UNIT (PREFIX b text-unit) commands. They can handle multiple boxes, so long as the cursor (not a pseudo-cursor) is located within the current box.
FIND BOX can find one box, anywhere on the screen (or active window), so long as it has at least two corner frame characters defining it. FIND BOX does not pay any attention to the location of the cursor.
READ BOX TEXT-UNIT doesn't read the box
Q: Why doesn't Video Voice's READ BOX TEXT-UNIT (PREFIX b text-unit) command read the contents of the current box?
A: Video Voice's READ BOX TEXT-UNIT (PREFIX b text-unit) commands can only read the contents of the box in which the cursor is located. Any boxes outside the box which contains the cursor, is ignored.
It is becoming common for application programs to use a "pseudo-cursor" that looks just like a real cursor, but is not. Video Voice can only follow the "real" cursor.
If Video Voice says "cursor outside active window", odds are the applications program is using a pseudo-cursor. In this case, any Video Voice command that depends on knowing where the cursor is (almost all of the right-hand text-unit commands) will not behave properly. However, if you ENTER REVIEW MODE (PREFIX function-10) and move the review mode cursor to within the box, the READ BOX TEXT-UNIT (PREFIX b text-unit) commands will behave as desired.
There are two variations of the READ BOX TEXT-UNIT (PREFIX b text-unit) commands. The box can be defined either by frame characters or color.
Use PREFIX B text-unit to read the text-unit within a box defined by frame-character.
Use PREFIX SHIFT-B text-unit to read the text-unit within a box defined by color.
Computer locks up if synthesizer not on
Q: How can I prevent my computer from locking up if the synthesizer on my parallel port is not turned on while Video Voice is trying to talk to it?
A: Keep your parallel port synthesizers attached and turned on anytime you are using Video Voice.
The parallel port is waiting for the synthesizer to send back an "it's OK to send more data" message. Since the synthesizer is turned off, the parallel port never receives the message. It just waits forever, halting any other computer activity.
This happens with some PC clones.
Video Voice too large
Q: Can I get a version of Video Voice which uses less memory?
A: Not at this time. Version 2.0 will be smaller because I am rewriting it in C and assembly language.
Searches too slow
Q: How do I make the searches work faster?
A: The Video Voice FIND ATTRIBUTE (PREFIX f a attribute-number ENTER), FIND BOX (PREFIX f b), and FIND TEXT (PREFIX f t text-to-find ENTER) commands are especially slow. Unfortunately, in programming, speed and space usually conflict with each other. I have written faster search procedures, but they take up too much precious memory.
Remember that when you have heard the information you want, you can stop the search by pressing the Video Voice QUIET (double-SHIFT) key combination command. This "shuts up" the synthesizer and aborts the search.
TOGGLE TO SECOND SYNTHESIZER and SELECT ENVIRONMENT don't work right
Q: How do I get the TOGGLE TO SECOND SYNTHESIZER (PREFIX SHIFT-7 1-or-2 ENTER) and SELECT ENVIRONMENT (PREFIX ALT-2 environment-number ENTER) commands to work?
A: Both of these commands require the VIDVOICE.000 overlay file to be available on the default drive. The file VIDVOICE.000 contains all the routines to control the various synthesizers and environments.
If you loaded Video Voice from your hard disk, Video Voice will remember to check the directory that you loaded from.
If VIDVOICE.000 is not accessible (because you had loaded Video Voice from a floppy and then swapped another floppy into the drive, for example), Video Voice will speak an error message. Video Voice will expect you to insert a disk containing VIDVOICE.000 and to press a key to retry.
On certain very incompatible computers, Video Voice might lock up if it cannot find its overlay files. If this happens, use these commands only when you are at the DOS ">" prompt and you are sure the VIDVOICE.000 file is accessible.
Macro programs don't work right
Q: How do I get my macro programs (SuperKey, ProKey, Keyworks, SmartKey, etc.) to control Video Voice commands?
A: It is not currently possible to put the PREFIX command in macros. Video Voice co-exists fine with macro programs, but they see the PREFIX key as Function key 9 (or whatever key you have assigned the PREFIX key to).
You might not really need the macros. Try the Video Voice SELECT ENVIRONMENT (PREFIX ALT-2 environment-number ENTER) command, which does a lot of the macro work for you. Using Video Voice Environments is almost like having a dedicated talking applications program!
DECtalk speaks too slowly
Q: Why does my DECtalk speak slower than my friend's DECtalk?
A: There are two versions of the DECtalk. DECtalk II will speak faster than 300 words per minute. Speech rate choices 8 and 9 are for rates 325 and 350 per minute respectively. If they don't work, then you have an older DECtalk. The Dectalk II also shuts up faster.
DECtalk or Accent spells too much
Q: When my DECtalk, or Accent, encounters a word with an embedded digit or punctuation like "COM1:", it spells the whole word. How do I turn the spelling off?
A: Set the Video Voice TOGGLE SPEAK SINGLE DIGITS (PREFIX ALT-1 1-or-2 ENTER) command to ON to hear it the way you want.
If TOGGLE SPEAK SINGLE DIGITS (PREFIX ALT-1 1-or-2 ENTER) is ON, Video Voice will separate the "COM" from the "1", which will cause the synthesizer to speak the word "COM".
If TOGGLE SPEAK SINGLE DIGITS (PREFIX ALT-1 1-or-2 ENTER) is OFF, Video Voice will send the whole "COM1" to the synthesizer. In the case of the DECtalk and the Accent, this will cause the word to be spelled because there is a digit in it. The other synthesizers are not "smart" enough to make the distinction.
Keyboard echo
Q: How do I get Video Voice to speak my keystrokes consistently. My keystrokes are echoed more quickly with some application programs than others.
A: When the Video Voice SELECT SPEECH SOURCE (PREFIX BACKSLASH source-number ENTER) command is set to KEYBOARD ECHO (choice 2), your keystrokes are spoken to you as they are typed.
Video Voice NEVER echoes a key until it knows the application program has grabbed it. Some programs take a long time to read the keyboard. Wordstar 3 and programs that do lots of disk and screen activity, are examples. These programs are sometimes slower at reading the keyboard than Video Voice.
If you type faster than the application program is able to process your keystrokes, you will not hear them echoed until the application program gets around to reading them.
ALT key doesn't work right
Q: Why doesn't the ALT key on my Toshiba computer work with some Video Voice key combinations?
A: Sometimes the keyboard of the Toshiba doesn't respond when you release the ALT key. The machine will appear locked up, but actually it just assumes that you are still holding down the ALT key.
Tap the ALT key a couple of times, and Video Voice and the Toshiba will work properly again.
This happens most frequently on the ALT punctuation key combinations.
Toshiba locks up
Q: How do I keep my Toshiba T1200-Plus from locking up when Video-Voice is loaded and a disk is being formatted or a drive door is left open?
A: This is a DOS bug found on versions of Toshiba DOS 2.1. It does not happen with the DOS 2.11 built in to the ROM of the Toshiba T1000, nor on the Toshibas which have the enhanced supertwist LCD.
Consider upgrading to the current version of Toshiba DOS.
Too many digits
Q: How can I hear multiple digits spoken as single digits, (for example, 135 spoken as one hundred thirty five)?
A: Set the Video Voice TOGGLE SPEAK SINGLE DIGITS (PREFIX ALT-4 1-or-2 ENTER) command to ON.
This command works with the Votrax, the Votrax-B, the Aicom Accent family of synthesizers, the Microvox, the IntexTalker, the Braille 'n' Speak, and the DECtalk I and DECtalk II.
It works with the SynPhonix or Votalker if you use PORTTALK.
It does not work with the Echo.
Preventing lock-ups
Q: How can I minimize computer lock ups?
A: Here are some hints to help you minimize computer lock-ups:
If using a SynPhonix (or Votalker), load the Sonix driver with the following command:
SONIX /k-
This disables the Sonix driver's keyboard handling routines.
Consider upgrading to an Accent synthesizer. The Accents don't lock-up a computer nearly as much as the SynPhonix or Votalker.
If you are using several resident programs, load them in the proper order. For example:
Memory Mate and Keyworks are loaded before Video Voice.
Turbo Lightning is loaded after Video Voice.
Most clock and calendar/reminder software needs to be loaded last.
Keyboard enhancers, such as KBFIX2 need to be loaded first.
RAM disks and print spoolers need to be loaded first.
Synthesizer drivers, such as those for the Micro Echo and the Soundingboard need to be loaded first.
WordPerfect Library's shell and other menu-shell type programs such as Norton Commander are always loaded last.
If you don't know what the proper order for loading programs is, you must read each program's manual carefully. You can also experiment to determine the optimum loading order for all of your resident software.
Remember: If you use a menu program, you must not load another resident program from within the menu.
Using terminal emulation programs
Q: Video Voice doesn't work with my terminal emulation program.
A: Terminal emulation software is usually very complex to set up. There are special techniques for using these programs with Video Voice.
Some terminal emulation programs, especially those emulating the IBM 3270 series of terminals, will not work at all with screen access programs.
Consult the Modem chapter for instructions on using Video Voice with terminal emulation software.
PORTTALK
Q: When should I use the PORTTALK program?
A: PORTTALK is a program which is sometimes included with the SynPhonix and Votalker cards.
Video Voice can communicate with the SynPhonix and Votalker synthesizers in either of two ways: through SONIX and TTS (native mode), or through PORTTALK.
We prefer to support synthesizers directly, rather than using PORTTALK, because it gives us better control of the synthesizer and because it consumes less memory.
If you want your synthesizer to speak numbers as separate digits, use SONIX and TTS. If you do not want your synthesizer to speak numbers as separate digits, use PORTTALK.
Specify which you want to use in your configuration file (GRCONFIG.YAK):
If you want to use the synthesizer's "native mode" (SONIX and TTS), include either:
FirstSyn = SynPhonix
or
FirstSyn = Votalker
Video Voice treats the SynPhonix and Votalker identically.
If, for example, you want to use PORTTALK on COM1: include:
FirstSyn = PORTTALK
FirstPort = COM1
IsSayingSingleDigits = OFF
Then load the synthesizer driver programs. In either case, you must load SONIX and TTS from the DOS prompt.
To use your SynPhonix or Votalker without PORTTALK, enter the following at the DOS prompt:
SONIX /k-
TTS
To use your SynPhonix with PORTTALK, enter the following at the DOS prompt:
SONIX /k-
TTS
PORTTALK
The PORTTALK configuration option is also handy if you want to use a PORTTALK compatible synthesizer which Video Voice doesn't yet support in its native mode.
The following synthesizers can be "addressed" either in its native mode or through the appropriate PORTTALK emulator:
Computer Aids SoundingBoard models
Automated Functions Laptalker
Artic SynPhonix models
Aicom Accent models
Votalker using Porttalk
If a PORTTALK emulator is used, the configuration file GRCONFIG.YAK should contain:
FirstSyn = PORTTALK
PORTTALK is becoming a standard so there will likely be other PORTTALK compatible synthesizers in the future.
Spelling checker
Q: How do I identify the misspelled word when I am within WordPerfect's spelling checker?
A: Use the Video Voice FIND ATTRIBUTE (PREFIX f a 0 ENTER) command sequence to search for reverse video. Video Voice will speak the highlighted misspelled word.
If you have reset your colors, use the Video Voice FIND ATTRIBUTE (PREFIX f a attribute-number ENTER) command to search for the appropriate color.
Video Voice required files
Q: I followed your instructions, but I can't get all the files to fit on a boot disk.
A: The files you must have to run Video Voice are:
VIDVOICE.COM the main program
VIDVOICE.000 its overlay
GRCONFIG.YAK the configuration file
WORDS.MSG the messages the program speaks
LCOM.EXE (needed only if your synthesizer is on a serial port)
You also need your synthesizer drivers, for example:
SynPhonix SONIX.COM and
TTS.COM
Votalker SONIX.COM and
TTS.COM
Soundingboard SB.COM and
SBLOAD.COM
Accent-PC SPKACNT.DVC
Accent-mini & SPKMINI.DVC
Accent-1200
In order for the disk to boot, it must be formatted and have the two hidden DOS system files plus COMMAND.COM.
In order for the system to boot up talking, you must have an AUTOEXEC.BAT to load the various files in the correct order.
In summary, your talking boot disk must contain:
COMMAND.COM on a formatted DOS system disk
the required Video Voice files
the required drivers for your synthesizer
an AUTOEXEC.BAT file to cause the files to load in the correct order.
You probably will not be able to fit the documentation, README and VVHELP files on your boot disk, so you'll need to make separate work disks with these supplementary files.
Find text
Q: How do I make the Video Voice FIND TEXT command work.
A: The Video Voice FIND TEXT (PREFIX f t text-to-find ENTER) doesn't work in all circumstances. See if it works in your application. If it does, fine. If it doesn't, we apologize.
Too many devices and not enough ports
Q: Can I share Video Voice with a printer (or a modem) on the same port?
A: This is one of those things that, yes, can technically be done. But it will drive you crazy! Ideally, you want to have one device per port.
Consider getting one or more of the following, in order to free up an existing port or to add an additional port.
an internal modem instead of a serial modem.
an internal synthesizer instead of an external synthesizer.
a parallel printer instead of a serial printer, or vice versa.
an extra parallel or serial port for your computer.
When you start sharing your synthesizer with another device on the same port, you will run into output conflicts. We don't advise it!
Speaking paragraphs
Q: How do I get Video Voice to speak paragraphs correctly?
A: The Video Voice PREFIX o, PREFIX l, and PREFIX . (period) commands cause Video Voice to speak the previous, current, and following paragraphs.
Two things can go wrong with paragraphs:
Video Voice will not find any paragraphs
Video Voice will find multiple paragraphs, but interpret as only one paragraph.
Video Voice interprets a space in the screen's column one as the beginning or end of a paragraph. Video Voice doesn't know anything about your document; it only knows what it "sees" on the screen.
Therefore, a space or tab must be in screen column one. This will ensure that Video Voice can find each paragraph.
There are two ways to indicate to Video Voice that you intend a paragraph:
Indent each new paragraph with at least one space or tab
separate paragraphs by a blank line.
Don't edit with the formatting set to double spacing. Edit in single space and then switch to double spacing just before printing.
If you are using a program (like WordPerfect version 4.1) which displays the left margin on screen, Video Voice will see all the spaces in that left margin and gets confused. Do one of the following:
set your word processor so that it doesn't display the left margin on screen
define an active window in Video Voice that excludes the left margin. For example, you might have a line in your configuration (GRCONFIG.YAK) file which looks like this:
Window1 = 1,11;25,80;wordperfect4.2
Remember to use the Video Voice ACTIVATE WINDOW command (PREFIX a w window-number ENTER) when you use Video Voice with the "offending" program.
See the Windows chapter for instructions on defining and activating windows.
The following instances can cause a space to be in column one, even if you do not intend it to begin or terminate a paragraph:
A blank line.
A tab or space at the beginning of the line, used to indent a paragraph.
A tab or space at the beginning of the line, used to align columns of text, as in a chart. Video Voice will incorrectly interpret each tabbed line as a new paragraph.
A tab or space at the beginning of the row, used to indent sections of an outline.
An "indent" at the beginning of the row. Some word processing programs differentiate between a "tab" and an "indent".
This is often used to create a paragraph with a "hanging indent." Video Voice would incorrectly interpret the second and following lines of a hanging indented paragraph as being a new paragraph. Interestingly, the first line of the paragraph would not be considered a new paragraph because the first line would be flush left, without a space in column one.
A space between words in a line that begins with a "margin release". This is a bit of a freak, but it does happen.
A line of text that is centered.
A line of text that is justified "flush right".
A space (or spaces) occupying the beginning of each screen line, used to represent the margin on a piece of paper. Some word processing packages use the left edge of the screen as the edge of the paper and show the whole body of text beginning several spaces to the right of column one. WordPerfect version 4.1 is an example.
A blank line alternating with a line of text, representing double-space line spacing on the screen. Triple spacing would have two blank lines alternating with a line of text, etc.
As you might imagine, many of these instances will cause Video Voice to incorrectly interpret the line with a space in column one as the beginning or end of a paragraph.
You might wonder why Video Voice doesn't simply look for carriage returns. WordPerfect, for example, knows that a paragraph is ended when it encounters a carriage return (sometimes called a hard return) in the text.
Video Voice knows nothing about hard returns because it can only access the screen's representation of the document, not the document itself. Unfortunately, carriage returns don't appear on your computer's screen in a identifiable form. There is no standard character which represents a carriage return; each word processing program represents the carriage return in its own proprietary manner.
So the only way Video Voice knows about a new paragraph is if it finds a space character in column 1.
If you indent all your paragraphs, or separate them each with a blank line, then Video Voice will find that space and know it is at the start of a new paragraph.
If your paragraphs have hanging indents, begin rows of text with a tab or a space, or use flush right or centering, then Video Voice will see each indented line as a separate paragraph.
Speaking sentences correctly
Q: How do I get Video Voice to speak sentences correctly.
A: The Video Voice PREFIX i, PREFIX k, and PREFIX , (comma) commands cause Video Voice to speak the previous, current, and following sentence.
Everything that can cause Video Voice to misinterpret a paragraph can also cause it to misinterpret a sentence. Video Voice assumes that a sentence cannot cross paragraph boundaries.
Speaking keystrokes as they are typed
Q: How do I get Video Voice to echo my keystrokes?
A: There are two Video Voice commands which affect how keystrokes are echoed. They are:
SELECT SPEECH UNIT (PREFIX ALT-1 source-number ENTER)
Use the Video Voice SELECT SPEECH SOURCE (PREFIX BACKSLASH source-number ENTER) command sequence to select your speech source. There are three options:
NO ECHO (choice 1) causes Video Voice to read neither the keyboard nor the screen. This choice is available so that sighted users can use the computer without having to listen to the speech. It does echo a few keystrokes (NumLock, CapsLock, and ScrollLock) which a sighted person might want to know had been pressed by accident.
KEYBOARD ECHO (choice 2) causes Video Voice to get its input only from the keyboard.
SCREEN ECHO (choice 3) causes Video Voice to get its input only from the screen.
SCREEN AND SOME KEYBOARD ECHO (choice 4) causes Video Voice to echo special keys (such as function keys) by reading the keyboard, and to echo letters, numbers, and punctuation marks by reading from the screen. This is the default choice.
Use the Video Voice SELECT SPEECH UNIT (PREFIX ALT-1 unit-number ENTER) command sequence to select your speech unit. There are three options:
CHARACTER UNIT causes individual characters to be spoken immediately after are typed at your keyboard. Character unit is the default.
WORD UNIT (choice 2) causes Video Voice to wait for a space or punctuation mark to be typed before speaking. It speaks the word, then doesn't speak anything else until it is appropriate to speak the next typed word.
LINE UNIT (choice 3) is really a misnomer. Line unit causes Video Voice to wait for a carriage return (ENTER key) to be typed before speaking. It speaks the whole unit of text up to the carriage return, then doesn't speak anything else until the next carriage return. In a data entry form, the unit of text might be the contents of a field. In a document, it might be a whole paragraph.
SELECT SPEECH SOURCE is not directly related to SELECT SPEECH UNIT. This can be a source of confusion. Remember:
SELECT SPEECH SOURCE determines where the information is coming from.
SELECT SPEECH UNIT determines the amount of text spoken, irrespective of its source.
Here are some sample situations:
SELECT SPEECH SOURCE on NO ECHO (choice 1):
Nothing is spoken except the CapsLock, NumLock, or ScrollLock, irrespective of speech unit.
SELECT SPEECH SOURCE on KEYBOARD ECHO (choice 2) and SELECT SPEECH UNIT on CHARACTER UNIT (choice 1):
Each keystroke will be echoed as it is typed.
SELECT SPEECH SOURCE on KEYBOARD ECHO (choice 2) and SELECT SPEECH UNIT on WORD UNIT (choice 2):
Individual characters will not be echoed as they are typed. The characters will be "saved up" and spoken as a whole word.
SELECT SPEECH SOURCE on KEYBOARD ECHO (choice 2) and SELECT SPEECH UNIT on LINE UNIT (choice 3):
All keystrokes will be "saved up" until the ENTER key is pressed. Then the whole hunk of text will be spoken.
SELECT SPEECH SOURCE on SCREEN ECHO (choice 3) and SELECT SPEECH UNIT on CHARACTER UNIT (choice 1):
If your application program uses the BIOS to write to the screen, characters typed on the keyboard will be spoken character by character, but any text put on the screen by the application program will be spoken word by word. DOS uses the BIOS to write text to the screen.
If your application program writes directly to screen memory, neither characters typed on the keyboard nor text put on the screen by the application program will be spoken. Most application programs, such as WordPerfect, write directly to screen.
If your application program writes some of the text using the BIOS and some of the text using direct screen writes, speech will be erratic and unpredictable. dBASE III is in this category.
SELECT SPEECH SOURCE on SCREEN ECHO (choice 3) and SELECT SPEECH UNIT on WORD UNIT (choice 2):
If your application program uses the BIOS to write to the screen, text displayed on the screen will be spoken word by word irrespective of whether it was entered from the keyboard or put on the screen by the application program
If your application program writes directly to screen memory, neither characters typed on the keyboard nor text put on the screen by the application program will spoken.
If your application program writes some of the text using the BIOS and some of the text using direct screen writes, speech will be erratic and unpredictable.
Program doesn't talk
Q: Why does DOS talk automatically, but my word processing program doesn't?
A: IBM compatible computers have two ways to put text on the screen. Which technique is used determines how much text is spoken.
If the BIOS is used to put text on the screen, the text can be automatically read by a screen access program. In this case, Video Voice can "see" the text as it is being put on the screen. Therefore Video Voice will speak the text automatically if SELECT SPEECH SOURCE is set to SCREEN (choice 3) or SCREEN AND SOME KEYBOARD (Choice 4).
If the application program writes directly to screen memory, the text cannot be automatically read by a screen access program. In this case Video Voice cannot "see" the text as it is being put on the screen. Therefore Video Voice will not speak the text automatically, irrespective of SELECT SPEECH SOURCE choice. Video Voice can be instructed to read the text using a Video Voice PREFIX command such as SPEAK TOP ROW (PREFIX HOME).
DOS writes to the screen using the BIOS, therefore it talks automatically.
Most word processors write directly to screen memory, therefore they do not talk automatically.
If you want the text your are typing with your word processing program to be spoken automatically, you must set the Video Voice SELECT SPEECH SOURCE (PREFIX BACKSLASH source-number ENTER) command to KEYBOARD ECHO (choice 2). Video Voice looks at the keyboard rather than the screen, therefore the technique used by your application to put text on the screen is irrelevant.
Turning off speech
Q: How can my sighted friends use the computer without hearing the unwanted speech, but without permanently turning off Video Voice?
A: Use the Video Voice SELECT SPEECH SOURCE -- NO ECHO (PREFIX BACKSLASH 1 ENTER) command when you want to temporarily turn off speech.
When you want to turn speech back on, use the PREFIX BACKSLASH 4 ENTER sequence to turn speech back on with screen and some keyboard echo. You might use the "backslash 2" option to turn on keyboard echo, or the "backslash 3" option to turn on screen echo.
