APPENDIX F SPEECH SYNTHESIZERS Vocal-Eyes is very easy to get along with. By that we mean it works with many voice synthesizers. Vocal-Eyes was not written for a specific synthesizer but takes advantage of each synthesizer's unique features. Some screen readers only work with one synthesizer and others require you to get a completely different version for each synthesizer you wish to use. In our minds, this is not acceptable. Here are the synthesizers Vocal-Eyes directly supports: - Sounding Board by GW Micro - All Accent boards by Aicom Corporation - All artic boards compatible with SONIX.COM and TTS.COM and the TransPort by Artic Technologies - Apollo by Dolphin Systems - Braille 'n Speak by Blazie Engineering - DECtalk by Digital Equipment Corporation - Portable DECtalk by The Childrens Hospital - MultiVoice by Childrens Hospital - DECtalk PC by Digital Equipment Corporation - Echo PC and GP by Street Electronics - Internal ECHO PCII, PC+, MC and 1000 by Street Electronics - DoubleTalk PC & LT by RC Systems - Audapter by Personal Data Systems - ASP by Automated Speech Functions - Personal Speech System versions A and B by Votrax - Prose 4000 by Speech Plus Incorporated - LiteTalk by Microtalk - Reading Edge by Xerox & Digital - Votalker by Votrax Did you see your synthesizer in this list? If not, all is not lost. There is a Generic option which will probably work. Many of you may say, "Sure it will work but I will not be able to use any of its features such as rate, pitch, volume and most of all the silence command." Fear not, you can set up what Vocal-Eyes should send for each of the ten possible rates, ten volume levels, ten pitch levels, twenty-six tones, silence string and even an initialization string. Chances are, we already fully support the synthesizer you are using. But if not, as you can see, you can customize the generic option to your individual synthesizer. If your synthesizer supports all of the above commands and it interfaces through a parallel or serial port, Vocal-Eyes will use them. When you first start up Vocal-Eyes, it scans memory for your synthesizer driver. First, Vocal-Eyes will search for the Sounding Board software. If it does not find it, it will look for one of the supplied synthesizer drivers. Each of these drivers will be discussed in detail later in this appendix. The Sounding Board is the only synthesizer which does not require one of the supplied drivers. If you are using a synthesizer other than the Sounding Board, the corresponding driver must be loaded prior to loading Vocal-Eyes. If no synthesizer was found, you will get the message: Synthesizer driver not loaded and Vocal-Eyes will not be loaded. If you are using the Sounding Board, issue the commands to get it going. If you are using another synthesizer load the proper driver. Now rerun Vocal-Eyes and this time it will work. Notice you do not have to tell Vocal-Eyes itself what port your synthesizer is connected through. Vocal-Eyes gets all the information it needs directly from the synthesizer driver automatically. Below is a list of each synthesizer and what is required in order to use it with Vocal-Eyes. Also, any unique features relating to each synthesizer will be discussed. If you require any further information relating to your synthesizer, please refer to your synthesizer's owners manual. SOUNDING BOARD The Sounding Board is by far the best choice if you require an extremely responsive synthesizer. Every feature Vocal-Eyes supports is handled by the Sounding Board. In fact, the Sounding Board hot keys can be used to enhance Vocal-Eyes performance. Fastforward and Rewind work great. If you press either the increase or decrease hot keys for the rate or volume, Vocal-Eyes will automatically adjust the screen voice setting accordingly. The only hot key which will not work with Vocal-Eyes is the punctuation hot key. This is because Vocal-Eyes uses a much more sophisticated punctuation scheme. Vocal-Eyes therefore automatically disables this hot key. Feel free to setup and use all of the hot keys as you see fit. The Rate, Volume, Pitch and Tone commands all map exactly into the Sounding Board commands. For example a Vocal-Eyes rate of seven matches exactly rate seven of the Sounding Board. In order to use the Sounding Board, you must load the programs SBLOAD.COM and SB.COM. These programs are supplied with the disk you received with your Sounding Board. However, the very latest version is also on your Vocal-Eyes disk. Vocal-Eyes requires version 2.5 or greater. If you try and use Vocal-Eyes with an earlier version of the Sounding Board software, you will get an error message and Vocal-Eyes will not be loaded. Simply replace the version you were using with the version supplied on your Vocal-Eyes disk. Therefore, assuming you don't require any Sounding Board command line parameters, you would have to type the following at the DOS prompt: SBLOAD SB Now simply start Vocal-Eyes in the usual manner. Vocal-Eyes will take over from here. If you are using macros and have the MACRO.COM program loaded you will notice when you enter Review Mode by pressing the fast review key, not only do the macro keys no longer work, the Sounding Board hot keys no longer work either. If you enter via the Review hot key, both the macro and the Sounding Board hot keys work as always. This is done intentionally for those using macros. NON SOUNDING BOARD USERS As we said earlier, all synthesizers except the Sounding Board require a supplied driver. Although we are referring to these as drivers, you do not load them in the CONFIG.SYS. These are actually .COM files which are loaded directly from the DOS prompt. The following lists all device drivers and what synthesizers they are to be used with. All ACCENT synthesizers ACCENT.COM Apollo I (external) APOLLO.COM Apollo I (internal) INAPOLLO.COM Artic boards which use SONIX.COM and TTS.COM ARTIC.COM Artic TransPort ACCENT.COM ASP VEASP.COM Audapter AUDAPTER.COM Braille 'n Speak BNSPEAK.COM DECtalk DECTALK.COM DECtalk PC DECPC.COM DoubleTalk PC & LT DBLTALK.COM Echo PC or Echo GP ECHOPC.COM Echo PC (new release) NEWECHO.COM Internal Echo PCII, PC+, MC or 1000 INTECHO.COM LiteTalk LITETALK.COM MultiVoice MVOICE.COM Prose 4000 PROSE4.COM Personal Speech System version A PSSA.COM Personal Speech System version B PSSB.COM Portable DECtalk PDECTALK.COM Reading Edge READEDGE.COM Votalker ARTIC.COM All other synthesizers GENERIC.COM The appropriate device driver must be loaded before you load Vocal-Eyes. The following describes the available command line parameters which can be used for any of the above drivers. /? (help) If you issue this command on any of the above drivers, that driver will display all valid command line parameters to the screen. The information will not be voiced unless you have already loaded Vocal-Eyes. Each synthesizer may contain different information depending on its individual capabilities. This command can be issued any time. For example, DECPC /? This will display all the available command line parameters for the DECtalk PC synthesizer. After each screen full of information, you will be prompted to "Press any key for more." Simply tap any key for the next screen full of information. /!Cn or /!Ln (Port assignment) Certain synthesizers require you to specify the port which they are connected to. If this is the case, you must include this command on the command line. If you do not specify this parameter and your particular synthesizer requires it, you will get the following error message. Speech port not specified! For a list of valid command line parameters issue a /? after the driver name Don't worry if you get this error message. Simply issue the command again, including the port assignment command or issue the /? for the syntax of the necessary command(s). The /!C or /!L parameter informs Vocal-Eyes where to find your synthesizer. Enter this parameter by typing /! and then following it with either a C or an L (C standing for COM port and L for LPT port) and a number designating the number of the port. /!C1, for instance, informs Vocal-Eyes that your synthesizer is connected to COM1. /!L3, on the other hand, designates LPT3 as the appropriate port. You can specify any of the three LPT ports one through three or any of the four COM ports one through four. Drivers which require a port assignment are: All Accent synthesizers, external Apollo, ASP, Audapter, Braille 'N Speak, DECtalk, Portable DECtalk, MultiVoice, Echo PC, Personal Speech System, LiteTalk, Reading Edge, Transport, and Generic. Drivers which do not require a port assignment are: Internal Apollo, Artic cards, DECtalk PC, Votalker, DoubleTalk PC & LT, Internal Echo, and Prose 4000. If you specify a port on any of these drivers, it will be ignored. /!Bn (Baud rate assignment) This parameter informs Vocal-Eyes the baud rate of your synthesizer. If you are using a parallel synthesizer, or one which did not require a port assignment, this command will be ignored. If you do not specify this parameter, the device driver will default to 9600 BAUD. If you wish to use the command, use one of the following forms: /!B1 = 1200 baud /!B2 = 2400 baud /!B3 = 300 baud /!B4 = 4800 baud /!B9 = 9600 baud /!Pn (Parity assignment) This parameter tells Vocal-Eyes the parity required by your speech synthesizer. If you are using a parallel synthesizer, or one which does not require a port assignment, this command will be ignored. Enter N for No parity, O for Odd parity or E for Even parity after the /!P. For example, /!PO would signal odd parity and /!PN would signal no parity. If you do not use this parameter, the device driver will default to no parity. /!Dn (Data bits assignment) This parameter informs Vocal-Eyes the number of data bits required by your speech synthesizer. If you are using a parallel synthesizer, or one which does not require a port assignment, this command will be ignored. Enter /!D7 for 7 data bits or /!D8 for 8 data bits. If you do not use this parameter, the device driver will default to 8 data bits. /!Sn (Stop bits assignment) This parameter tells Vocal-Eyes the number of stop bits required by your speech synthesizer. If you are using a parallel synthesizer, or one which does not require a port assignment, this command will be ignored. Enter /!S1 for 1 stop bit or /!S2 for 2 stop bits. If you do not use this parameter, the device driver will default to 1 stop bit. EXCEPTION DICTIONARY With Vocal-Eyes, even if your synthesizer does not directly support an exception dictionary, you can still set one up. Each driver supports a user exception dictionary. Those of you familiar with the Sounding Board will feel at home with this feature as it was modeled after your synthesizer's exception dictionary. Of course no matter which voice synthesizer you are using, there will always be unusual words, specialized terms, proper names, and symbols that it will not pronounce exactly as you might like. For that reason, we've given you an ability to add your own custom dictionaries of exceptional words. You must first create a standard ASCII file containing the exceptional word(s) and a mis-spelling of the word(s) that sounds proper. For example, if you would like to have the abbreviation "lb" to be spoken as "pounds" and the roman numeral VI to be spoken as "6", enter these four lines in an ASCII file: lb pounds vi 6 Note: Upper or lower case is not important. Enter the correct spelling that is mis-pronounced on the first line, and the mis-spelling that sounds better on the following line. Give this file a meaningful name, and save it on disk. How about "VE.DIC." You'll probably find the exception dictionary feature quite useful for abbreviations, roman numerals, acronyms, and proper names. As was said earlier, this file must be in a standard ASCII format. If you are using a word processor, make sure you pick this option when saving. If you are using Noteworthy, simply save it as you would save any file. There are a few rules about what can and can not be in this file. 1) Each entry must take two lines in the file, the correctly spelled word on line one and the misspelled word on line two. 2) The first line (correctly spelled word) can ONLY contain letters A-Z and the apostrophe. No other symbols are allowed. 3) The second line (misspelled word) can contain any standard printable characters including spaces. Let us see how to use the feature. When you first startup your specific synthesizer driver for your current session, it will attempt to load the dictionary file called "VE.DIC" by default. If the file exists, it will automatically be loaded and used. Nothing is required other than to load the specified driver as you have always done. However, if you wish to load a file with a different name you will need to use the following command line parameter: Parameter: /Efilename The synthesizer driver will load an ASCII file containing new rules for text- to-speech translation by using the "E" parameter. "filename" can be any valid DOS file specification, including drive and directory information. For example, if you are using an ECHO PC synthesizer and you want to load a dictionary file called 'LOTUS.DIC' on the C: drive under the directory called 'LOTUS', you could issue the command: ECHOPC /EC:\LOTUS\LOTUS.DIC After the dictionary is loaded successfully, the driver will confirm its acceptance of the new rules with a message. Remember, by default, the first time the driver is issued it will look for a file called "VE.DIC" on the DOS default log path, and load it, if it's located there. The number of new rules that can be added depends on a buffer setting. The default buffer size will accommodate approximately thirty to forty entries. If there is a problem while trying to load the dictionary, you will receive one of the two error messages: * Dictionary load error * * Dictionary format error * The Dictionary load error means the driver could not find the specified file name or the file was too large for the supplied buffer. Verify the file exists and your buffer is large enough to hold the supplied dictionary. The format error indicates that you did not follow the above rules while creating the dictionary file. Verify that the file is in ASCII format and the two entries contain valid characters. Previously, we said the default buffer size would allow around thirty to forty entries. This number of course depends on the size of the words you are defining but it also depends on the buffer size. The drivers allocate 512 characters of storage for dictionary entries by default. Depending on the size of each entry, this allocation will accommodate up to approximately forty entries. The arithmetic here is 512 divided by twelve equals forty-two. We are using six characters as the average word size, and we are doubling that to encompass the mis-pronounced word and the properly pronounced substitute. Should you find this allocation to be too small, use the exception dictionary buffer parameter to increase the size of the buffer as needed. To adjust the buffer size, use the following command: Parameter: /!Ex Notice the ! (exclamation point). As always, since this command contains the exclamation point it can only be issued the first time you startup your driver each session. The x is a number between zero and forty-nine. The number determines how many 512 character chunks to allocate as part of the buffer. The default is one or 512 characters. For example, if you wanted to triple the number of possible entries, the command would read like this: DBLTALK /!E3 The "E" in this parameter stands or "exception", and the number corresponds to some multiple of 512 characters. Notice that the dictionary buffer can be set to zero characters. This is because if you do not wish to use an exception dictionary file, you should not be penalized by taking valuable memory. Therefore if you are not using an exception dictionary, use the command: DBLTALK /!E0 This will set the buffer size to zero, freeing up a valuable 512 characters. Of course, you will not be able to load any exception dictionary. You can, if you wish, combine both commands to increase the buffer size and load the actual file on one command line. Because you will be issuing the /!Ex parameter, this must be done the first time the driver is loaded each session. For example: DBLTALK /!E5 /EWP This will load the DoubleTalk driver and increase the exception dictionary buffer to five times 512 bytes or two and a half kilobytes. It will then load the file WP.DIC. Parameter: /U This command can be used to remove the synthesizer driver from memory. If you want to completely remove Vocal-Eyes from memory, you can use the following commands. Substitute your own synthesizer driver if you're not using a DECTalk PC. VE /U DECPC /U The first of these commands will remove Vocal-Eyes itself. You must remove it before you can remove the synthesizer driver. After Vocal-Eyes has been removed, the second line will remove the synthesizer driver, DECtalk PC in this case. You can only remove the driver if nothing is loaded after it. This means you must uninstall anything loaded after the driver before you can uninstall the driver itself. As shown above, this is why you must first remove Vocal-Eyes. If the driver can not uninstall itself because you have loaded something after it and have not removed it, the driver will give you an error message and ignore the /U command. REFRESH COMMAND The final command line parameter allows you to fine-tune how the driver communicates to your synthesizer. The driver is constantly sending commands to your synthesizer. However, with interruptability enabled with Vocal-Eyes, those commands may not be noticed by the synthesizer. Therefore, the driver will occasionally send the entire command sequence every so often to ensure the synthesizer is in the correct mode. Actually, after every ten silence commands this happens. If you are using a synthesizer which takes some time to digest the new commands, you may notice an annoying pause at certain times. Therefore, the following command can be used to disable this feature. Be aware however, it may be possible for your synthesizer to get stuck in, let's say, a wrong rate or pitch. Normally, this is not a problem unless you have your keyboard and screen voices different from each other. If they are the same, your synthesizer should always be in the correct mode. Parameter /!Rx (x=0 or 1) 0=disable, 1=enable. The default is enable. This means Vocal-Eyes will occasionally send out the entire command sequence to the synthesizer to guarantee it is set properly. To cause the driver to disable this feature, issue the following command line parameter when the driver is first loaded each session: AUDAPTER /!C1 /!R0 The 'R' stands for refresh and the 0 disables the feature. If your synthesizer issues a slight pause about every ten keystrokes, you might want to give this command a try. When you have typed in the name of your driver followed by whichever switches are appropriate--don't forget to press ENTER after typing in the complete command-- a message should be printed to the screen and also spoken through your synthesizer. The message contains the name of the synthesizer and the version number of the device driver. If you do not wish to hear this message, add a ; (semicolon) to the end of the command line parameter. This signals the driver not to speak the initial message. For example: ECHOPC /!C1 /!B4 /!E5 /!R0 /EBADWORDS; This will load the device driver for the external echo synthesizer using COM1 at 4800 baud, 8 data bits, 1 stop bit and no parity. The exception dictionary buffer is increased to 2-1/2K and the exception dictionary BADWORD.DIC is loaded. Finally the refresh command has been disabled. The welcome message will be printed to the screen but will not be spoken since the command line ended with the ;. PSSB /!L1 This will load the device driver for the Personal Speech System version B synthesizer using LPT1. Since this is being used on a parallel synthesizer, there is no need to specify the baud, data bits, stop bits or parity. The initial message will be printed to the screen and spoken. PROSE4 This simply starts up the driver for the Prose 4000 synthesizer. This is a synthesizer which does not require a port assignment. Now that you understand all of the commands available for each of the synthesizer drivers, lets talk specifically about each of the drivers in turn. INDEXING Vocal-Eyes supports a feature called indexing. Indexing basically is a two way communication between the voice synthesizer and Vocal-Eyes. This two way communication allows Vocal-Eyes to determine exactly what word the synthesizer is speaking at any particular moment. Why is this important? If you are using the read to end feature of Vocal-Eyes and your synthesizer supports indexing, you will notice when you cancel the read to end with the ESCAPE key, Vocal-Eyes routes the applications cursor to the last word you heard. If your synthesizer does not support indexing, the cursor will be anywhere from a few lines to many lines below what you last heard. Since Vocal-Eyes does not know what the synthesizer is speaking, it can't move the cursor to the last spoken word. Indexing is also used if you are using the visual tracking feature located on the screen menu. This feature will highlight each word as it is being spoken. Without the synthesizer telling Vocal-Eyes what word is being spoken, there is no way the highlighted word would be synchronized with the spoken word. Not all synthesizers support indexing. The following list will describe each synthesizer in detail including whether it supports indexing or not. ACCENT.COM (Supports indexing) As we said, this should be used if you plan on using any of the ACCENT synthesizers or the TransPort. The accent boards support all ten rates, pitch values, and volumes but they do not support all twenty-six possible tone settings. They only support ten possible tones. Therefore, Vocal- Eyes tones A through D are mapped into the accent's tone zero. Vocal- Eyes tones E through L are mapped into the accent's tones one through eight respectively. Finally Vocal-Eyes tones M through Z are mapped into the accent tone nine. This means all tones between M and Z will sound the same. The same is true for tones A through D. All other features work as expected. Before you can use the internal accent, you have to load the Aicom device driver. Please refer to your accent user's manual for a complete description of this process. All necessary accent files should be loaded before you attempt to load the ACCENT.COM driver. By default, the internal Accent synthesizer intercepts LPT3. This means you should supply /!L3 on the ACCENT command line. If the data bits do not match, the driver will not function properly and you may notice a severe delay between keystrokes and speaking and/or garbage characters being spoken. The Artic TransPort emulates all Accent commands. Therefore, use this driver for the TransPort synthesizer as well. Note: The external Accent synthesizer will probably default to seven data bits. However, the ACCENT.COM driver for Vocal-Eyes defaults to eight data bits. Therefore, you must either adjust the Accent itself or issue a /!D7 on the ACCENT.COM command line. Please make sure the data bits are identical. APOLLO.COM (does not support indexing) This driver will work with the external Apollo I synthesizers. The synthesizer itself does not require any setup. Simply execute the APOLLO driver. The Apollo supports the ten rate levels and ten pitch settings but does not support the volume or tone settings. The Apollo synthesizers do not support any handshaking. This means when the 8K buffer in the synthesizer fills, any new data will be lost. It is suggested that you not try and send more than 8,000 characters to the synthesizer at one time. With true handshaking, the synthesizer will signal to the computer that its internal buffer is full so to stop sending data. As the synthesizer speaks, the buffer slowly empties. When there is enough room for more data, the synthesizer will tell the computer to start sending data again. Without this ability, the computer has no idea when the synthesizer buffer is full. Again, this will only be a problem if you send more than 8K of data to the synthesizer at once. INAPOLLO.COM (does not support indexing) This driver will work with the internal Apollo I synthesizers. The synthesizer itself does not require any set-up. Simply execute the INAPOLLO driver. The Apollo supports the ten rate levels and ten pitch settings but does not support the volume or tone settings. This driver will work for the default settings of the standard desktop internal. However, you can adjust the port address on this board. Also, the Toshiba model has a different address yet. Therefore, you may have to include the following command line parameter to specify the correct speech base address. /!Bx x=0-2 (0=308HEX, 1=318HEX, 2=Toshiba model) By default, Vocal-Eyes will assume you are using the desktop with the standard address of 308HEX. If you have changed the address to 318H, you must supply: INAPOLLO /!B1 If you are using the internal Toshiba version, you must supply: INAPOLLO /!B2 The Apollo synthesizers do not support any handshaking. This means when the 8K buffer in the synthesizer fills, any new data will be lost. It is suggested that you not try and send more than 8,000 characters to the synthesizer at one time. With true handshaking, the synthesizer will signal to the computer that its internal buffer is full so to stop sending data. As the synthesizer speaks, the buffer slowly empties. When there is enough room for more data, the synthesizer will tell the computer to start sending data again. Without this ability, the computer has no idea when the synthesizer buffer is full. Again, this will only be a problem if you send more than 8K of data to the synthesizer at once. You do not have to specify a speech port. ARTIC.COM (does not support indexing) This is to be used for all ARTIC synthesizers which support the SONIX.COM and TTS.COM. Newer ARTIC cards combine these two programs into one. Every card produced by ARTIC is capable of being used by Vocal-Eyes. However, you may have to purchase their software interface which allows non-Artic products to function with the board. Also, the Votalker board from Votrax should use this driver. You will recall, the Votalker also uses SONIX.COM and TTS.COM. Before you can load the ARTIC.COM driver, you must first load the above two programs. Assuming you do not require any command line parameters the following can be typed from the DOS prompt: SONIX TTS If you try and load the ARTIC.COM driver before loading the above two files, the driver will print "SONIX not loaded!" Simply load SONIX and TTS and reload ARTIC.COM. Those of you familiar with PORTTALK.COM will be happy to know it is not required with Vocal-Eyes. The ARTIC.COM driver will support all of Vocal-Eyes' features. This driver does not require you to specify a speech port. AUDAPTER.COM (supports indexing on serial version only) This driver is to be used with the Audapter synthesizer. The Audapter supports all ten rate, pitch, and volume commands as well as all twenty- six tone settings. BNSPEAK.COM (supports indexing on later Braille 'n Speaks) This driver is to be used with the Braille 'n Speak from Blazie Engineering. Many of the speech features of the Braille 'n Speak have been added over time. Depending on which ROM you have from Blazie you may or may not have all the features required by Vocal-Eyes. If you have an old ROM, you may want to contact Blazie Engineering for an upgrade. With this BNSPEAK.COM driver, you can set it up for two versions, one which only supports the silence command and one which supports all commands including indexing. The default is for all commands. However, if you have an old ROM, you may get garbage spoken when Vocal-Eyes tries to issue a command your Braille 'n Speak does not support. If this is the case, you should issue the following command: BNSPEAK /!V0 when you first load the driver. This will instruct the driver that you have an older ROM and to only use the silence command. If you do not issue this command or issue a /!V1 the driver will assume you have the latest ROM which supports all rate, volume, pitch, and tone commands as well as indexing. In order for the Braille 'n Speak to be used as an external synthesizer, you first must enable the serial port and place it into what is called "Speech box mode." The following can be used to enable the Braille 'n Speak: First perform a warm reset to ensure that you have the default settings: Dots 1, 2, 3, 4, 5, 6-chord Dots 1, 2, 3, 4, 5, 6-chord Next enable the speech port: P-chord Dots 2, 6 Y Next enable the Speech Box Mode Dots 3, 4, 5-chord Dots 1, 2, 3, 4, 5, 6 E-chord At this point, you can load the BNSPEAK.COM driver. If you wish to adjust the rate, pitch, or volume you should use Vocal-Eyes to do so unless you have setup the driver with the /!V0 command. If you have issued /!V0 on the command line, you will have to use the Braille 'n Speak to adjust these settings. If you have any further questions relating to the Braille 'n Speak, please refer to your Braille 'n Speak Users Guide. DECTALK.COM (does not support indexing) This will allow you to use the well-known DECtalk voice synthesizer. This synthesizer probably has the most pleasing voice quality but is by no means a very responsive synthesizer. However, Vocal-Eyes uses it to its full potential. We are sure you will be pleased with the results. If you like the DECtalk quality voice but don't like the unresponsiveness, you may want to consider the DECtalk PC. Of all the DECtalk synthesizers, the DECtalk PC is the most responsive. Since the DECtalk is a completely self-contained external synthesizer, you don't have to load any additional software other than the DECTALK.COM driver. It supports all ten rates and pitch settings but does not support any of the volume settings. Changing the volume from within Vocal-Eyes will have no effect at all. To adjust the volume, you must use the volume control on the DECtalk itself. Also, we decided to map the six different voices the DECtalk offers into the tone settings. Therefore, Vocal-Eyes tones A- H will use Huge Harry. Tone I will use Perfect Paul. Tone J will use Frail Frank. Tone K will use Rough Rita. Tone L will use Beautiful Betty. Finally tones M through Z will use Kit the Kid. The DECtalk is probably the most complicated speech synthesizer available. Because of this, it seems to respond somewhat more slowly than other synthesizers. For example, you may silence the speech but the DECtalk may keep talking for a few seconds before it stops speaking. Rest assured, Vocal-Eyes told the DECtalk to stop speaking the moment you requested it to. PDECTALK.COM (does not support indexing) This will allow you to use the Portable DECtalk voice synthesizer. This synthesizer uses the same voice as the original DECtalk and has the same characteristics. Since the Portable DECtalk is a completely self-contained external synthesizer, you don't have to load any software other than the PDECTALK.COM driver. It supports all ten rates, pitch and volume settings. Also, we decided to map the ten different voices the Portable DECtalk offers into the tone settings. Therefore, Vocal-Eyes tones A through H will use Huge Harry. Tone I will use Perfect Paul. Tone J will use Frail Frank. Tone K will use Rough Rita. Tone L will use Beautiful Betty. Tone M will use Kit The Kid. Tone N will use Uppity Ursula. Tone O will use Doctor Dennis. Tone P will use Whispering Wendy. Finally tones Q through Z will use Variable Val. The Portable DECtalk is slightly different from the original DECtalk. First, it is capable of speaking at a much faster rate. You will notice rate nine for the original DECtalk does not speak nearly as fast as the Portable DECtalk rate nine. The other difference is a flaw in the Portable DECtalk itself. It has a feature which will automatically powerdown the DECtalk if no character has been received through the serial port within a configurable time. This is to conserve battery power. Once a character is received, it was suppose to power up and continue on normally. However, if the Portable DECtalk is in the process of speaking something but nothing has been sent to it through the serial port for some time, it will shut down and not finish speaking the remainder of the text. The PDECTALK driver will disable this feature when it is first loaded. However, if you power down the Portable DECTalk and power it back up again, the auto-powerdown feature may be re-enabled. Therefore, you should never power down the Portable DECtalk unless you are finished with it. The Portable DECtalk is among the most complicated speech synthesizers available. Because of this, it seems to respond somewhat more slowly than other synthesizers. For example, you may silence the speech but the DECtalk may keep talking for a few seconds before it decides to stop speaking. Rest assured, Vocal-Eyes told the DECtalk to stop speaking the moment you requested it to. DECPC.COM (supports indexing) We recommend you use version 4.1C or greater of the DECtalk PC software from Digital Equipment Corporation. Although 4.1 will work, they have made some improvements in the 4.1C version. This will allow you to use the internal DECtalk PC voice synthesizer. This is the latest DECtalk voice synthesizer on the market. It is more responsive than all the other DECtalk synthesizers. However, you will find it still is not as responsive as other synthesizers such as the Sounding Board. You must first load the software that came with your DECtalk PC. Please follows its instruction manual for details. Once this software is up and running, you can simply execute the DECPC.COM program supplied on your Vocal-Eyes disk. You do not need to specify any command line parameters. The DECtalk PC supports all ten rates, pitch, and volume settings. We have mapped the ten different voices into the tone settings. Therefore, Vocal-Eyes tones A through H will use Huge Harry. Tone I will use Perfect Paul. Tone J will use Frail Frank. Tone K will use Rough Rita. Tone L will use Beautiful Betty. Tone M will use Kit the Kid. Tone N will use Uppity Ursula. Tone O will use Doctor Dennis. Tone P will use Whispering Wendy. Finally tones Q through Z will use Variable Val. DBLTALK.COM (supports indexing) This is used for the internal DoubleTalk PC and external DoubleTalk LT synthesizers. The DoubleTalk supports all ten speech rates, pitch, and volume settings. However, it only supports two tone settings. Vocal- Eyes tones A through M are mapped into the first tone setting and Vocal- Eyes tones N through Z are mapped into the second. You will notice using a rate of nine will cause the DoubleTalk PC to talk much faster but use a much higher pitch. This is sort of a hyper-rate. If you do not feel comfortable with this rate, feel free to slow it down to a standard rate. This is only true for rate nine. If you are using the DoubleTalk PC, no other software is needed and you should not specify a speech port on the command line. The driver will find the board automatically. This driver is also used if you are using the external DoubleTalk LT. If you specify a speech port, the driver will assume you are using the LT on the specified port and not bother looking for the internal PC. This driver is actually capable of searching for both the internal and external versions automatically. If you issue the DBLTALK.COM driver with no speech port specified, it will first scan for the internal. If it does not find it, it will scan all serial ports for the external. The driver will use which ever it finds. If neither is located, you will get an error message. ECHOPC.COM (does not support indexing) Although this is called ECHOPC.COM it can be used with the Echo GP as well but it should not be used with the latest release of the ECHO PC. If you are using the newest release of the ECHO PC, you should use the NEWECHO.COM driver discussed later. The older ECHO PC & GP synthesizers, like the DECtalk, are a completely self-contained external synthesizer. You don't have to load any software other then the ECHOPC.COM driver. The Echos support all ten pitch settings and volume settings. However, it only offers two speech rates. Therefore, Vocal-Eyes speech rates zero through four will issue the slower of the two speeds and rates five through nine will issue the faster of the two. Also, the Echos do not support any of the tone settings. Changing the tone value in Vocal-Eyes will have no effect at all on the speech. Unlike the DECtalk, the external Echos will stop speaking immediately when requested to. However, they have a side effect of squeaking or "hiccuping." This can be annoying if you are using interruptability. You may wish to turn interruptability off when working with Echo. INTECHO.COM (does not support indexing) This driver will work with the Echo PCII, PC+, MC, or 1000. All four of these are treated the same. The only difference is the machines they work on. In order to use these synthesizers, you need to load the program TALK.EXE. This is supplied with your synthesizer. Please refer to your synthesizer manual if you have any questions on how to load this program. If you try and load the INTECHO.COM driver prior to loading the TALK.EXE program, the driver will print "The ECHO software was not found." Simply load TALK.EXE and reload INTECHO.COM. One of the options the TALK program offers is the ability to choose the key you would like to use to silence the speech. Do not assign a key for this feature. Allow Vocal-Eyes to handle this. These synthesizers support all ten rates, pitch settings, and volume levels. However, they do not support any of the tone settings. Vocal-Eyes' tone options will have no effect at all. These synthesizers require the complete attention of the machine while they are speaking text. Therefore, you may notice some sluggishness while using any of these synthesizers. These synthesizer do not require a port assignment. Vocal-Eyes will automatically find the synthesizer. NEWECHO.COM (does not support indexing) This driver will work with the latest release of the Echo PC. This synthesizer supports all ten rates, pitch settings and volume levels. However, it does not support any of the tone settings. Vocal-Eyes' tone options will have no effect at all. This synthesizer has a very small buffer and therefore you may notice some sluggishness while using it. LITETALK.COM (supports indexing on the serial version only) The LiteTalk is very much like the DoubleTalk LT synthesizer. Like the DoubleTalk LT, the LiteTalk supports all ten speech rates, pitch, and volume settings. However, it only supports two tone settings. Vocal- Eyes tones A through M are mapped into the first tone setting and Vocal- Eyes tones N through Z are mapped into the second. You will notice using a rate of nine will cause the LiteTalk to talk much faster but use a much higher pitch. This is sort of a hyper-rate. If you do not feel comfortable with this rate, feel free to slow it down to a standard rate. This is only true for rate nine. MVOICE.COM (does not support indexing) This will allow you to use the MultiVoice voice synthesizer. This synthesizer uses the same voice as the original DECtalk and has very similar characteristics. Since the MultiVoice is a completely self-contained external synthesizer, you don't have to load any software other than the MVOICE.COM driver. It supports all ten rates, pitch, and volume settings. Also, we decided to map the ten different voices the MultiVoice offers into the tone settings. Therefore, Vocal-Eyes tones A through H will use Huge Harry. Tone I will use Perfect Paul. Tone J will use Frail Frank. Tone K will use Rough Rita. Tone L will use Beautiful Betty. Tone M will use Kit The Kid. Tone N will use Uppity Ursula. Tone O will use Doctor Dennis. Tone P will use Whispering Wendy. Finally tones Q through Z will use Variable Val. The MultiVoice is slightly different from the original DECtalk. First, it is capable of speaking at a much faster rate. You will notice rate nine for the original DECtalk does not speak nearly as fast as the MultiVoice rate nine. The DECtalk class of synthesizers are among the most complicated speech synthesizers available. Because of this, the MultiVoice seems to respond somewhat more slowly than other synthesizers. For example, you may silence the speech but the MultiVoice may keep talking for a few seconds before it decides to stop speaking. Rest assured, Vocal-Eyes told it to stop speaking the moment you requested it to. PROSE4.COM (supports indexing) This should be used if you wish to use Vocal-Eyes with the PROSE-4000 synthesizer. Even though this is an internal synthesizer, nothing else is required to be loaded. The PROSE4.COM contains all necessary information. If you have the PROSE.COM utility, you should not use it. This would interfere with PROSE4.COM. All ten rates, pitch settings and volume levels are supported. However, the PROSE does not support the tone settings. Therefore, changing Vocal-Eyes' tone settings will have no effect at all. Like the DECtalk, if you silence the speech it may take a moment for the Prose to stop speaking. Rest assured Vocal-Eyes sent the proper commands to the PROSE. Sometimes, it takes it a moment or so to actually silence the speech however. A speech port assignment is not required for this synthesizer. However, if you have changed the on-board IRQ from the default of two or have changed the speech port address from the default of 300 HEX, you must supply a command to the driver indicating the change. /!Ix - x = 2-7. This parameter will supply to the PROSE4 driver which IRQ the Prose 4000 has been setup for. By default the driver will assume IRQ two. This is the default for the Prose 4000 board. If you have changed the IRQ from two, you must supply this parameter. For example, if the IRQ has been changed to seven, you must supply: PROSE4 /!I7 /!Bx - x = 0-4. (0=300HEX, 1=308HEX, 2=310HEX, 3=318HEX, 4=320HEX) This parameter will supply to the PROSE4 driver the speech base address. By default the driver will assume 300 HEX. This is the default for the Prose 4000 board. If you have changed the base address from 300 HEX, you must supply this parameter. For example, if the speech address has been changed to 318 HEX, you must supply: PROSE4 /!B3 If necessary, you can combine both the IRQ command and the speech base address. This driver does not require a port address. PSSA.COM (does not support indexing) This is a self contained external synthesizer so nothing other then PSSA.COM needs to be loaded. This should be used for the Personal Speech System version A only. This is a relatively old synthesizer and was not really designed for the blind community. You will notice as you speed the voice up, the voice becomes very choppy and difficult to understand. It supports all ten rates, pitch settings, and volume levels. However, it does not support any tone settings. Changing Vocal-Eyes' tone option will have no effect whatsoever. PSSB.COM (does not support indexing) This is just like the PSSA.COM except it is for Personal Speech System version B. It has the same features and limitations as version A. READEDGE.COM (does not support indexing) This will allow you to use the DECtalk synthesizer of the Reading Edge. This synthesizer uses the same voice as the original DECtalk and has very similar characteristics. In order to use the Reading Edge as a voice synthesizer, you must first set it up using the keypad for "receive from screen reader." Once this is done, you can treat it as an external synthesizer. It supports all ten rates, pitch, and volume settings. Also, we decided to map the ten different voices the DECtalk offers into the tone settings. Therefore, Vocal-Eyes tones A through H will use Huge Harry. Tone I will use Perfect Paul. Tone J will use Frail Frank. Tone K will use Rough Rita. Tone L will use Beautiful Betty. Tone M will use Kit The Kid. Tone N will use Uppity Ursula. Tone O will use Doctor Dennis. Tone P will use Whispering Wendy. Finally tones Q through Z will use Variable Val. The Reading Edge is slightly different from the original DECtalk. First, it is capable of speaking at a much faster rate. You will notice rate nine for the original DECtalk does not speak nearly as fast as the Reading Edge rate nine. The DECtalk is probably the most complicated speech synthesizer available. Because of this, it seems to respond somewhat more slowly than other synthesizers. For example, you may silence the speech but the Reading Edge may keep talking for a few seconds before it decides to stop speaking. Rest assured, Vocal-Eyes told the DECtalk to stop speaking the moment you requested it to. VEASP.COM (does not support indexing) This driver is to be used with the ASP synthesizer. It supports all of the Vocal-Eyes commands. GENERIC.COM (does not support indexing) This can be used if the synthesizer you are using is not directly supported with its own driver. The synthesizer must, however, be communicated to through one of the four COM ports or three LPT ports. You will need to setup your synthesizer as required in its users manual. Once all its required software is loaded, simply load the GENERIC.COM driver. At this point, you should be able to use your synthesizer with Vocal-Eyes. However, you will not have any control over the rate, pitch, volume, tone and worst of all, you will not be able to silence the voice from Vocal-Eyes. If your synthesizer supports any of these features, you can tell the generic driver how to use it. When you first load GENERIC.COM, it looks for a file called "GENERIC.DAT." If this file does not exist, it will default to not supporting any features at all. However, if it does exist, the information will be used giving Vocal-Eyes control it normally would not have. The format of the GENERIC.DAT file is very important. There is not much room for error. First of all, the file must be in a standard ASCII format. It must contain exactly fifty-nine lines. The first ten lines relate to the ten rate commands, followed by the ten pitches, ten volumes, twenty-six tones, silence string, initialization string and finally a list of punctuation characters that can be sent to the synthesizer. This all adds up to fifty- nine lines. Even if your synthesizer does not support all features, you must make an entry for all fifty-nine lines. If there is a problem in the format of the file, the GENERIC driver will print "Format error in GENERIC.DAT" and abort the load process. You should correct the problem and attempt to load the driver again. In the following example, we will create a GENERIC.DAT driver for the Sounding Board. Of course, this is not needed for the Sounding Board but it will make a good demonstration. The first ten lines of the file relate to rates zero through nine. Each rate command must be on a line by itself. First you need to figure out what command your synthesizer needs for the ten rates. The Sounding Board uses CTRL-ER0-9. Therefore, the first ten lines of your file would look like this: /5/R0 /5/R1 /5/R2 /5/R3 /5/R4 /5/R5 /5/R6 /5/R7 /5/R8 /5/R9 You are probably wondering what the /5/ is used for at the beginning of each line. Many word processors or text editors will not let you place ESCAPE or control characters directly in a file. Therefore, whenever you wish to enter one of these codes, use the format: /n/ Where n is the ASCII value of the character you wish entered. For example, to enter an 'ESCAPE Control-Z A' as one of your entries, the following line could be used: /27//26/A The /27/ is the ASCII value of the ESCAPE. /26/ is the ASCII value of Control-Z and of course the A can be enter directly. What if you really wanted to enter a "/" in your string? Simply place two slashes together. For example: /5////6/B This line would place a Control-E (/5/) followed by a / (//) followed by a Control-F (/6/) and finally a B (B). So in our Sounding Board example, the /5/ is simply the required Control-E character. Now that we have defined the first ten lines as the rate, the next ten lines need to define pitches zero through nine. The Sounding Board requires CTRL-EP0-9. Therefore, the following would be lines eleven through twenty. /5/P0 /5/P1 /5/P2 /5/P3 /5/P4 /5/P5 /5/P6 /5/P7 /5/P8 /5/P9 Now, we need to define the ten possible volumes zero through nine. The Sounding Board requires CTRL-EV0-9. Therefore, the following would be lines twenty-one through thirty: /5/V0 /5/V1 /5/V2 /5/V3 /5/V4 /5/V5 /5/V6 /5/V7 /5/V8 /5/V9 The next twenty-six lines need to define the twenty-six possible tones A through Z. The Sounding Board requires CTRL-EA-Z. Therefore, the following would be lines thirty-one through fifty-six: /5/TA /5/TB /5/TC . . . /5/TX /5/TY /5/TZ The next line in the GENERIC.DAT file should be the string which, when sent to the synthesizer, will silence the speech. On the Sounding Board this is CTRL-X. Therefore, the following would be line fifty-seven in our GENERIC.DAT file. /24/ Remember, /24/ simply means ASCII 24 which is a CTRL-X. The next line contains the initialization string. This string will only be sent to your synthesizer one time. You can use this feature to make sure certain settings are always set the way they should be. The Sounding Board really does not require such a string. Therefore, leave this line completely blank. Line fifty-eight in our GENERIC.DAT file will simply be a blank line. The final line holds all the punctuation characters which can be sent to the synthesizer. On the Sounding Board this is every character. However, for example, the Personal Speech System uses both the ! (exclamation point) and the @ (at sign) for command strings. If you send these characters, the synthesizer will give an error. So for the final line fifty-nine, you would type the following if you are using the Personal Speech System: ~`()_^-{}|:[]"?\;,.' Notice the leading space. Make sure you include this character. You should try and send as many punctuation characters as possible so your synthesizer can use them for intonation. Notice the @ and ! have been left out. There are a few others which have been left out since they also can be used as command characters. Sounding Board users could have included each and every punctuation character. Now that you have all fifty-nine lines typed in your file, save the file in a standard ASCII format. If you have no format errors, each time you load GENERIC.COM, it will load the GENERIC.DAT file and use the information specified. It may take a bit to get this file setup but once it is done, you will no longer have to worry about it and your synthesizer will work forever.