home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Audio Version 4.94
/
audioversion4.94knowledgemediaresourcelibraryoctober1994.iso
/
amiga
/
speech
/
spechty2.doc
< prev
next >
Wrap
Text File
|
1991-12-28
|
6KB
|
136 lines
---------------------------------------------------------------------------
SPEECHTOY2
by Chris P. Demiris
---------------------------------------------------------------------------
---------------------------------------------------------------------------
Introduction
---------------------------------------------------------------------------
**** This requires version 2.0 or higher of the Amiga O.S. ****
This is the first release of a new SpeechToy program that allows you to
play with all of the features of the narrator device under 2.0. The speech
abilities of the Amiga have been greatly enhanced with 2.0, and this program
will give you a chance to see just what has been added.
I didn't originally intend to release it until it was finished, but I
thought everyone would enjoy it even as is.
This program is freely distributable, as long as it is kept intact with this
document and the program remains unaltered.
If you have any suggestions or problems, let me know:
Genie: C.DEMIRIS1
BIX: cdemiris
or on the Armadillo Brothers BBS of Salt Lake: (801) 484-2766
---------------------------------------------------------------------------
Using SpeechToy2
---------------------------------------------------------------------------
Double-click on SpeechToy2 icon to start the program.
Notice that many of the gadgets have keyboard equivalents. Note that for
the slider gadgets and the cycle gadgets, these keyboard equivalents move
forward when the unshifted key is pressed, and backwards when the shifted
key is pressed.
The TEXT string gadget allows you to enter any english text to be spoken.
After text is entered, it is translated into phonemes in the PHONE gadget.
The PHONE gadget allows you to change the actual phonemic text to be spoken
by the narrator. Here you can adjust stressed syllables, and even change
phonemes entirely.
Click on the SPEAK gadget or press 'S' to have the Amiga speak whatever text
is in the PHONE string.
The DEFAULTS gadget returns all of the voice settings to their default
values.
The remainder of the gadgets adjust the Amiga voice. They are described
briefly below. A more detailed description will come with the final
release. If you want more info now, get the Third Edition of the Amiga RKM
Devices Manual. New features for 2.0 are marked with a *.
Speaker: This adjusts which speaker(s) the voice comes out of.
Sex: Self-explanatory (I hope).
Mode: Natural - The narrator device uses punctuation marks
to adjust speech.
Robotic - No stress is added to any syllables.
Manual - This narrator will use only the stresses
indicated in the phonemic text, and will not
make adjustments for punctuation.
Rate: Speaking rate in words-per-minute.
Volume: Sets the volume of the speaking voice.
Pitch: The baseline pitch of the speaking voice in Hertz.
*Perturb: Non-zero values in this field cause random modulation
of the pitch. This can cause the pitch to shake
similar to an elderly person's voice.
*Enthus: Enthusiasm. This causes greater pitch changes on
accented syllables.
*F1Adj - *F3Adj: Formant frequencies.
Rather than explain these, just try them. Negative
values are more useful, since positive values can
cause clipping of sound.
*A1Adj - *A3Adj: Amplitudes of formants.
Once again, just try different values.
*Articulate: Normally our articulators (jaw, tongue, lips, etc.)
move from one position to the next smoothly. This
field allows you to adjust how smoothly the
transitions are made. Smaller values will cause
shorter transitions, making more clipped speech, while
larger values will create slurring effects.
*AVBias, *AFBias: Adjusts the clarity of voiced and unvoiced
sounds. Try a higher AFBias and you'll notice that
the consonants are voiced more clearly.
*Cent% and *Centphon: These are the really neat features. You can create
regional accents easily with this feature. When
any vowel is spoken, the narrator "pulls" the
spoken vowel towards the target vowel specified in
Centphon. The amount which the original vowel is
pulled is specified in Cent%.
---------------------------------------------------------------------------
*** COMING SOON ***
---------------------------------------------------------------------------
The final release of this little program will contain the following
additional features:
* A graphic face that moves while speaking.
* The ability to save a set of parameters as a 'Character'.
* The ability to load and save Character databases.
* The ability to speak text files, including files with special codes
to select parameters and Characters.
* A full Arexx port to allow control from other programs.
* ??? If you think of anything else, let me know.