home *** CD-ROM | disk | FTP | other *** search
- VOICES EDITOR !VoiceEdit
- ========================
-
- !VoiceEdit allows you to:
- - select a voice to be used by the Speak module
- - edit existing voices and create new ones.
-
- Load the Speak module (eg. by running the !Speak application) before
- running !VoiceEd.
-
- Click on the !VoiceEd icon bar icon to open the Voice Editor window
- and to edit the “voices” file which is being used by the Speak
- module.
-
- Voices are numbered from 1 to 31. In addition, Voice number 0 is
- used as the default voice.
-
- Select a voice either by
- - clicking with the Menu button over the “Voice” field.
- (this only allows named voices to be selected)
- - clicking the arrow icons labelled “Next”.
-
- The functions of the various fields in the Voice Editor window are
- now described:
-
-
- VOICE. This indicates the currently selected voice number. Clicking
- with Menu over this field will display a menu from with any names
- voice can be selected.
-
- NAME. This is a writable field which is used to give a name to the
- voice.
-
- RENUMBER. Enter a number (0 to 31) to renumber the current voice to
- this number. The voice which previously had this number is
- swapped with this one.
-
- PHONEMES. This specified which phoneme set is used by the voice.
- The phoneme set determines the “accent” used by the voice. Click
- with the Menu button to show a list of the phoneme sets which are available.
-
- DICTIONARY. This field is used to specify certain pronunciation
- rules which may differ for different voices. See below for
- details.
-
- PITCH BASE. Determines the pitch of the voice. A low number gives a
- low pitched voice.
-
- PITCH RANGE. Determines the range of pitches used by the voice. A
- value of zero will give a monotone. This field should generally
- have a value which is similar to that of PITCH BASE.
-
- PITCH DROP. Not currently used.
-
- INTONATION. Not currently used.
-
- SPEED. Determines the normal speed of the voice. A value of 256
- gives the standard speed.
-
- FUNDAMENTL. This field can be used to adjust the strength of the
- lowest frequency harmonic of the voice. A larger value will give
- a deeper, more mellow, sound.
-
- CONSONANTS. Determies the strength of consonants (such as “t” and
- “s”). A lower value can be used to reduce sibilance. It does
- not affect “liquid” or “sonorant” consonants (such as m, n, r, l,
- y, w).
-
- FLUTTER. Introduces pseudo-random pitch variations into the voice
- while it is speaking. Values 0 to 6. A low value (above zero)
- can give a slight quiver to make the voice sound more natural.
- A larger value gives a hoarse or croaky voice.
-
- HISS. Adds some hiss to the vowel sounds. This was intended to make
- the voice sound less mechanical, but is not particularly
- successful.
-
- INTERPOLATE WAVE. Reduces the amount of processing power needed, by
- using some interpolation in calculating the speech waveform. It
- reduces the quality, but is needed when running on a low speed
- processor (ARM 2).
-
- FORMANTS - AMP. Adjusts the strength of the various frequency bands
- of the voice. This can be used as a form of tone control. For
- example, decreasing the strength of the higher formants will give
- a more muffled sound. The default values are all 32.
-
- FORMANTS - SHIFT. This simulates altering the shape of the mouth by
- shifting the various formants up or down in frequency. This
- alters the voice quality. For example, increasing the values for
- the higher formants will give the effect of a smaller mouth.
- The default values are: 16, 20, 44, 40, 40, 50, 64.
-
- FORMANTS - DEFAULT. This sets all the Formant-Amp and Formant-Shift
- values to their defaults.
-
- DEFAULTS. Sets all the fields to their defaults.
-
- COPY. Copies the values from another voice into this one. It can be
- used to build a new voice which is based on an existing one.
-
- SAVE. Saves the current values to file, and instructs the Speak
- module to start using this voice.
-
-
-