eSpeak: Phoneme tables

INTONATION

In eSpeak's standard intonation model, a "tune" is applied to each clause depending on its punctuation. Other intonation models may be used for some languages, such as tone languages.

Named tunes are defined in the text file: phsource/intonation. This file must be compiled for use by eSpeak by using the espeakedit program, using the menu option: Compile -> Compile intonation data.

Clauses

The tunes which are used for a language can be specified by using a tunes statement in a voice file in espeak-data/voices. eg:

tunes s1 c1 q1 e1

It's parameters are four tune names which are used for clauses which end in:

Full-stop.
Comma.
Question mark.
Exclamation mark.

A clause consists of the following parts:

Pre-head.
These are any unstressed syllables before the first stressed syllable.
Head.
This is the part from the first stressed syllable up to the last syllable before the nucleus.
Nucleus.
This is stressed syllable which is the focus of the clause. eSpeak chooses the last stressed syllable of the clause.
Tail.
These are the syllables after the nucleus.

Tune definitions

Here is an example tune definition from the file phsource/intonation.

tune s1
prehead   46 57
headenv   fall 16
head       4 80 55 -8 -5
headextend 0 63 38 13 0
nucleus  fall 70 18 24 12
nucleus0 fall 64 8
endtune

It contains:

tune <tune name>

Starts the definition of a tune. The tune name can be used in a tunes statements in voice files.

endtune <tune name>

Ends the definition of a tune.

prehead <start pitch> <end pitch>

Gives the pitch path for any series of unstressed syllables before the first stressed syllable.

headenv <envelope> <height>

Gives the pitch envelope which is used for stressed syllables in the head (before the nucleus), including onset and headlast syllables if these are specified. height gives a pitch range for the envelope.

head <steps> <start pitch> <end pitch> <unstressed start> <unstressed end>

start pitch and end pitch give a pitch path for the stressed syllables of the head. steps is the maximum number of stressed syllables for which this applies. If there are additional stressed syllables, then the headextend statement is used for them.

unstressed start and unstressed end give a pitch path for unstressed syllables between two stressed syllables. Their values are relative to the pitch of the previous stressed syllable. Values are usually negative, meaning that the unstressed syllables have lower pitch than the previous stressed syllable.

headextend <percentage list>

If the head contains more stressed syllables than is specified by steps, then percentage list is used. It contains up to 8 numbers which are used repeatedly for the additional stressed syllables. A value of 0 corresponds to the lower the start pitch and end pitch values of the head statement. 100 corresponds to the higher value. Negative values and values greater than 100 are allowed.

nucleus <envelope> <top pitch> <bottom pitch> <tail start> <tail end>

This gives the pitch envelope and pitch range of the last stressed syllable of the clause. tail start and tail end give a pitch path for the unstressed syllables which are after the last stressed syllable.

nucleus0 <envelope> <top pitch> <bottom pitch>

This is used instead of nucleus if there are no unstressed syllables after the last stressed syllable. In this case, the pitch changes of the nucleus and the tail and both included in the nucleus.

The following attributes may also be included:

onset <pitch> <unstressed start> <unstressed end>: This specifies the pitch for the first stressed syllable of the head. If the onset statement is present, then the head statement used for the stressed syllables after the first.
headlast <pitch> <unstressed start> <unstressed end>: This specifies the pitch for the last stressed syllable of the head (i.e. the stressed syllable before the nucleus).