A phrase consists of a series of syllables, separated by spaces, written in X-SAMPA notation. For example, to sing the words "Happy birthday to you," enter the phrase h{ pi b3T dEj tu ju. Each successive note is sung on the next syllable. When it reaches the end of the current phrase, it loops back to the start again.

You can enter up to 128 phrases. To select the phrase to sing, click on it in the table. To automate it in a DAW, set the parameter "Selected Phrase" to the index of the phrase to sing. You can use the "Advance Syllable" parameter to temporarily disable advancing to the next syllable. When this option is off, every note continues to repeat the same syllable until it is turned back on. Alternatively, if you play legato (see below), it extends a single syllable across multiple notes. This is useful when a phrase is sung repeatedly, but the number of notes each syllable is used for changes between repetitions.

A syllable consists of

• Zero or more initial consonants
• One or more vowels
• Zero or more final consonants

When a syllable contains multiple vowels, one of them is the main vowel that is held for the duration of the note. The others are transient vowels that are touched only briefly at the beginning or end. Chorus Ex Machina tries to make a reasonable guess about which one should be the main vowel. If it guesses wrong, you can indicate the main vowel by adding a - immediately after it. Compare A-i and Ai-.

The following vowels are supported.

Symbol	Pronunciation	Notes
A	father
E	met
I	kit
N	thing
O	off
U	foot
V	strut
Y	hübsch (German)
a	dame (French)
e	beauté (French)
i	be	Sometimes changed to y when singing
l	lot
m	him
n	nap
o	veau (French)
u	boot
y	über (German)
{	cat
@	arena
&	skörd (Swedish)
2	deux (French)
3	nurse	See note on the letter R below
9	neuf (French)

The following consonants are supported.

Symbol	Pronunciation	Notes
C	ich (German)
D	this
S	ship
T	thin
Z	vision
b	bed
d	dog
f	five
g	game
h	house
j	you
k	cut
p	pit
r	perro (Spanish)	See note on the letter R below
s	seem
t	tin
v	vest
w	west
x	loch
z	zoo
4	caro (Spanish)	See note on the letter R below
dZ	jeans
ts	Zeit (German)
tS	cheap

Note on the letter R

Several different sounds can be used for the letter R, depending on the language and context.

• The "flipped R" 4 is the most commonly used R sound in choral singing. When a spoken word uses a different sound, it often is changed to 4 when sung.
• The "rolled R" r is common in Spanish and Italian. It often is changed to 4 in choral singing.
• For the non-rolled R common in English, use the vowel 3.
• In spoken French, R is often pronounced as the "uvular trill" R\. This sound is not commonly used in choral singing, and is not supported by Chorus Ex Machina. It usually is changed to 4, or occasionally to r.
• When R appears as a final consonant, it sometimes is omitted. For example, the Latin word "eterna" is often sung as E tE nA. Alternatively, it may be replaced by a neutral vowel such as @ or 9.

Choosing vowels

Your choices of what vowels to use are important for creating a natural sound. The vowels used in singing, and especially in choral singing, are often different from the ones used when speaking. There are many vowels that sound very similar to each other. Some examples are

• A and a
• i and y
• U and u
• @, V, and 9

Which one sounds best depends not only on the word being sung but also on the pitch, the voice part, the surrounding words, and the style of music. Consider the English word "I". It can be pronounced in many subtly different ways: Ai, Ae, AI, aI, etc. If a note sounds harsh or unnatural, experiment with different vowels to find the combination that sounds best.

This is especially important for high soprano notes, which can sound unpleasantly bright when sung on closed vowels. A singer will often change to a more open vowel for these notes, such as changing i to e or I to E.

Placing consonants and transient vowels

Consonants can appear both at the beginning and the end of a syllable. Many consonants sound slightly different depending on the position. Initial consonants tend to be slightly louder and longer than final consonants.

When a consonant appears between two vowels that are sung legato (see below), you can choose which position to put it in. For example, the word "open" can be pronounced either as op En or o pEn. Each one involves the same series of sounds, but the "p" is pronounced differently.

The same is true of transient vowels. Consider the word "around". It can be pronounced either as @3 aUnd or @ 3aUnd. The latter puts more emphasis on the 3, which corresponds to the letter "r".

Playing

Chorus Ex Machina is a monophonic instrument: each instance plays only one note at a time. To create splits within a section use multiple tracks, each with its own instance of the plugin. If you press a new note before releasing the previous one, it is played legato, smoothly blending between them. Smooth transitions are only possible between vowels, however. If there are any consonants in between, they necessarily create a break in the sound, but it will shorten the consonants to minimize the gap.

You can optionally add an accent to the beginning of each separated (non-legato) note. When this option is enabled, the strength of the accent is determined by the key velocity.

The Vowel Delay parameter is an important tool for creating accurate timing, especially in fast passages. When a syllable begins with one or more consonants, the consonants are normally sung before the beat, and the vowel begins right on the beat. This requires the start of the note to be shifted earlier. How far it must be shifted depends on the number and type of consonants, possibly including final consonants from the previous note. Adjusting note start times by hand is slow and imprecise.

Instead you can tell it to delay the start of the first vowel by up to 250 ms. This allows the vowel to always follow the start of the note by exactly the same amount, regardless of what consonants precede it. The ends of notes are shifted by the same amount to avoid shortening notes or creating a break in the sound. When singing legato, the first vowel of the new note begins on the beat. When ending non-legato notes, the first final consonant begins on the beat. This gives precise timing with no extra effort. If you need to synchronize the chorus with other instruments, you can adjust the MIDI offset in your DAW to shift the start of the notes earlier by the same amount as the delay, so vowels begin exactly on the beat.

There are several parameters you can automate in a DAW to control the performance.

• Selected Phrase. The index of the phrase to sing.
• Dynamics. How loud to sing. This is not simply a volume control. Voices sound different depending on how loudly they are singing.
• Vibrato. The amount of vibrato in the voices.
• Intensity. How relaxed or intense the voices sound.
• Brightness. The overall vowel color. Reduce this to get a darker sound.
• Consonant Volume. How loud consonants are relative to vowels. This can be automated for fine control over individual consonants.
• Attack Rate. How quickly notes reach full volume.
• Release Rate. How quickly the sound stops at the end of a note.
• Stereo Width. How widely the singers are spread out in space.
• Exciter Strength. The strength of the harmonic exciter effect to apply to the output. This increases the amount of high frequency harmonics. When used with restraint, it can improve clarity and produce a fuller sound.
• Time Spread. The amount of delay between voices in the chorus.
• Vowel Delay. The delay in milliseconds from the start of each note to the beginning of the first vowel.
• Accent. Whether to add an accent to each note based on its velocity.
• Advance Syllable. Whether to advance to the next syllable in the phrase for the next note.