@c -*- coding: utf-8; mode: texinfo; -*-
@ignore
    Translation of GIT committish: FILL-IN-HEAD-COMMITTISH

    When revising a translation, copy the HEAD committish of the
    version that you are working on.  See TRANSLATION for details.
@end ignore


@node Pitches
@section Pitches

@lilypondfile[ragged-right,line-width=16\cm,staffsize=16,quote]
{pitches-headword.ly}

This section discusses how to specify the pitch of notes.  There
are three steps to this process: input, modifying, and output.

@menu
* Writing pitches::             
* Changing multiple pitches::   
* Displaying pitches::          
@end menu


@node Writing pitches
@subsection Writing pitches

This section discusses how to input pitches.

@menu
* Absolute octave entry::       
* Relative octave entry::       
* Accidentals::                 
* Note names in other languages::  
@end menu


@node Absolute octave entry
@unnumberedsubsubsec Absolute octave entry

@cindex Pitch names
@cindex pitches

@c don't use @samp here because that might get confused with
@c octave marks
A pitch name is specified using lowercase letters @code{a} through
@code{g}.  An ascending C-major scale is engraved with

@lilypond[quote,fragment,verbatim,ragged-right]
\clef bass
c d e f g a b c'
@end lilypond

There are two different ways to enter these note names with regard
to octave placement: absolute and relative.  The relative mode
will in most cases be the most convenient, but the absolute mode
is more fundamental and will be presented first.

@cindex Absolute
@cindex Absolute octave specification

@c don't use @samp here because that might get confused with
@c octave marks
In absolute mode, every note is placed explicitly in a certain
octave.  The note names @code{c} to @code{b} are engraved in the
octave below middle C, as in the previous example.

@lilypond[quote,fragment,verbatim,ragged-right]
\clef treble
c1
\clef bass
c1
@end lilypond

@funindex '
@funindex ,

If a note should be in another octave than this, an octave
specification must be given.  It takes the form of a series of
single quote (@code{'}) or comma (@code{,}) characters.  Each
@code{'} raises the pitch by one octave; each @code{,} lowers the
pitch by an octave.

@lilypond[quote,ragged-right,fragment,verbatim]
\clef treble
c' c'' e' g d'' d' d c
\clef bass
c, c,, e, g d,, d, d c
@end lilypond

@seealso

Snippets: @lsrdir{pitch}

Program reference: @internalsref{LedgerLineSpanner},
@internalsref{NoteHead}.


@node Relative octave entry
@unnumberedsubsubsec Relative octaves entry

@cindex Relative
@cindex Relative octave specification
@funindex \relative

The absolute mode requires a lot of octave specifications, which
not only becomes tedious but also easily leads to errors.
Therefore, it is usually more convenient to let notes be entered
relative to the previous note.  This is done with the relative
mode.  

@example
\relative @var{startpitch} @var{musicexpr}
@end example

In relative mode, each note is assumed to be as close to the
previous note as possible.  This means that the octave of notes
that appear in @var{musicexpr} are calculated as follows:

@itemize
@item 
The pitch of the first note is relative to @code{\relative
@var{startpitch} }.  

@item
If no octave changing marks are used, the basic interval between a
note and the one that precedes it is always taken to be a fourth
or less.

@item
This distance is determined without regarding alterations or the
actual sounding pitches; a @code{fisis} following a @code{ceses}
will be put above the @code{ceses}.  In other words, a
doubly-augmented fourth is considered a smaller interval than a
diminished fifth, even though the doubly-augmented fourth spans
seven semitones while the diminished fifth only spans six
semitones.

@item
The octave changing marks @code{'} and @code{,} can be added to
raise or lower the pitch by an extra octave from the basic
interval.

@item
Multiple octave changing marks can be used.  @code{''} will raise
the pitch by two octaves.

@end itemize 

Here is the relative mode shown in action:

@lilypond[quote,fragment,ragged-right,verbatim]
\relative c'' {
  b c d c b c bes a
}
@end lilypond

Octave changing marks are used for intervals greater than a
fourth:

@lilypond[quote,ragged-right,fragment,verbatim]
\relative c'' {
  c g c f, c' a, e''
}
@end lilypond

A note sequence without a single octave mark can nevertheless span
large intervals:

@lilypond[quote,ragged-right,fragment,verbatim]
\relative c {
  c f b e a d g
}
@end lilypond

If the preceding item is a chord, the first note of the chord is
used as the reference point for the octave placement of a
following note or chord:

@lilypond[quote,ragged-right,fragment,verbatim]
\relative c' {
  c <c e g>
  <c' e g>
  <c, e' g> c
}
@end lilypond

The @var{startpitch} (after @code{\relative}) is a note name in
absolute mode.  In principle it can be any note name, but common
practice is to use @code{c} in various octave positions. 

When octaves are specified as above by adding @code{'} and
@code{,} to pitch names, it is easy to accidentally put a pitch in
the wrong octave.  The relative octave mode prevents these errors
since most of the time it is not necessary to indicate any octaves
at all.  Furthermore, in absolute mode, a single mistake may be
difficult to spot; in relative mode, a single error puts the rest
of the piece off by one octave.

@seealso

Snippets: @lsrdir{pitch}


@refbugs

The relative conversion will not affect @code{\transpose},
@code{\chordmode} or @code{\relative} sections in its argument.
To use relative within transposed music, an additional
@code{\relative} must be placed inside @code{\transpose}.

@c DEPRECATED
If no @var{startpitch} is specified for @code{\relative}, then
@code{c'} is assumed.  However, this is a deprecated option and
may disappear in future versions, so its use is discouraged.



@node Accidentals
@unnumberedsubsubsec Accidentals

@warning{New users are sometimes confused about accidentals and
key signatures.  In LilyPond, note names are the raw input; key
signatures and clefs determine how this raw input is displayed.
An unaltered note like @code{c} means @q{C natural}, regardless of
the key signature or clef.  For more information, see
@rlearning{Accidentals and key signatures}.}

@cindex note names, Dutch
@cindex note names, default

A @notation{sharp} pitch is made by adding @code{is} to the name,
and a @notation{flat} pitch by adding @code{es}.  As you might
expect, a @notation{double sharp} or @notation{double flat} is
made by adding @code{isis} or @code{eses}.  This syntax derived
from note naming conventions in Nordic and Germanic languages,
like German and Dutch.  To use other names for accidentals, see
@ref{Note names in other languages}.

@lilypond[quote,ragged-right,fragment,verbatim,relative=2]
a2 ais a aes
a2 aisis a aeses
@end lilypond

@noindent
These are the Dutch note names.  In Dutch, @code{aes} is
contracted to @code{as}, but both forms are accepted.  Similarly,
both @code{es} and @code{ees} are accepted:

@lilypond[fragment,quote,ragged-right,verbatim,relative=2]
a2 as e es
@end lilypond

A natural will cancel the effect of an accidental or key
signature.  However, naturals are not encoded into the note name
syntax with a suffix; a natural pitch is shown as a simple note
name:

@lilypond[quote,ragged-right,fragment,verbatim,relative=2]
a4 aes a2
@end lilypond


@cindex accidental, reminder
@cindex accidental, cautionary
@cindex accidental, parenthesized
@cindex reminder accidental
@funindex ?
@cindex cautionary accidental
@cindex parenthesized accidental
@funindex !

Normally accidentals are printed automatically, but you may also
print them manually.  A reminder accidental can be forced by
adding an exclamation mark @code{!} after the pitch.  A cautionary
accidental (i.e., an accidental within parentheses) can be
obtained by adding the question mark @code{?} after the pitch.
These extra accidentals can be used to produce natural signs, too.

@lilypond[quote,ragged-right,fragment,verbatim,relative=1]
cis cis cis! cis? c c? c! c
@end lilypond


@cindex quarter tones
@cindex semi-flats, semi-sharps

Half-flats and half-sharps are formed by adding @code{eh} and
@code{ih}; the following is a series of Cs with increasing
pitches

@lilypond[verbatim,ragged-right,quote,relative=2,fragment]
\set Staff.extraNatural = ##f
ceseh ceh cih cisih
@end lilypond

Micro tones are also exported to the MIDI file.


@commonprop

In accordance with standard typesetting rules, a natural sign is
printed before a sharp or flat if a previous accidental needs to
be cancelled.  To change this behavior, use

@lilypond[fragment,quote,ragged-right,verbatim,relative=2]
ceses4 ces cis c
\set Staff.extraNatural = ##f
ceses4 ces cis c
@end lilypond


@seealso

Music glossary: @rglos{sharp}, @rglos{flat}, @rglos{double sharp},
@rglos{double flat}.

Notation reference: @ref{Automatic accidentals}, @ref{Musica ficta
accidentals}.

Snippets: @lsrdir{pitch}


@refbugs

There are no generally accepted standards for denoting
three-quarter flats, so LilyPond's symbol does not conform to any
standard.



@node Note names in other languages
@unnumberedsubsubsec Note names in other languages

There are predefined sets of note names for various other
languages.  To use them, include the language specific init file.
For example, add @code{\include "english.ly"} to the top of the
input file.  The available language files and the note names they
define are:

@cindex note names, other languages

@example
@multitable @columnfractions .2 .6 .05 .05 .05 .05
@headitem Language
  @tab Note names
  @tab sharp @tab flat @tab double sharp @tab double flat
@item netherlands.ly
  @tab c d e f g a bes b
  @tab -is @tab -es @tab -isis @tab -eses
@item english.ly
  @tab c d e f g a bf b 
  @tab -s/-sharp @tab -f/-flat @tab -ss/-x/-sharpsharp
    @tab -ff/-flatflat
@item deutsch.ly
  @tab c d e f g a b h 
  @tab -is @tab -es @tab -isis @tab -eses
@item norsk.ly
  @tab c d e f g a b h 
  @tab -iss/-is @tab -ess/-es @tab -ississ/-isis
    @tab -essess/-eses
@item svenska.ly
  @tab c d e f g a b h 
  @tab -iss @tab -ess @tab -ississ @tab -essess
@item italiano.ly
  @tab do re mi fa sol la sib si 
  @tab -d @tab -b @tab -dd @tab -bb
@item catalan.ly
  @tab do re mi fa sol la sib si 
  @tab -d/-s @tab -b @tab -dd/-ss @tab -bb
@item espanol.ly
  @tab do re mi fa sol la sib si 
  @tab -s @tab -b @tab -ss @tab -bb
@end multitable
@end example

@noindent
Note that in Dutch, German, Norwegian, and Swedish, the flat
alterations of @code{a} like for example @code{aes} and
@code{aeses} are usually contracted to @code{as} and @code{ases}
(or more commonly @code{asas}).  Sometimes only these contracted
names are defined in the corresponding language files (this also
applies to the suffixes for quartertones below).

@noindent
Some music uses microtones whose alterations are fractions of a
@q{normal} sharp or flat.  The note names for quartertones defined
in the various language files are listed in the following table.
Here the prefixes @q{semi-} and @q{sesqui-} mean @q{half} and
@q{one and a half}, respectively.  For Norwegian, Swedish, Catalan
and Spanish no special names have been defined yet.


@c What about Turkish Maquam music and similar microtonal systems?
@c
@c Note that the term "three-quarter-sharp/-flat" used in lilypond's source code 
@c is actually misleading since the alteration is in fact one and a half
@c of a regular sharp/flat.  Whence the naming "sesqui-sharp/-flat" used below.

@example
@multitable @columnfractions .2 .6 .05 .05 .05 .05
@headitem Language
  @tab Note names
  @tab semi- @tab semi- @tab sesqui- @tab sesqui-

@item nederlands.ly
  @tab c d e f g a bes b 
  @tab -ih @tab -eh @tab -isih @tab -eseh
@item english.ly
  @tab c d e f g a bf b 
  @tab -qs @tab -qf @tab -tqs @tab -tqf
@item deutsch.ly
  @tab c d e f g a b h 
  @tab -ih @tab -eh @tab -isih @tab -eseh
@c @item norsk.ly
@c   @tab c d e f g a b h 
@c @item svenska.ly
@c   @tab c d e f g a b h 
@item italiano.ly
  @tab do re mi fa sol la sib si 
  @tab -sd @tab -sb @tab -dsd @tab -bsb
@c @item catalan.ly 
@c   @tab do re mi fa sol la sib si 
@c @item  espanol.ly 
@c   @tab do re mi fa sol la sib si 
@end multitable
@end example

@seealso

Snippets: @lsrdir{pitch}



@node Changing multiple pitches
@subsection Changing multiple pitches

This section discusses how to modify pitches.

@menu
* Octave check::                
* Transpose::                   
@end menu

@node Octave check
@unnumberedsubsubsec Octave check

@cindex Octave check

In relative mode, it is easy to forget an octave changing mark.
Octave checks make such errors easier to correct. A note may be
followed by @code{=}@var{quotes} to indicate what its absolute
octave should be.  In the following example,

@example
\relative c'' @{ c g d='' @}
@end example

@noindent
the @code{d} will generate a warning, because a
@code{d''} is expected, as indicated by the octave check
@code{=''}, but a @code{d'} is found, since both @code{c}-@code{g}
and @code{g}-@code{d} without octavation marks are descending
fourths.  In the output, the octave is corrected to be a
@code{d''} and the next note is calculated relative to @code{d''}
instead of @code{d'}.

There is also an octave check that produces no visible output.
The syntax is:

@example
@var{pitch} \octave @var{controlpitch}
@end example

where @var{pitch} is a regular note within a music expression in
relative mode and @var{controlpitch} is a pitch in absolute mode.
This checks that @var{pitch} lies within the octave surrounding
@var{controlpitch}, i.e. not more than a fourth above or below it.
If this is not the case, a warning is printed and the octave is
corrected.  The @var{controlpitch} is not printed as a note.

In the example below, the first check passes without incident,
since the @code{e} (in @code{relative} mode) is within the octave
range of @code{a'}.  However, the second check produces a warning,
since the @code{e} is not within the octave range of @code{b'}.
The warning message is printed, and the octave is adjusted so that
the following notes are in the correct octave once again.

@example
\relative c' @{
  e
  \octave a'
  \octave b'
@}
@end example


The octave of a note following an @code{\octave} check is
determined with respect to the note preceding it, as modified by
the outcome of the octave check.  In the next fragment, the first
note, an @code{e'}, lies within the octave of @code{b}, and the
@code{\octave} check passes successfully. The second note is then
calculated relative to (absolute) @code{e'}, which yields an 
@code{a'}, above middle C.  In this case, the check had no
influence on the output of the piece.

The second @code{\octave} check fails: @code{a'} is not within the
range of @code{b}. A warning is issued, and the last note is
calculated relative to @code{a}, not to @code{a'}, and we get a
@code{d'}.  Without the last @code{\octave} check, the last note
would have been a @code{d''}

@lilypond[quote,ragged-right,verbatim,fragment]
\relative c' {
  e
  \octave b
  a
  \octave b
  d
}
@end lilypond

@seealso

Snippets: @lsrdir{pitch}


@node Transpose
@unnumberedsubsubsec Transpose

@cindex Transpose
@cindex Transposition of pitches
@funindex \transpose

A music expression can be transposed with @code{\transpose}.  The
syntax is

@example
\transpose @var{from} @var{to} @var{musicexpr}
@end example

This means that @var{musicexpr} is transposed by the interval
between the pitches @var{from} and @var{to}: any note with pitch
@var{from} is changed to @var{to} and any other note is transposed
by the same interval.

Consider a piece written in the key of D-major.  If this piece is
a little too low for its performer, it can be transposed up to
E-major with

@example
\transpose d e @dots{}
@end example

Consider a part written for violin (a C instrument).  If this part
is to be played on the A clarinet (for which an A is notated as a
C, and thus sounds a minor third lower than notated), the
following transposition will produce the appropriate part:

@example
\transpose a c @dots{}
@end example

@code{\transpose} will also transpose key signatures:

@lilypond[quote,ragged-right,verbatim]
mus = { \key d \major cis d fis g }
\new Staff {
  \clef "F" \mus
  \clef "G"
  \transpose c g' \mus
  \transpose c f' \mus
}
@end lilypond

@code{\transpose} distinguishes between enharmonic pitches: both
@code{\transpose c cis} or @code{\transpose c des} will transpose
up half a tone.  The first version will print sharps and the notes
will remain on the same scale step, the second version will print
flats on the scale step above.

@lilypond[quote,ragged-right,verbatim]
mus = { c d e f }
\new Staff {
  \transpose c cis { \relative c' \mus }
  \transpose c des { \relative c' \mus }
}
@end lilypond

@code{\transpose} may also be used to input written notes for a
transposing instrument.  The previous examples show how to enter
pitches in C (or @notation{concert pitch}) and typeset them for a
transposing instrument, but the opposite is also possible if you
for example have a set of instrumental parts and want to print a
conductor's score. For example, when entering music for a B-flat
trumpet which begins on a notated E (concert D), one would write:

@example
musicInBflat = @{ e4 @dots{} @}
\transpose c bes \musicInBflat
@end example

To print this music in F (e.g., rearranging to a french horn) you
would wrap the existing music with another @code{\transpose}

@example
musicInBflat = @{ e4 @dots{} @}
\transpose f c @{ \transpose c bes \musicInBflat @}
@end example

@seealso

Snippets: @lsrdir{pitch}, @lsr{scheme,
transpose-pitches-with-minimum-accidentals.ly}.

Program reference: @internalsref{TransposedMusic}.

@refbugs

If you want to use both @code{\transpose} and @code{\relative},
you must put @code{\transpose} outside of @code{\relative}, since
@code{\relative} will have no effect on music that appears inside
a @code{\transpose}. See the previous example. 


@node Displaying pitches
@subsection Displaying pitches

This section discusses how to alter the output of pitches.

@menu
* Clef::                        
* Key signature::               
* Ottava brackets::             
* Instrument transpositions::   
* Automatic accidentals::       
@end menu

@node Clef
@unnumberedsubsubsec Clef

@funindex \clef

@cindex G clef
@cindex C clef
@cindex F clef
@cindex treble clef
@cindex violin clef
@cindex alto clef
@cindex tenor clef
@cindex bass clef
@cindex french clef
@cindex soprano clef
@cindex mezzosoprano clef
@cindex baritone clef
@cindex varbaritone clef
@cindex subbass clef

The clef is set with the @code{\clef} @var{clefname} 
command:

@lilypond[quote,ragged-right,fragment,verbatim]
<<
  \new Staff {
    % treble clef by default
    c'1^"middle C"
  }
  \new Staff {
    \clef alto
    c'1^"middle C"
  }
  \new Staff {
    \clef bass
    c'1^"middle C"
  }
>>
@end lilypond

The clef can also be changed inside the staff:

@lilypond[quote,ragged-right,fragment,verbatim]
\new PianoStaff <<
  \new Staff {
    \time 2/4
    \clef treble
    r16
    \clef bass
    e g c' e' e g c'
  }
  \new Staff {
    \clef bass
    c,8 r \clef treble g''4
  }
>>
@end lilypond

These same clef symbols are used in different positions on
the staff to change the range of notes shown by that staff.
The treble (alto, bass) clef is always positioned to show
the line on which a g' (c', f) note is printed.
 
Supported clefs include

@multitable @columnfractions .4 .6
@headitem Clef @tab Position
@item @code{treble, violin, G, G2} @tab
G clef (g') on 2nd line
@item @code{alto, C} @tab
C clef (c') on 3rd line
@item @code{tenor} @tab
C clef (c') on 4th line.
@item @code{bass, F} @tab
F clef (f) on 4th line
@item @code{french} @tab
G clef (g') on 1st line, so-called French violin clef
@item @code{soprano} @tab
C clef (c') on 1st line
@item @code{mezzosoprano} @tab
C clef (c') on 2nd line
@item @code{baritone} @tab
C clef (c') on 5th line
@item @code{varbaritone} @tab
F clef (f) on 3rd line
@item @code{subbass} @tab
F clef (f) on 5th line
@item @code{percussion} @tab
percussion clef
@item @code{tab} @tab
tablature clef
@end multitable

Further supported clefs are described under @ref{Ancient clefs}.

By adding @code{_8} or @code{^8} to the clef name, the clef is
transposed one octave down or up, respectively, and @code{_15} and
@code{^15} transposes by two octaves.  The argument @var{clefname}
must be enclosed in quotes when it contains underscores or digits.
For example,

@cindex choral tenor clef

@lilypond[quote,ragged-right,verbatim,fragment,relative=1]
\clef "G_8" c4
@end lilypond

@commonprop

When a clef change takes place at a line break the new clef symbol
is printed at both the end of the previous line and the beginning
of the new line by default.  If the warning clef at the end of the
previous line in not required it can be suppressed by setting the
@code{explicitClefVisibility} Staff property to the value
@var{end-of-line-invisible}:  @code{\set
Staff.explicitClefVisibility = #end-of-line-invisible} The default
behaviour can be recovered with @code{\unset
Staff.explicitClefVisibility}

The command @code{\clef "treble_8"} is equivalent to setting
@code{clefGlyph}, @code{clefPosition} (which controls the Y
position of the clef), @code{middleCPosition} and
@code{clefOctavation}.  A clef is printed when either the
@code{clefGlyph}, @code{clefPosition} or @code{clefOctavation}
properties are changed.  Note that changing the glyph, the
position of the clef, or the octavation, does not in itself change
the position of subsequent notes on the staff: the position of
middle C must also be specified to do this.  The positional
parameters are relative to the staff centre line, positive numbers
displacing upwards, counting 1 for each line and space.  The
clefOctavation value would normally be set to 7, -7, 15 or -15,
but other values are not invalid.

The following examples show the possibilities when setting these
properties manually.  First some manual changes which preserve the
standard relative positioning of clefs and notes:

@lilypond[quote,ragged-right,verbatim]
{
  % The default treble clef
  c'4
  % The standard bass clef
  \set Staff.clefGlyph = #"clefs.F"
  \set Staff.clefPosition = #2
  \set Staff.middleCPosition = #6
  c'4
  % The baritone clef
  \set Staff.clefGlyph = #"clefs.C"
  \set Staff.clefPosition = #4
  \set Staff.middleCPosition = #4
  c'4
  % The standard choral tenor clef
  \set Staff.clefGlyph = #"clefs.G"
  \set Staff.clefPosition = #-2
  \set Staff.clefOctavation = #-7
  \set Staff.middleCPosition = #1
  c'4
  % A non-standard clef
  \set Staff.clefPosition = #0
  \set Staff.clefOctavation = #0
  \set Staff.middleCPosition = #-4
  c'4
}
@end lilypond

and some which do not:

@lilypond[quote,ragged-right,verbatim]
{
  \set Staff.clefGlyph = #"clefs.F"
  \set Staff.clefPosition = #2
  c'4
  \set Staff.clefGlyph = #"clefs.G"
  c'4
  \set Staff.clefGlyph = #"clefs.C"
  c'4
  \set Staff.clefOctavation = #7
  c'4
  \set Staff.clefOctavation = #0
  \set Staff.clefPosition = #0
  c'4
  \set Staff.middleCPosition = #4
  c'4
}
@end lilypond

@seealso

Notation reference: @ref{Ancient clefs}.

Snippets: @lsrdir{pitch}

Program reference: @internalsref{Clef}.


@node Key signature
@unnumberedsubsubsec Key signature

@cindex Key signature
@funindex \key

@warning{New users are sometimes confused about accidentals and
key signatures.  In LilyPond, note names are the raw input; key
signatures and clefs determine how this raw input is displayed.
An unaltered note like @code{c} means @q{C natural}, regardless of
the key signature or clef.  For more information, see
@rlearning{Accidentals and key signatures}.}

The key signature indicates the tonality in which a piece is
played.  It is denoted by a set of alterations (flats or sharps)
at the start of the staff.

Setting or changing the key signature is done with the @code{\key}
command:

@example
\key @var{pitch} @var{type}
@end example

@funindex \minor
@funindex \major
@funindex \minor
@funindex \ionian
@funindex \locrian
@funindex \aeolian
@funindex \mixolydian
@funindex \lydian
@funindex \phrygian
@funindex \dorian
@cindex church modes

Here, @var{type} should be @code{\major} or @code{\minor} to get
@var{pitch}-major or @var{pitch}-minor, respectively.  You may
also use the standard mode names (also called @q{church modes}):
@code{\ionian}, @code{\locrian}, @code{\aeolian},
@code{\mixolydian}, @code{\lydian}, @code{\phrygian}, and
@code{\dorian}.

@lilypond[quote,ragged-right,verbatim,relative=2,fragment]
\key g \major
f1
fis
@end lilypond


@commonprop

A natural sign is printed to cancel any previous accidentals.
This may be altered:

@lilypond[quote,fragment,ragged-right,verbatim,relative=2]
\key d \major
a b cis d
\key g \minor
a bes c d
\set Staff.printKeyCancellation = ##f
\key d \major
a b cis d
\key g \minor
a bes c d
@end lilypond

The @code{\key} command sets the context property
@code{Staff.keySignature}.  Non-standard key signatures can be
specified by setting this property directly.  The format of this
command is a list:

@example
\set Staff.keySignature = #'((@var{octave} .  @var{step}) . @var{alter})
@end example

@noindent
where, for each element in the list, @var{octave} specifies the
octave (0 being the octave from middle C to the B above),
@var{step} specifies the note within the octave (0 means C and 6
means B), and @var{alter} is ,SHARP ,FLAT ,DOUBLE-SHARP etc.
(Note the leading comma.)  Here is an example of a possible key
signature for generating a whole-tone scale:

@lilypond[quote,ragged-right,verbatim]
\relative c'
  \new Staff {
    \set Staff.keySignature = 
      #`(((0 .  3) . ,SHARP) ((0 . 5) . ,FLAT) ((0 . 6) . ,FLAT))
    c d e fis aes bes c
}
@end lilypond

@seealso

Music glossary: @rglos{Church mode}, @rglos{scordatura}.

Snippets: @lsrdir{pitch},
FIXME link to snippet 248,
FIXME link to a snippet about scordatura.

Program reference: @internalsref{KeyCancellation},
@internalsref{KeySignature}, @internalsref{Key_engraver}


@node Ottava brackets
@unnumberedsubsubsec Ottava brackets

@cindex ottava
@cindex 15ma
@cindex 8va
@cindex 8ve
@cindex octavation

@notation{Ottava brackets} introduce an extra transposition of an
octave for the staff:

@lilypond[quote,ragged-right,verbatim,fragment]
\relative c''' {
  a2 b
  #(set-octavation 1)
  a b
  #(set-octavation 0)
  a b
}
@end lilypond

The @code{set-octavation} function also takes -1 (for 8va bassa),
2@tie{}(for 15ma), and -2 (for 15ma bassa) as arguments.
Internally the function sets the properties @code{ottavation}
(e.g., to @code{"8va"} or @code{"8vb"}) and
@code{middleCPosition}.  To override the text of the bracket, set
@code{ottavation} after invoking @code{set-octavation}, i.e.,

@lilypond[quote,ragged-right,verbatim]
{
  #(set-octavation 1)
  \set Staff.ottavation = #"8"
  c'''
}
@end lilypond


@seealso

Snippets: @lsrdir{pitch}

Program reference: @internalsref{OttavaBracket}.


@refbugs

@code{set-octavation} will get confused when clef changes happen
during an octavation bracket.


@node Instrument transpositions
@unnumberedsubsubsec Instrument transpositions

@cindex transposition, MIDI
@cindex transposition, instrument

The key of a @notation{transposing instrument} can be specified.
This applies to many wind instruments, for example, clarinets
(B-flat, A, and E-flat), horn (F) and trumpet (B-flat, C, D, and
E-flat).

The transposition is entered with the command

@example
\transposition @var{pitch}
@end example

@noindent
where @var{pitch} specifies the key in which the instrument plays.
For example, a part for B-flat clarinet would include

@example
\transposition bes
@end example

This command sets the property @code{instrumentTransposition}.
The value of this property is used for MIDI output and quotations.
MIDI output is transposed to play at concert pitch, and cue notes
and quotations are automatically transposed to the key of the
instrument in which they are quoted.  For more information about
quotations, see @ref{Quoting other voices}.

The pitch to use for @code{\transposition} should correspond to
the real sound heard when a @code{c'} written on the staff is
played by the transposing instrument.  @code{\transposition}
should be used when the music is entered from an orchestral part
written for the instrument.  For example, in classical horn parts,
the tuning of the instrument is often changed during a piece.
Here are a few notes for a violin and a B-flat clarinet where the
parts have been entered using the notes and key as they appear in
each part or the conductor's score:

@lilypond[quote,ragged-right,verbatim]
\new GrandStaff <<
    \new Staff = "Violin I" {
      \set Staff.instrumentName = "Vln"
      \set Staff.midiInstrument="violin"
      \key c \major
      \transposition c
      g'4( c''8) r c'' r c''4
    }
    \new Staff = "Clarinet" {
      \set Staff.instrumentName = "Cl"
      \set Staff.midiInstrument="clarinet"
      \key d \major
      \transposition bes
      a'4( d''8) r d'' r d''4
    }
>>
@end lilypond

Without the @code{\transposition} command the midi output of this
music would play the notes a tone apart; with it the notes are
played correctly in unison.

However, when entering the notes of a score in concert pitch, when
all voices are entered in C, they may be prefaced by (the default)
@code{\transposition c'}:

@example
clarinet = @{
  \transposition c'
  ...
@}
saxophone = @{
  \transposition c'
  ...
@}
@end example

@noindent
to serve as a reminder that these parts are written in C.


@seealso

Notation reference: @ref{Quoting other voices}, @ref{Transpose}.

Snippets: @lsrdir{pitch}


@node Automatic accidentals
@unnumberedsubsubsec Automatic accidentals
@cindex Automatic accidentals

Common rules for typesetting accidentals have been placed in a
function.  This function is called as follows

@funindex set-accidental-style
@example
#(set-accidental-style 'STYLE)
@end example

@c TODO: check the context stuff below
@c -does it *really* work?
@c -the default contexts as specified in
@c  scm/music-function.scm seem to be different -vv

Optionally, the function can take two arguments: the name of the
accidental style, and an optional argument that denotes the context that
should be changed:

@example
#(set-accidental-style 'STYLE #('CONTEXT#))
@end example

If no context name is supplied, @code{Staff} is the default,
but you may wish to apply the accidental style to a single @code{Voice}
instead.

The following accidental styles are supported:

@table @code
@item default
This is the default typesetting behavior.  It corresponds
to 18th century common practice: Accidentals are
remembered to the end of the measure in which they occur and
only on their own octave.

@lilypond[quote,ragged-right]
musicA = {  << \relative {  cis'8 fis, d'4 <a cis>8 f bis4 | cis2. <c, g'>4 | } \\
       \relative { ais'2 cis, | fis8 b a4 cis2 | } >> }

musicB = { \clef bass  \new Voice { \voiceTwo \relative { < fis, a cis>4 
	\change Staff = up cis' \change Staff = down <fis, a>  
	\change Staff = up dis' | \change Staff = down <fis, a cis>4 gis 
	<f a d>2 |  } }}

\score { 
	\new PianoStaff {
       << \context Staff = "up" {
	       %#(set-accidental-style 'default) 
       \musicA }
       \context Staff = "down"{
	       %#(set-accidental-style 'default)
       \musicB } >> }
       \header { piece = \markup {\fill-line { \fontsize #3  "'default" }}}
}
@end lilypond

@item voice
The normal behavior is to remember the accidentals on
Staff-level.  This variable, however, typesets accidentals
individually for each voice.  Apart from that, the rule is similar to
@code{default}.

@example
 \new Staff <<
        #(set-accidental-style 'voice)
	@{ @dots{} @}
	>>
@end example

As a result, accidentals from one voice do not get canceled in other
voices, which is often an unwanted result: in the following example, it
is hard to determine whether the second @samp{a} should be played
natural or sharp.

@lilypond[quote,ragged-right]
musicA = {  << \relative {  cis'8 fis, d'4 <a cis>8 f bis4 | cis2. <c, g'>4 | } \\
       \relative { ais'2 cis, | fis8 b a4 cis2 | } >> }

musicB = { \clef bass  \new Voice { \voiceTwo \relative { < fis, a cis>4 
	\change Staff = up cis' \change Staff = down <fis, a>  
	\change Staff = up dis' | \change Staff = down <fis, a cis>4 gis 
	<f a d>2 |  } }}

\score { 
	\new PianoStaff {
       << \context Staff = "up" {
	       #(set-accidental-style 'voice) 
       \musicA }
       \context Staff = "down"{
	       #(set-accidental-style 'voice)
       \musicB } >> }
       \header { piece = \markup {\fill-line { \fontsize #3  "'voice" }}}
}
@end lilypond

The @code{voice} option should be used if the voices
are to be read solely by individual musicians.  If the staff is to be
used by one musician (e.g., a conductor) then
@code{modern} or @code{modern-cautionary}
should be used instead.

@item modern
@funindex modern style accidentals
This rule corresponds to the common practice in the 20th century.  This rule
prints the same accidentals as @code{default}, but temporary
accidentals also are canceled in other octaves.  Furthermore,
in the same octave, they also get canceled in the following
measure: in the following example, notice the two natural signs which appear
in the second bar of the upper staff.

@lilypond[quote,ragged-right]
musicA = {  << \relative {  cis'8 fis, d'4 <a cis>8 f bis4 | cis2. <c, g'>4 | } \\
       \relative { ais'2 cis, | fis8 b a4 cis2 | } >> }

musicB = { \clef bass  \new Voice { \voiceTwo \relative { < fis, a cis>4 
	\change Staff = up cis' \change Staff = down <fis, a>  
	\change Staff = up dis' | \change Staff = down <fis, a cis>4 gis 
	<f a d>2 |  } }}

\score { 
	\new PianoStaff {
       << \context Staff = "up" {
	       #(set-accidental-style 'modern) 
       \musicA }
       \context Staff = "down"{
	       #(set-accidental-style 'modern)
       \musicB } >> }
       \header { piece = \markup {\fill-line { \fontsize #3  "'modern" }}}
}
@end lilypond

@item @code{modern-cautionary}
@funindex modern-cautionary
This rule is similar to @code{modern}, but the @q{extra} accidentals
(the ones not typeset by @code{default}) are typeset as cautionary
accidentals.  They are printed in reduced size or (by default)
with parentheses -- this can be set by definig the @code{cautionary-style}
property of the @internalsref{AccidentalSuggestion} object.

@lilypond[quote,ragged-right]
musicA = {  << \relative {  cis'8 fis, d'4 <a cis>8 f bis4 | cis2. <c, g'>4 | } \\
       \relative { ais'2 cis, | fis8 b a4 cis2 | } >> }

musicB = { \clef bass  \new Voice { \voiceTwo \relative { < fis, a cis>4 
	\change Staff = up cis' \change Staff = down <fis, a>  
	\change Staff = up dis' | \change Staff = down <fis, a cis>4 gis 
	<f a d>2 |  } }}

\score { 
	\new PianoStaff {
       << \context Staff = "up" {
	       #(set-accidental-style 'modern-cautionary) 
       \musicA }
       \context Staff = "down"{
	       #(set-accidental-style 'modern-cautionary)
       \musicB } >> }
       \header { piece = \markup {\fill-line { \fontsize #3  "'modern-cautionary" }}}
}
@end lilypond

@funindex modern-voice
@item modern-voice
This rule is used for multivoice accidentals to be read both by musicians
playing one voice and musicians playing all voices.  Accidentals are
typeset for each voice, but they @emph{are} canceled across voices in
the same @internalsref{Staff}.

@lilypond[quote,ragged-right]
musicA = {  << \relative {  cis'8 fis, d'4 <a cis>8 f bis4 | cis2. <c, g'>4 | } \\
       \relative { ais'2 cis, | fis8 b a4 cis2 | } >> }

musicB = { \clef bass  \new Voice { \voiceTwo \relative { < fis, a cis>4 
	\change Staff = up cis' \change Staff = down <fis, a>  
	\change Staff = up dis' | \change Staff = down <fis, a cis>4 gis 
	<f a d>2 |  } }}

\score { 
	\new PianoStaff {
       << \context Staff = "up" {
	       #(set-accidental-style 'modern-voice) 
       \musicA }
       \context Staff = "down"{
	       #(set-accidental-style 'modern-voice)
       \musicB } >> }
       \header { piece = \markup {\fill-line { \fontsize #3  "'modern-voice" }}}
}
@end lilypond

@funindex modern-voice-cautionary
@item modern-voice-cautionary
This rule is the same as @code{modern-voice}, but with the extra
accidentals (the ones not typeset by @code{voice}) typeset
as cautionaries.  Even though all accidentals typeset by
@code{default} @emph{are} typeset by this variable,
some of them are typeset as cautionaries.

@lilypond[quote,ragged-right]
musicA = {  << \relative {  cis'8 fis, d'4 <a cis>8 f bis4 | cis2. <c, g'>4 | } \\
       \relative { ais'2 cis, | fis8 b a4 cis2 | } >> }

musicB = { \clef bass  \new Voice { \voiceTwo \relative { < fis, a cis>4 
	\change Staff = up cis' \change Staff = down <fis, a>  
	\change Staff = up dis' | \change Staff = down <fis, a cis>4 gis 
	<f a d>2 |  } }}

\score { 
	\new PianoStaff {
       << \context Staff = "up" {
	       #(set-accidental-style 'modern-voice-cautionary) 
       \musicA }
       \context Staff = "down"{
	       #(set-accidental-style 'modern-voice-cautionary)
       \musicB } >> }
       \header { piece = \markup {\fill-line { \fontsize #3  "'modern-voice-cautionary" }}}
}
@end lilypond

@item piano
@funindex piano accidentals
This accidental style takes place in a GrandStaff context.  However, you have to 
explicitly set it for @emph{each} individual Staff of the GrandStaff:

@example
\new GrandStaff @{ <<
  \new Staff = "up" @{ <<
    #(set-accidental-style 'piano)
    @{ @dots{} @}
  >> @}
  \new Staff = "down"@{ <<
    #(set-accidental-style 'piano)
  @{ @dots{} @}
  >> @}
>> @}
@end example

This rule reflects 20th century practice for piano notation.  Its behavior is very
similar to @code{modern} style, but here accidentals also get canceled
across the staves in the same @internalsref{GrandStaff} or
@internalsref{PianoStaff}.

@lilypond[quote,ragged-right]
musicA = {  << \relative {  cis'8 fis, d'4 <a cis>8 f bis4 | cis2. <c, g'>4 | } \\
       \relative { ais'2 cis, | fis8 b a4 cis2 | } >> }

musicB = { \clef bass  \new Voice { \voiceTwo \relative { < fis, a cis>4 
	\change Staff = up cis' \change Staff = down <fis, a>  
	\change Staff = up dis' | \change Staff = down <fis, a cis>4 gis 
	<f a d>2 |  } }}

\score { 
	\new PianoStaff {
       << \context Staff = "up" {
	       #(set-accidental-style 'piano) 
       \musicA }
       \context Staff = "down"{
	       #(set-accidental-style 'piano)
       \musicB } >> }
       \header { piece = \markup {\fill-line { \fontsize #3  "'piano" }}}
}
@end lilypond

@item piano-cautionary
@funindex #(set-accidental-style 'piano-cautionary)
Same as @code{#(set-accidental-style 'piano)} but with the extra
accidentals typeset as cautionaries.

@lilypond[quote,ragged-right]
musicA = {  << \relative {  cis'8 fis, d'4 <a cis>8 f bis4 | cis2. <c, g'>4 | } \\
       \relative { ais'2 cis, | fis8 b a4 cis2 | } >> }

musicB = { \clef bass  \new Voice { \voiceTwo \relative { < fis, a cis>4 
	\change Staff = up cis' \change Staff = down <fis, a>  
	\change Staff = up dis' | \change Staff = down <fis, a cis>4 gis 
	<f a d>2 |  } }}

\score { 
	\new PianoStaff {
       << \context Staff = "up" {
	       #(set-accidental-style 'piano-cautionary) 
       \musicA }
       \context Staff = "down"{
	       #(set-accidental-style 'piano-cautionary)
       \musicB } >> }
       \header { piece = \markup {\fill-line { \fontsize #3  "'piano-cautionary" }}}
}
@end lilypond

@item no-reset
@funindex no-reset accidental style
This is the same as @code{default} but with accidentals lasting
@q{forever} and not only until the next measure:
@lilypond[quote,ragged-right]
musicA = {  << \relative {  cis'8 fis, d'4 <a cis>8 f bis4 | cis2. <c, g'>4 | } \\
       \relative { ais'2 cis, | fis8 b a4 cis2 | } >> }

musicB = { \clef bass  \new Voice { \voiceTwo \relative { < fis, a cis>4 
	\change Staff = up cis' \change Staff = down <fis, a>  
	\change Staff = up dis' | \change Staff = down <fis, a cis>4 gis 
	<f a d>2 |  } }}

\score { 
	\new PianoStaff {
       << \context Staff = "up" {
	       #(set-accidental-style 'no-reset) 
       \musicA }
       \context Staff = "down"{
	       #(set-accidental-style 'no-reset)
       \musicB } >> }
       \header { piece = \markup {\fill-line { \fontsize #3  "'no-reset" }}}
}
@end lilypond

@item forget
This is sort of the opposite of @code{no-reset}: Accidentals
are not remembered at all -- and hence all accidentals are
typeset relative to the key signature, regardless of what was
before in the music

@lilypond[quote,ragged-right]
musicA = {  << \relative {  cis'8 fis, d'4 <a cis>8 f bis4 | cis2. <c, g'>4 | } \\
       \relative { ais'2 cis, | fis8 b a4 cis2 | } >> }

musicB = { \clef bass  \new Voice { \voiceTwo \relative { < fis, a cis>4 
	\change Staff = up cis' \change Staff = down <fis, a>  
	\change Staff = up dis' | \change Staff = down <fis, a cis>4 gis 
	<f a d>2 |  } }}

\score { 
	\new PianoStaff {
       << \context Staff = "up" {
	       #(set-accidental-style 'forget) 
       \musicA }
       \context Staff = "down"{
	       #(set-accidental-style 'forget)
       \musicB } >> }
       \header { piece = \markup {\fill-line { \fontsize #3  "'forget" }}}
}
@end lilypond
@end table


@seealso

Program reference: @internalsref{Accidental_engraver},
@internalsref{Accidental}, @internalsref{AccidentalSuggestion} and @internalsref{AccidentalPlacement}.


@refbugs

Simultaneous notes are considered to be entered in sequential
mode.  This means that in a chord the accidentals are typeset as if the
notes in the chord happen one at a time, in the order in which
they appear in the input file.  This is a problem when accidentals
in a chord depend on each other,
which does not happen for the default accidental style.  The problem
can be solved by manually inserting @code{!} and @code{?} for the
problematic notes.