* GNUmakefile.in (EXTRA_DIST_FILES): remove VIM stuff.

[lilypond.git] / Documentation / user / running.itely

@node Using LilyPond
@chapter Using LilyPond

@menu
* Setting variables::           
* Fine tuning layout::          
* Tuning output::               
* Text markup::                 
* Global layout::               
* Sound::                       
@end menu

@node Setting variables
@section Setting variables

When the music is converted from notes to print it is interpreted
in left-to-right order. This is similar to what happens when we read
music. During this step context-sensitive information such as the
accidentals to print, and where bar lines must be placed, are stored in
variables. These variables are called @emph{context properties}.
The properties can also be manipulated from input files. Consider this input:
@example
\set Staff.autoBeaming = ##f
@end example 

@noindent
It sets the property named @code{autoBeaming} in the current staff at
this point in the music to @code{##f}, which means `false'. This
property controls whether beams are printed automatically:
@c
@lilypond[relative=1,fragment,verbatim]
  c8 c c c
  \set Staff.autoBeaming = ##f
  c8 c c c  
@end lilypond

@noindent
LilyPond includes a built-in programming language, namely, a dialect
of Scheme.  The argument to @code{\set}, @code{##f}, is an
expression in that language.  The first hash-mark signals that a piece
of Scheme code follows. The second hash character is part of the
boolean value true (@code{#t}).  Values of other types may be
entered as follows:
@itemize @bullet
@item a string, enclosed in double quotes, for example,
@example
  \set Staff.instrument = #"French Horn"
@end example
@item a boolean: either @code{#t} or @code{#f}, for true and false
respectively, e.g.
@example
  \set autoBeaming = ##f
  \set Score.skipBars = ##t
@end example

@item a number, such as
@example
  \set Score.currentBarNumber = #20
@end example

@item a symbol, which is introduced by a quote character, as in 
@example
  \set Staff.crescendoSpanner = #'dashed-line
@end example

@item a pair, which is also introduced by a quote character, like in
the following statements, which set properties to the pairs (-7.5, 6) 
and (3, 4) respectively:

@example
  \set Staff.minimumVerticalExtent = #'(-7.5 . 6)
  \set Staff.timeSignatureFraction = #'(3 . 4)
@end example

@item a list, which is also introduced by a quote character. In the
following example, the @code{breakAlignOrder} property is set to a
list of symbols:
@example
  \set Score.breakAlignOrder = 
 #'(left-edge time-signature key-signatures)
@end example


@end itemize

There are many different properties.  Not all of them are listed in
this manual. However, the program reference lists them all in the
section @internalsref{Context-properties}, and most properties are
demonstrated in one of the
@ifhtml
@uref{../../../input/test/out-www/collated-files.html,tips-and-tricks}
@end ifhtml
@ifnothtml
tips-and-tricks
@end ifnothtml
examples.


@node Fine tuning layout
@section Fine tuning layout

Sometimes it is necessary to change music layout by hand.  When music
is formatted, layout objects are created for each symbol.  For
example, every clef and every note head is represented by a layout
object.  These layout objects also carry variables, which we call
@emph{layout properties}. By changing these variables from their
values, we can alter the look of a formatted score:

@lilypond[verbatim,relative]
  c4
  \override Stem #'thickness = #3.0
  c4 c4 c4 
@end lilypond

@noindent
In the example shown here, the layout property @code{thickness} (a
symbol) is set to 3 in the @code{Stem} layout objects of the current
As a result, the notes following @code{\override} have thicker
stems.

For the most part, a manual override is needed only on a case by
case basis and not for all subsequent instances of the altered
property. To accomplish this, simply prefix @code{\once} to the
@code{\override} statement and the override will apply only once,
immediately reverting to its default setting, i.e.

@example
 \once \override Stem #'thickness = #3.0
@end example

@lilypond[relative]
  c4
  \once \override Stem #'thickness = #3.0
  c4 c4 c4 
@end lilypond

@noindent
Some overrides are so common that predefined commands are provided as
a short cut.  For example, @code{\slurUp} and @code{\stemDown}. These
commands are described in
@ifhtml
the
@end ifhtml
@ref{Notation manual}, under the sections for slurs and stems
respectively.

The exact tuning possibilities for each type of layout object are
documented in the program reference of the respective
object. However, many layout objects share properties, which can be
used to apply generic tweaks.  We mention a couple of these:

@itemize @bullet
@item The @code{extra-offset} property, which
@cindex @code{extra-offset}
has a pair of numbers as value, moves around objects in the printout.
The first number controls left-right movement; a positive number will
move the object to the right.  The second number controls up-down
movement; a positive number will move it higher.  The units of these
offsets are staff-spaces.  The @code{extra-offset} property is a
low-level feature: the formatting engine is completely oblivious to
these offsets.

In the following example, the second fingering is moved a little to
the left, and 1.8 staff space downwards:

@cindex setting object properties

@lilypond[relative=1,verbatim]
\stemUp
f-5
\once \override Fingering
    #'extra-offset = #'(-0.3 . -1.8) 
f-5
@end lilypond

@item
Setting the @code{transparent} property will cause an object to be printed
in `invisible ink': the object is not printed, but all its other
behavior is retained. The object still takes up space, it takes part in
collisions, and slurs, and ties and beams can be attached to it.

@cindex transparent objects
@cindex removing objects
@cindex invisible objects
The following example demonstrates how to connect different voices
using ties. Normally, ties only connect two notes in the same
voice. By introducing a tie in a different voice, and blanking a stem
in that voice, the tie appears to cross voices:

@lilypond[fragment,relative=1,verbatim]
  c4 << {
      \once \override Stem #'transparent = ##t
      b8~ b8
  } \\ {
       b[ g8]
  } >>
@end lilypond

@item
The @code{padding} property for objects with
@cindex @code{padding}
@code{side-position-interface} can be set to increase distance between
symbols that are printed above or below notes. We only give an
example; a more elaborate explanation is in @ref{Constructing a
tweak}:

@lilypond[relative=1,verbatim]
  c2\fermata
  \override Script #'padding = #3
  b2\fermata
@end lilypond

@end itemize

More specific overrides are also possible.  The notation manual
discusses in depth how to figure out these statements for yourself, in
@ref{Tuning output}.




@node Tuning output
@section Tuning output

There are situations where default layout decisions are not
sufficient.  In this section we discuss ways to override these
defaults.

Formatting is internally done by manipulating so called objects
(graphic objects). Each object carries with it a set of properties
(object or layout properties) specific to that object.  For example, a
stem object has properties that specify its direction, length and
thickness.

The most direct way of tuning the output is by altering the values of
these properties. There are two ways of doing that: first, you can
temporarily change the definition of one type of object, thus
affecting a whole set of objects.  Second, you can select one specific
object, and set a layout property in that object.

Do not confuse layout properties with translation
properties. Translation properties always use a mixed caps style
naming, and are manipulated using @code{\set} and @code{\unset}: 
@example
  \set Context.propertyName = @var{value}
@end example

Layout properties are use Scheme style variable naming, i.e.  lower
case words separated with dashes. They are symbols, and should always
be quoted using @code{#'}.  For example, this could be an imaginary
layout property name:
@example
  #'layout-property-name
@end example

@seealso

The introduction of the @ref{Technical manual} gives a more in-depth
treatment of the difference between translation and layout.

@menu
* Tuning objects::              
* Constructing a tweak::        
* Applyoutput::                 
* Font selection::              
@end menu



@node Tuning objects
@subsection Tuning objects 

@cindex object description

The definition of an object is a list of default object
properties. For example, the definition of the Stem object (available
in @file{scm/define-grobs.scm}), includes the following definitions
for @internalsref{Stem}:

@example
        (thickness . 1.3)
        (beamed-lengths . (3.5 3.5 3.5 4.5 5.0))
        (Y-extent-callback . ,Stem::height)
        @var{...}
@end example


Adding variables on top of this existing definition overrides the
system default, and alters the resulting appearance of the layout
object.

@syntax


Changing a variable for only one object is commonly achieved with
@code{\once}:

@example
\once \override @var{context}.@var{objectname}
    @var{symbol} = @var{value}
@end example
Here @var{symbol} is a Scheme expression of symbol type, @var{context}
and @var{objectname} is a string and @var{value} is a Scheme expression.
This command applies a setting only during one moment in the score.

In the following example, only one @internalsref{Stem} object is
changed from its original setting:

@lilypond[verbatim,fragment,relative=1]
  c4 
  \once \override Voice.Stem #'thickness = #4
  c4
  c4
@end lilypond
@cindex @code{\once}

For changing more objects, the same command, without @code{\once} can
be used:
@example
\override @var{context}.@var{objectname}   @var{symbol} = @var{value}
@end example
This command adds @code{@var{symbol} = @var{value}} to the definition
of @var{objectname} in the context @var{context}, and this definition
stays in place until it is removed.

An existing definition may be removed by the following command:
@c
@example
\property @var{context}.@var{objectname} \revert @var{symbol}
@end example
@c

Some examples: 
@lilypond[verbatim]
c'4 \override Stem   #'thickness = #4.0
c'4
c'4 \revert Stem #'thickness
c'4
@end lilypond

The following example gives exactly the same result as the previous
one (assuming the system default for stem thickness is 1.3):
@c
@lilypond[verbatim]
  c'4 \override Stem   #'thickness = #4.0
  c'4
  c'4 \override Stem   #'thickness = #1.3
  c'4
@end lilypond

Reverting a setting which was not set in the first place has no
effect.


@seealso

Internals: @internalsref{OverrideProperty}, @internalsref{RevertProperty},
@internalsref{PropertySet}, @internalsref{All-backend-properties}, and
@internalsref{All-layout-objects}.


@refbugs

The back-end is not very strict in type-checking object properties.
Cyclic references in Scheme values for properties can cause hangs
and/or crashes.

@menu
* Constructing a tweak::        
* Applyoutput::                 
* Font selection::              
* Text markup::                 
@end menu

@node Constructing a tweak
@subsection Constructing a tweak


@cindex internal documentation
@cindex finding graphical objects
@cindex graphical object descriptions 
@cindex tweaking
@cindex @code{\override}
@cindex @code{\set}
@cindex internal documentation



Three pieces of information are required to use @code{\override} and
@code{\set}: the name of the layout object, the context and the name
of the property.  We demonstrate how to glean this information from
the notation manual and the generated documentation.

The generated documentation is a set of HTML pages which should be
included if you installed a binary distribution, typically in
@file{/usr/share/doc/lilypond}.  They are also available on the web:
go to the @uref{http://lilypond.org,LilyPond website}, click
``Documentation'', select the correct version, and click then
``Program reference.'' It is advisable to bookmark the local HTML
files. They will load faster than the ones on the web.  If you use the
version from the web, you must check whether the documentation matches
the program version: it is generated from the definitions that the
program uses, and therefore it is strongly tied to the LilyPond
version.


@c  [TODO: revise for new site.]

Suppose we want to move the fingering indication in the fragment below:

@lilypond[relative=2,verbatim]
c-2
\stemUp
f
@end lilypond

If you visit the documentation of @code{Fingering} (in @ref{Fingering
instructions}), you will notice that there is written:

@quotation
@seealso

Internals: @internalsref{FingerEvent} and @internalsref{Fingering}.

@end quotation

@separate

@noindent
In other words, the fingerings once entered, are internally stored as
@code{FingerEvent} music objects. When printed, a @code{Fingering}
layout object is created for every @code{FingerEvent}.

The Fingering object has a number of different functions, and each of
those is captured in an interface. The interfaces are listed under
@internalsref{Fingering} in the program reference.



The @code{Fingering} object has a fixed size
(@internalsref{item-interface}), the symbol is a piece of text
(@internalsref{text-interface}), whose font can be set
(@internalsref{font-interface}).  It is centered horizontally
(@internalsref{self-alignment-interface}), it is placed next to other
objects (@internalsref{side-position-interface}) vertically, and its
placement is coordinated with other scripts
(@internalsref{text-script-interface}).  It also has the standard
@internalsref{grob-interface} (grob stands for Graphical object)
@cindex grob
@cindex graphical object
@cindex layout object
@cindex object, layout 
with all the variables that come with
it.  Finally, it denotes a fingering instruction, so it has
@internalsref{finger-interface}.

For the vertical placement, we have to look under
@code{side-position-interface}:
@quotation
@code{side-position-interface}

  Position a victim object (this one) next to other objects (the
  support).  In this case, the property @code{direction} signifies where to put the
  victim object relative to the support (left or right, up or down?)
@end quotation

@cindex padding
@noindent
below this description, the variable @code{padding} is described as
@quotation
@table @code
@item padding
 (dimension, in staff space)

   add this much extra space between objects that are next to each
other. Default value: @code{0.6}
@end table
@end quotation

By increasing the value of @code{padding}, we can move away the
fingering.  The following command inserts 3 staff spaces of white
between the note and the fingering:
@example
\once \override Fingering   #'padding = #3
@end example

Inserting this command before the Fingering object is created,
i.e. before @code{c2}, yields the following result:

@lilypond[relative=2,fragment,verbatim]
\once \override Fingering
    #'padding = #3
c-2
\stemUp
f
@end lilypond

The context name @code{Voice} in the example above can be determined
as follows. In the documentation for @internalsref{Fingering}, it says
@quotation
Fingering grobs are created by: @internalsref{Fingering_engraver} @c
@end quotation

Clicking @code{Fingering_engraver} shows the documentation of
the module responsible for interpreting the fingering instructions and
translating them to a @code{Fingering} object.  Such a module is called
an @emph{engraver}.  The documentation of the @code{Fingering_engraver}
says
@example
Fingering_engraver is part of contexts: Voice 
@end example
so tuning the settings for Fingering should be done with
@example
  \override Fingering @dots{}
@end example

Of course, the tweak may also done in a larger context than
@code{Voice}, for example, @internalsref{Staff} or
@internalsref{Score}.

@seealso

Internals: the program reference also contains alphabetical lists of
@internalsref{Contexts}, @internalsref{All-layout-objects} and
@internalsref{Music-expressions}, so you can also find which objects
to tweak by browsing the internals document.


@node Applyoutput
@subsection Applyoutput

The most versatile way of tuning an object is @code{\applyoutput}. Its
syntax is
@example
\applyoutput @var{proc}
@end example

@noindent
where @var{proc} is a Scheme function, taking three arguments.

When interpreted, the function @var{proc} is called for every layout
object found in the context, with the following arguments:
@itemize @bullet
@item the layout object itself,
@item the context where the layout object was created, and
@item the context where @code{\applyoutput} is processed.
@end itemize


In addition, the cause of the layout object, i.e.  the music
expression or object that was responsible for creating it, is in the
object property @code{cause}.  For example, for a note head, this is a
@internalsref{NoteHead} event, and for a @internalsref{Stem} object,
this is a @internalsref{NoteHead} object.

Here is a simple example of @code{\applyoutput}; it blanks note-heads on the
center-line:
@example
(define (blanker grob grob-origin context)
  (if (and (memq (ly:grob-property grob 'interfaces)
                 note-head-interface)
           (eq? (ly:grob-property grob 'staff-position) 0))

           (ly:grob-set-property! grob 'transparent #t)))
@end example



@node Font selection
@subsection Font selection

The most common thing to change about the appearance of fonts is their
size. The font size of any context can be easily changed by setting
the @code{fontSize} property for that context.  Its value is a number:
negative numbers make the font smaller, positive numbers larger. An
example is given below:
@c
@lilypond[fragment,relative=1,verbatim]
  c4 c4 \set fontSize = #-1
  f4 g4
@end lilypond
This command will set @code{font-size} (see below), and does
not change the size of variable symbols, such as beams or slurs.

One of the uses of @code{fontSize} is to get smaller symbols for cue
notes. An elaborate example of those is in
@inputfileref{input/test,cue-notes.ly}.

@cindex magnification
@cindex cue notes

The font used for printing a object can be selected by setting
@code{font-name}, e.g.
@example
  \override Staff.TimeSignature
      #'font-name = #"cmr17"
@end example

@noindent
Any font can be used, as long as it is available to @TeX{}. Possible
fonts include foreign fonts or fonts that do not belong to the
Computer Modern font family.  The size of fonts selected in this way
can be changed with the @code{font-magnification} property.  For
example, @code{2.0} blows up all letters by a factor 2 in both
directions.

@cindex font size
@cindex font magnification

Font selection for the standard fonts, @TeX{}'s Computer Modern fonts,
can also be adjusted with a more fine-grained mechanism.  By setting
the object properties described below, you can select a different font;
all three mechanisms work for every object that supports
@code{font-interface}:


@table @code
@item font-family
 is a symbol indicating the general class of the typeface.  Supported are
@code{roman} (Computer Modern), @code{braces} (for piano staff
braces), @code{music} (the standard music font, including ancient
glyphs), @code{dynamic} (for dynamic signs) and @code{typewriter}.
  
@item font-shape
  is a symbol indicating the shape of the font, there are typically several
  font shapes available for each font family. Choices are @code{italic},
  @code{caps} and @code{upright}.

@item font-series
is a  symbol indicating the series of the font. There are typically several
font series for each font family and shape. Choices are @code{medium}
and @code{bold}. 

@end table

For any of these properties, the value @code{*} (i.e. the symbol
@code{*}, entered as @code{#'*}), acts as a wildcard. This can be used
to override default setting, which are always present. For example:
@example
  \override Lyrics .LyricText   #'font-series = #'bold
  \override Lyrics .LyricText   #'font-family = #'typewriter
  \override Lyrics .LyricText   #'font-shape = #'*
@end example

@cindex @code{font-style}

The font size is set by modifying the @code{font-size} property.  Its
value is a number indicating the size relative to the standard size.
Each step up is an increase of approximately 12% of the font size. Six
steps is exactly a factor two. The Scheme function @code{magstep}
converts a @code{font-size} number to a scaling factor.

LilyPond has fonts in different design sizes: the music fonts for
smaller sizes are chubbier, while the text fonts are relatively wider.
Font size changes are achieved by scaling the design size that is
closest to the desired size.

The @code{font-size} mechanism does not work for fonts selected
through @code{font-name}. These may be scaled with
@code{font-magnification}.

@refcommands

The following commands set @code{fontSize} for the current voice.

@cindex @code{\tiny}
@code{\tiny}, 
@cindex @code{\small}
@code{\small}, 
@cindex @code{\normalsize}
@code{\normalsize}.

@seealso

Init files: @file{ly/declarations-init.ly} contains hints how new
fonts may be added to LilyPond.

@refbugs

There is no style sheet provided for other fonts besides the @TeX{}
Computer Modern family.

@cindex font selection
@cindex font magnification
@cindex @code{font-interface}


@node Text markup
@section Text markup
@cindex text markup
@cindex markup text


@cindex typeset text

LilyPond has an internal mechanism to typeset texts. You can access it
with the keyword @code{\markup}. Within markup mode, you can enter texts
similar to lyrics: simply enter them, surrounded by spaces:
@cindex markup

@lilypond[verbatim,fragment,relative=1]
 c1^\markup { hello }
 c1_\markup { hi there }
 c1^\markup { hi \bold there, is \italic anyone home? }
@end lilypond

@cindex font switching

The markup in the example demonstrates font switching commands.  The
command @code{\bold} and @code{\italic} only apply to the first
following word; enclose a set of texts with braces to apply a command
to more words:
@example
  \markup @{ \bold @{ hi there @} @}
@end example

@noindent
For clarity, you can also do this for single arguments, e.g.

@verbatim
  \markup { is \italic { anyone } home }
@end verbatim

@cindex font size, texts


In markup mode you can compose expressions, similar to mathematical
expressions, XML documents and music expressions.  The braces group
notes into horizontal lines. Other types of lists also exist: you can
stack expressions grouped with @code{<}, and @code{>} vertically with
the command @code{\column}. Similarly, @code{\center-align} aligns
texts by their center lines:

@lilypond[verbatim,fragment,relative=1]
 c1^\markup { \column < a bbbb c > }
 c1^\markup { \center-align < a bbbb c > }
 c1^\markup { \line < a b c > }
@end lilypond


Markups can be stored in variables, and these variables
may be attached to notes, like
@verbatim
allegro = \markup { \bold \large { Allegro } }
\notes { a^\allegro b c d }
@end verbatim


Some objects have alignment procedures of their own, which cancel out
any effects of alignments applied to their markup arguments as a
whole.  For example, the @internalsref{RehearsalMark} is horizontally
centered, so using @code{\mark \markup @{ \left-align .. @}} has no
effect. Similarly, whole texts over notes cannot be moved vertically
with @code{\raise}. For moving and aligning complete objects, grob
properties should be used.



@seealso

Init files:  @file{scm/new-markup.scm}.


@refbugs

Text layout is ultimately done by @TeX{}, which does kerning of
letters.  LilyPond does not account for kerning, so texts will be
spaced slightly too wide.

Syntax errors for markup mode are confusing.

Markup texts cannot be used in the titling of the @code{\header}
field. Titles are made by La@TeX{}, so La@TeX{} commands should be used
for formatting.



@menu
* Overview of text markup commands::  
* Markup construction in scheme::  
* Markup command definition::   
@end menu

@node  Overview of text markup commands
@subsection Overview of text markup commands

@include markup-commands.tely

@node Markup construction in scheme
@subsection Markup construction in scheme

@cindex defining markup commands 

The @code{markup} macro builds markup expressions in Scheme while
providing a LilyPond-like syntax. For example,
@example
(markup #:column (#:line (#:bold #:italic "hello" #:raise 0.4 "world")
                  #:bigger #:line ("foo" "bar" "baz")))
@end example

@noindent
is equivalent to:
@example
\markup \column < @{ \bold \italic "hello" \raise #0.4 "world" @}
                  \bigger @{ foo bar baz @} >
@end example

@noindent
This example exposes the main translation rules between regular
LilyPond markup syntax and scheme markup syntax, which are summed up
is this table:
@multitable @columnfractions .5 .5
@item @b{LilyPond} @tab @b{Scheme}
@item @code{\command} @tab @code{#:command}
@item @code{\variable} @tab @code{variable}
@item @code{@{ ... @}} @tab @code{#:line ( ... )}
@item @code{\center-align < ... >} @tab @code{#:center ( ... )}
@item @code{string} @tab @code{"string"}
@item @code{#scheme-arg} @tab @code{scheme-arg}
@end multitable

Besides, the whole scheme language is accessible inside the
@code{markup} macro: thus, one may use function calls inside
@code{markup} in order to manipulate character strings for
instance. This proves useful when defining new markup commands (see
@ref{Markup command definition}).

@refbugs

One can not feed the @code{#:line} (resp @code{#:center},
@code{#:column}) command with a variable or the result of a function
call. Eg:
@lisp
(markup #:line (fun-that-returns-markups))
@end lisp
is illegal. One should use the @code{make-line-markup} (resp
@code{make-center-markup}, @code{make-column-markup}) function
instead:
@lisp
(markup (make-line-markup (fun-that-returns-markups)))
@end lisp

@node Markup command definition
@subsection Markup command definition

New markup commands can be defined thanks to the @code{def-markup-command} scheme macro.
@lisp
(def-markup-command (@var{command-name} @var{paper} @var{props} @var{arg1} @var{arg2} ...)
            (@var{arg1-type?} @var{arg2-type?} ...)
  ..command body..)

    @var{argi}: i@var{th} command argument
    @var{argi-type?}: a type predicate for the i@var{th} argument
    @var{paper}: the `paper' definition
    @var{props}: a list of alists, containing all active properties. 
@end lisp

As a simple example, we show how to add a @code{\smallcaps} command,
which selects @TeX{}'s small caps font.  Normally, we could select the
small caps font as follows:

@verbatim
  \markup { \override #'(font-shape . caps)  Text-in-caps }
@end verbatim

This selects the caps font by setting the @code{font-shape} property to
@code{#'caps} for interpreting @code{Text-in-caps}.

To make the above available as @code{\smallcaps} command, we have to
define a function using @code{def-markup-command}. The command should
take a single argument, of markup type. Therefore, the start of the
definition should read
@example
  (def-markup-command (smallcaps paper props argument) (markup?)
@end example

@noindent

What follows is the content of the command: we should interpret
the @code{argument} as a markup, i.e.

@example
    (interpret-markup paper  @dots{} argument)
@end example

@noindent
This interpretation should add @code{'(font-shape . caps)} to the active
properties, so we substitute the the following for the @dots{} in the
above example:

@example
 (cons (list '(font-shape . caps) ) props)
@end example

@noindent
The variable @code{props} is a list of alists, and we prepend to it by
consing a list with the extra setting.

However, suppose that we are using a font that does not have a
small-caps variant. In that case, we have to fake the small caps font,
by setting a string in upcase, with the first letter a little larger:

@example
#(def-markup-command (smallcaps paper props str) (string?)
   "Print the string argument in small caps. Syntax: \\smallcaps #\"string\""
   (interpret-markup paper props
    (make-line-markup
     (map (lambda (s)
            (if (= (string-length s) 0)
                s
                (markup #:large (string-upcase (substring s 0 1))
                        #:translate (cons -0.6 0)
                        #:tiny (string-upcase (substring s 1)))))
          (string-split str #\Space)))))
@end example

The @code{smallcaps} command first splits its string argument into
tokens separated by spaces (@code{(string-split str #\Space)}); for
each token, a markup is built with the first letter made large and
upcased (@code{#:large (string-upcase (substring s 0 1))}), and a
second markup built with the following letters made tiny and upcased
(@code{#:tiny (string-upcase (substring s 1))}). As LilyPond
introduces a space between markups on a line, the second markup is
translated to the left (@code{#:translate (cons -0.6 0) ...}). Then,
the markups built for each token are put in a line
(@code{(make-line-markup ...)}). Finally, the resulting markup is
passed to the @code{interpret-markup} function, with the @code{paper}
and @code{props} arguments.

Finally, suppose that we are typesetting a recitative in an opera, and
we would like to define a command that will show character names in a
custom manner. Names should be printed with small caps and translated a
bit to the left and top.  We will define a @code{\character} command
that takes into account the needed translation, and uses the newly
defined @code{\smallcaps} command:

@verbatim
#(def-markup-command (character paper props name) (string?)
   "Print the character name in small caps, translated to the left and
   top. Syntax: \\character #\"name\""
   (interpret-markup paper props 
    (markup "" #:translate (cons -4 2) #:smallcaps name)))
@end verbatim

There is one complication that needs explanation: texts above and below
the staff are moved vertically to be at a certain distance (the
@code{padding} property) from the staff and the notes. To make sure
that this mechanism does not annihilate the vertical effect of our
@code{#:translate}, we add an empty string (@code{""}) before the
translated text.  Now the @code{""} will be put above the notes, and the
@code{name} is moved in relation to that empty string. The net effect is
that the text is moved to the upper left.

The final result is as follows:
@verbatim
\score {
    \notes { \fatText
        c''^\markup \character #"Cleopatra"
        e'^\markup \character #"Giulio Cesare"
    }
}
@end verbatim

@lilypond[raggedright]
#(def-markup-command (smallcaps paper props str) (string?)
   "Print the string argument in small caps. Syntax: \\smallcaps #\"string\""
   (interpret-markup paper props
    (make-line-markup
     (map (lambda (s)
            (if (= (string-length s) 0)
                s
                (markup #:large (string-upcase (substring s 0 1))
                        #:translate (cons -0.6 0)
                        #:tiny (string-upcase (substring s 1)))))
          (string-split str #\Space)))))

#(def-markup-command (character paper props name) (string?)
   "Print the character name in small caps, translated to the left and
   top. Syntax: \\character #\"name\""
   (interpret-markup paper props 
    (markup "" #:translate (cons -4 0) #:smallcaps name)))

\score {
    \notes { \fatText
        c''^\markup \character #"Cleopatra"
        e'^\markup \character #"Giulio Cesare"
    }
}
@end lilypond



@node Global layout
@section Global layout

The global layout determined by three factors: the page layout, the
line breaks and the spacing. These all influence each other. The
choice of spacing determines how densely each system of music is set,
which influences where line breaks breaks are chosen, and thus
ultimately how many pages a piece of music takes. This section
explains how to tune the algorithm for spacing.

Globally spoken, this procedure happens in three steps: first,
flexible distances (``springs'') are chosen, based on durations. All
possible line breaking combination are tried, and the one with the
best results---a layout that has uniform density and requires as
little stretching or cramping as possible---is chosen. When the score
is processed by @TeX{}, each page is filled with systems, and page breaks
are chosen whenever the page gets full.



@menu
* Vertical spacing::            
* Horizontal spacing::          
* Font Size::                   
* Line breaking::               
* Page layout::                 
@end menu


@node Vertical spacing
@subsection Vertical spacing

@cindex vertical spacing
@cindex distance between staves
@cindex staff distance
@cindex between staves, distance
@cindex staffs per page
@cindex space between staves

The height of each system is determined automatically by LilyPond, to
keep systems from bumping into each other, some minimum distances are
set.  By changing these, you can put staves closer together, and thus
put more  systems onto one page.

Normally staves are stacked vertically. To make
staves maintain a distance, their vertical size is padded. This is
done with the property @code{minimumVerticalExtent}. It takes a pair
of numbers, so if you want to make it smaller from its, then you could
set
@example
  \set Staff.minimumVerticalExtent = #'(-4 . 4)
@end example
This sets the vertical size of the current staff to 4 staff spaces on
either side of the center staff line.  The argument of
@code{minimumVerticalExtent} is interpreted as an interval, where the
center line is the 0, so the first number is generally negative.  The
staff can be made larger at the bottom by setting it to @code{(-6
. 4)}.

The piano staves are handled a little differently: to make cross-staff
beaming work correctly, it is necessary that the distance between staves
is fixed beforehand.  This is also done with a
@internalsref{VerticalAlignment} object, created in
@internalsref{PianoStaff}. In this object the distance between the
staves is fixed by setting @code{forced-distance}. If you want to
override this, use a @code{\translator} block as follows:
@example
  \paper @{
    \translator @{
      \PianoStaffContext
      \override VerticalAlignment #'forced-distance = #9
    @}
    @dots{}
  @}
@end example
This would bring the staves together at a distance of 9 staff spaces,
measured from the center line of each staff.

@seealso

Internals: Vertical alignment of staves is handled by the
@internalsref{VerticalAlignment} object.




@node Horizontal spacing
@subsection Horizontal Spacing

The spacing engine translates differences in durations into
stretchable distances (``springs'') of differing lengths. Longer
durations get more space, shorter durations get less.  The shortest
durations get a fixed amount of space (which is controlled by
@code{shortest-duration-space} in the @internalsref{SpacingSpanner} object). 
The longer the duration, the more space it gets: doubling a
duration adds a fixed amount (this amount is controlled by
@code{spacing-increment}) of space to the note.

For example, the following piece contains lots of half, quarter and
8th notes, the eighth note is followed by 1 note head width (NHW). 
The quarter note is followed by 2 NHW, the half by 3 NHW, etc.
@lilypond[fragment,verbatim,relative=1] c2 c4. c8 c4. c8 c4. c8 c8
c8 c4 c4 c4
@end lilypond

Normally, @code{shortest-duration-space} is set to 1.2, which is the
width of a note head, and @code{shortest-duration-space} is set to
2.0, meaning that the shortest note gets 2 NHW (i.e. 2 times
@code{shortest-duration-space}) of space. For normal notes, this space
is always counted from the left edge of the symbol, so the shortest
notes are generally followed by one NHW of space.

If one would follow the above procedure exactly, then adding a single
32th note to a score that uses 8th and 16th notes, would widen up the
entire score a lot. The shortest note is no longer a 16th, but a 32nd,
thus adding 1 NHW to every note. To prevent this, the
shortest duration for spacing is not the shortest note in the score,
but the most commonly found shortest note.  Notes that are even
shorter this are followed by a space that is proportional to their
duration relative to the common shortest note.  So if we were to add
only a few 16th notes to the example above, they would be followed by
half a NHW:

@lilypond[fragment,verbatim,relative=2]
 c2 c4. c8 c4. c16[ c] c4. c8 c8 c8 c4 c4 c4
@end lilypond

The most common shortest duration is determined as follows: in every
measure, the shortest duration is determined. The most common short
duration, is taken as the basis for the spacing, with the stipulation
that this shortest duration should always be equal to or shorter than
1/8th note. The shortest duration is printed when you run lilypond
with @code{--verbose}.  These durations may also be customized. If you
set the @code{common-shortest-duration} in
@internalsref{SpacingSpanner}, then this sets the base duration for
spacing. The maximum duration for this base (normally 1/8th), is set
through @code{base-shortest-duration}.

@cindex @code{common-shortest-duration}
@cindex @code{base-shortest-duration}
@cindex @code{stem-spacing-correction}
@cindex @code{spacing}

In the introduction it was explained that stem directions influence
spacing. This is controlled with @code{stem-spacing-correction}
property in @internalsref{NoteSpacing}, which are generated for every
@internalsref{Voice} context. The @code{StaffSpacing} object
(generated at @internalsref{Staff} context) contains the same property
for controlling the stem/bar line spacing. The following example
shows these corrections, once with default settings, and once with
exaggerated corrections:

@lilypond
    \score { \notes {
      c'4 e''4 e'4 b'4 |
      b'4 e''4 b'4 e''4|
      \override Staff.NoteSpacing   #'stem-spacing-correction
   = #1.5
      \override Staff.StaffSpacing   #'stem-spacing-correction
   = #1.5
      c'4 e''4 e'4 b'4 |
      b'4 e''4 b'4 e''4|      
    }
    \paper { raggedright = ##t } }
@end lilypond

@cindex SpacingSpanner, overriding properties

Properties of the  @internalsref{SpacingSpanner} must be overridden
from the @code{\paper} block, since the @internalsref{SpacingSpanner} is
created before any property commands are interpreted.
@example
\paper @{ \translator  @{
  \ScoreContext
  SpacingSpanner \override #'spacing-increment = #3.0
@} @}
@end example


@seealso

Internals: @internalsref{SpacingSpanner}, @internalsref{NoteSpacing},
@internalsref{StaffSpacing}, @internalsref{SeparationItem}, and
@internalsref{SeparatingGroupSpanner}.

@refbugs

Spacing is determined on a score wide basis. If you have a score that
changes its character (measured in durations) halfway during the
score, the part containing the longer durations will be spaced too
widely.

There is no convenient mechanism to manually override spacing.



@node Font Size
@subsection Font size
@cindex font size, setting
@cindex staff size, setting
@cindex @code{paper} file

The Feta font provides musical symbols at eight seven different
sizes. Each font is tuned for a different staff size: at smaller sizes
the font gets heavier, to match the relatively heavier staff lines.
The recommended font sizes are listed in the following table:

@multitable @columnfractions  .25 .25 .25 .25

@item @b{name}
@tab @b{staff height (pt)}
@tab @b{staff height (mm)}
@tab @b{use}

@item feta11
@tab 11.22
@tab 3.9 
@tab pocket scores

@item feta13
@tab 12.60pt
@tab 4.4mm
@tab

@item feta14
@tab 14.14pt
@tab 5.0mm
@tab 

@item feta16
@tab 15.87pt
@tab 5.6mm
@tab 

@item feta18
@tab 17.82pt
@tab 6.3mm
@tab song books

@item feta20
@tab 17.82pt
@tab 7.0mm
@tab standard parts 

@item feta23
@tab 22.45 pt
@tab 7.9mm
@tab 

@item feta20
@tab 25.2 pt
@tab 8.9mm
@tab
@c modern rental material  ?

@end multitable

These fonts are available in any sizes. The context property
@code{fontSize} and the layout property @code{staff-space} (in
@internalsref{StaffSymbol}) can be used to tune size for individual
staffs. The size of individual staffs are relative to the global size,
which can be set   in the following manner:

@example
  #(set-global-staff-size 14)
@end example

This sets the global default size to 14pt staff height, and scales all
fonts accordingly.



@node Line breaking
@subsection Line breaking

@cindex line breaks
@cindex breaking lines

Line breaks are normally computed automatically. They are chosen such
that lines look neither cramped nor loose, and that consecutive lines
have similar density.

Occasionally you might want to override the automatic breaks; you can
do this by  specifying @code{\break}. This will force a line break at
this point.  Line breaks can only occur at places where there are bar
lines.  If you want to have a line break where there is no bar line,
you can force an invisible bar line by entering @code{\bar
""}. Similarly, @code{\noBreak} forbids a line break at a 
point.


@cindex regular line breaks
@cindex four bar music. 

For line breaks at regular intervals  use @code{\break} separated by
skips and repeated with @code{\repeat}:
@example
<<  \repeat unfold 7 @{
         s1 \noBreak s1 \noBreak
         s1 \noBreak s1 \break  @}
   @emph{the real music}
>> 
@end  example

@noindent
This makes the following 28 measures (assuming 4/4 time) be broken every
4 measures, and only there.

@refcommands

@code{\break}, @code{\noBreak}
@cindex @code{\break}
@cindex @code{\noBreak}

@seealso

Internals: @internalsref{BreakEvent}.


@node Page layout
@subsection Page layout

@cindex page breaks
@cindex breaking pages

@cindex @code{indent}
@cindex @code{linewidth}

The most basic settings influencing the spacing are @code{indent} and
@code{linewidth}. They are set in the @code{\paper} block. They
control the indentation of the first line of music, and the lengths of
the lines.

If  @code{raggedright} is set to true in the @code{\paper}
block, then the lines are justified at their natural length. This
useful for short fragments, and for checking how tight the natural
spacing is.

@cindex page layout
@cindex vertical spacing

The page layout process happens outside the LilyPond formatting
engine: variables controlling page layout are passed to the output,
and are further interpreted by @code{lilypond} wrapper program. It
responds to the following variables in the @code{\paper} block.  The
variable @code{textheight} sets the total height of the music on each
page.  The spacing between systems is controlled with
@code{interscoreline}, its default is 16pt.  The distance between the
score lines will stretch in order to fill the full page
@code{interscorelinefill} is set to a positive number.  In that case
@code{interscoreline} specifies the minimum spacing.

@cindex @code{textheight}
@cindex @code{interscoreline}
@cindex @code{interscorelinefill}

If the variable @code{lastpagefill} is defined,
@c fixme: this should only be done if lastpagefill= #t 
systems are evenly distributed vertically on the last page.  This
might produce ugly results in case there are not enough systems on the
last page.  The @command{lilypond-book} command ignores
@code{lastpagefill}.  See @ref{lilypond-book manual} for more
information.

@cindex @code{lastpagefill}

Page breaks are normally computed by @TeX{}, so they are not under
direct control of LilyPond.  However, you can insert a commands into
the @file{.tex} output to instruct @TeX{} where to break pages.  This
is done by setting the @code{between-systems-strings} on the
@internalsref{NonMusicalPaperColumn} where the system is broken.
An example is shown in @inputfileref{input/regression,between-systems.ly}.
The predefined command @code{\newpage} also does this.

@cindex paper size
@cindex page size
@cindex @code{papersize}

To change the paper size, use the following Scheme code:
@example
        \paper@{
           #(set-paper-size "a4")
        @}
@end example


@refcommands

@cindex @code{\newpage}
@code{\newpage}. 


@seealso

In this manual: @ref{Invoking lilypond}.

Examples: @inputfileref{input/regression,between-systems.ly}.

Internals: @internalsref{NonMusicalPaperColumn}.

@refbugs

LilyPond has no concept of page layout, which makes it difficult to
reliably choose page breaks in longer pieces.




@node Sound
@section Sound
@cindex Sound

Entered music can also be converted to MIDI output.  The performance
is good enough for proof-hearing the music for errors.

Ties, dynamics and tempo changes are interpreted.  Dynamic marks,
crescendi and decrescendi translate into MIDI volume levels.  Dynamic
marks translate to a fixed fraction of the available MIDI volume
range, crescendi and decrescendi make the volume vary linearly between
their two extremities.  The fractions can be adjusted by
@code{dynamicAbsoluteVolumeFunction} in @internalsref{Voice} context.
For each type of MIDI instrument, a volume range can be defined.  This
gives a basic equalizer control, which can enhance the quality of
the MIDI output remarkably.  The equalizer can be controlled by
setting @code{instrumentEqualizer}.

@refbugs

Many musically interesting effects, such as swing, articulation,
slurring, etc., are not translated to MIDI.

Since slurs are not interpreted, @code{\lyricsto} and
@code{\addlyrics} sections will be interpreted wrongly.

The MIDI output allocates a channel for each Staff, and one for global
settings.  Hence, the MIDI file should not have more than 15 staves
(or 14 if you do not use drums).


@menu
* MIDI block::                  
* MIDI instrument names::       
@end menu


@node MIDI block
@subsection MIDI block
@cindex MIDI block


The MIDI block is analogous to the paper block, but it is somewhat
simpler.  The @code{\midi} block can contain:
@cindex MIDI block

@itemize @bullet
  @item a @code{\tempo} definition, and
  @item context definitions.
@end itemize

Assignments in the @code{\midi} block are not allowed.

A number followed by a period is interpreted as a real number, so
for setting the tempo for dotted notes, an extra space should be
inserted, for example:

@example
  \midi @{ \tempo 4 . = 120 @} 
@end example


@cindex context definition

Context definitions follow precisely the same syntax as within the
\paper block.  Translation modules for sound are called performers.
The contexts for MIDI output are defined in @file{ly/performer-init.ly}.


@node MIDI instrument names
@subsection MIDI instrument names

@cindex instrument names
@cindex @code{Staff.midiInstrument}

The MIDI instrument name is set by the @code{Staff.midiInstrument}
property.  The instrument name should be chosen from the list in
@ref{MIDI instruments}.

@refbugs

If the selected string does not exactly match, then the default is
used, which is the Grand Piano.