Philosophy of algorithmic composition (how to generate a music...)
The music program code is built to resemble as much as possible the true structure of music, as it is recognized by contemporary music theory… Thus, some enumerated objects and classes directly correspond to some musical elements or properties.
to achieve simplicity and clarity ... (reduction of combinatorial system variety ...)
Composers never use all possible musical elements - in a given musical phrase, even in the whole song. Reducing infinite variability simplifies music and makes it easier to understand. Simplification is also beneficial because that allows the composer to create patterns. Something that covers the whole set is not as clear as the relationship of two smaller similar subsets.
1. Selection of systemThe selection of pitches (frequencies) for musical compositions does not (with some exceptions) form a continuous spectrum, but the spectrum is discreet. Only certain selected tones are used. The same is true for "tics" (or tone lengths) in rhythmic patterns. On this most general level, we are talking about a harmonic and rhythmic system.
Not all tones (harmonic systems) or tics (rhythmic systems) should be used in a musical phrase. Musical phrases often respect harmonic modality (mode, scale) and rhythmic modality (selection of possible onset tones). Modality is a subset of a set of all tones or tics. All possible structures can be represented as binary (or ternal) numbers of length k.
Restrictions that were initiated using systems and modalities continue ...
Not all harmonic structures are to be used.
Modality together with harmonic variety and harmonic functions determines the key.
Each individual composition (written by a certain composer in a given stylistic period) uses a limited number of musical (harmonic, rhythmic and melodic) schemes, which are variously repeated, changed and combined. The abstract description (image) of a composition is called a musical model and the individual basic schemes used are the so-called motifs. Each motif is a sequence of certain musical elements. The harmonic motif consists of measures filled with harmonies, the rhythmic motif consists of measures with a determined rhythm. (The harmonic motif itself also has a rhythmic, but this is simplified - only a few basic patterns are considered. )
The melodic motif is not (unlike the harmonic and rhythmic motif) divided into measures,
but into segments, each segment consisting of one or more melodic fragments.
One melodic fragment generally spans several bars.
Only harmonic and rhythmic motifs can be divided into individual measures in advance,
melodic motifs are not so independent -
the number of tones in a bar is determined by the rhythm
and the pitch of the tones is conditioned by harmony.
If the melodic motif is exhausted earlier, it is repeated again (figural movement).
If, on the other hand, it is longer than the number of possible rhythmic locations,
it remains unfinished and at the desired time a new motif (part of the motif, fragment) enters.
When compiling motifs, it is possible to limit the musical material by modality (harmonic, rhythmic)
and possibly also by a qualifier (condition).
It is reasonable to assign melodic motives to some voises in
given phrase. But this voices need not to respect given motives exactly.
All voices should respect given harmonic stream (according to actual
value of harmonic cover) and also other rules (acoustical qualities of tone clusters...)
Music should respects some given form, e.g. some bars of musical stream have to be similar to other bars... But not all. Music must change!
Music that have the same number of tones in all bars is boring.
Here is a place for application of physical laws (velocity,
acceleration,...).
Usage of accents and passages of higher loudness is needed
when the inner dissonance (impulse) of chords increases.
Energy input should be always increased by consumed energy.
(see Music Theory Excerpts - Dynamics)
Unstable chords, i.e. chords having lower potential energy
can have (like dissonant chords) a bit higher loudness.
Unregular rhytmical patterns also need more energy.
Ending of phrases
The last bars of phrases ussualy release energy.
This is done with help of simplification, using sequences of continuity
and using tonic and consonant chords.
Reducing of complicated rhytmical pattterns, increase of continuity
and increase of musical potential or increase of consonance amount appears at the end of many phrases.
Music as a whole should be nice. This does not mean that musician can't use e.g. dissonant chords. But some formations of tones are very unpleasant and these have to be eliminated from music..
Tone clusters
Every cluster of tones (tones sounding at the same time) have to be
checked for collisions of impulse.
Lower tones cannot be treated the same way as higher tones.
Higher and middle tones allow higher density of chords.
Density of harmonic stream should be within a given limits in
a musical phrase.
E.g. interchange of some 2-tone and 5-tone clusters or some very
consonant and very dissonant chords is unpleasant.
Homogenity of music cohere with musical styles, e.g. triads
dominate in clasical music, while four or more tone chords
in impresionism...
Auxiliary voices are usually inner voices that fill and highlight
harmonic stream.
The progress of these voices should (in every harmonic connection)
minimize sum of distances in particular voices.
Only such leading do not disrupt impression and enable the other (main) voices
to be free in motion.
The basic concept for creating new songs is the idea that they can be composed of 4 main components - harmony, tectonics, style and orchestration: MusicalComponentType (Harmony, Tectonic, Style, Orchestration). Tectonics here determines the general construction of the composition and the style of melodic-rhythmic patterns used (see MelodicPattern, MelodicPatternVoice, RhythmicPattern, RhythmicPatternVoice).
Transformation means the modification of a given composition, usually by adjusting individual tones in the voices. MusicalTransformation (InvertOctaves, InvertScore (VerticalInversion), VerticalExtension, VerticalNarrowing, ModularDeformation, OctaveUp, OctaveDown). However, in order to preserve the harmony of the music, it is necessary to follow certain rules - for example, to transform only one selected voice, to recompose the song again to complete the exact positions of the tones, etc.
Mutation means the replacement of selected motifs or structures of a given composition with motifs / structures from another composition. When deciding which rhythm pattern to insert into a given measure, the "energy" (eg rhythm mobility) of the given measure in the initial composition can be taken into account. Substitution variants are determined by MutationPairing (Motives, Structures, Energy).
New harmonic / rhythmic / melodic combinations...
The compositions are created in a regular (tempered) 12-tone system, the aim is to enable refinement to multitone microtonal systems in the future (see PitchWheelStep).