Introduction to Hybridized Interactive Algorithmic Composition Model

Music composition (by machine) requires the solution of a number of difficult problems in the fields of algorithm design, data representation, human interface design, and software engineering in general. These aforementioned problems led to the formulation of an algorithmic composition model from the existing algorithmic composition models – where their strengths were harnessed and their weaknesses transparently subdued.This blog post provides an Introduction to Hybridized Interactive Algorithmic Composition Model.

This brought about the hybridization of the existing models that gave birth to Hybridized Interactive Algorithmic Composition model. This hybridized model was analyzed using Unified Modeling Language (UML) as an Object-Oriented Analysis and Design tool. The analysis gave rise to the development of an Object-Oriented software framework for the hybridized model, presented as Java objects with generated JavaDoc hypertext documentation for further developmental references.

These Java objects, packaged as API with JavaDoc documentations, have succeeded in providing music/multimedia developers with necessary building blocks for implementing and realizing applications which require musical features.

The Hybridized Interactive Algorithmic Composition model was developed by Dr. Etemi Joshua Garba for Ethereal Multimedia Technology.

 

Introduction to Hybridized Interactive Algorithmic Composition Model

Music is the skillful arrangement and alternation of sound and silence within a given period of time with appreciable aural perception. When musicians play instruments, sound waves are produced (as the sound-producing parts of the instruments vibrate/oscillate). These vibrations cause air molecules to displace one another in a systematic continuous flow, which moves away from its source. In the end, our ears perceive these air pressure fluctuations as sounds that are translated by our brains as music.

Musically speaking, the sounds produced (which are pleasant to our ears) are musical notes and the silences/pauses between the notes are called rests. The musical notes and rests vary in duration. That is, a note could be played for a long or short time. Thus, in contemporary music, there are six types of notes – whole, half, quarter, eighth, sixteenth and thirty-second.

Table 1 Musical Notes/Rests Interpretation

musical_notes_rests_interpretation Ethereal Multimedia Technology www.ethereal.ng

Computability of Music

Music is the alternation of sound and silence. It is therefore obvious that the concept of music conforms to binary theory; where only the values of 0 and 1 are considered. The binary value of “0” indicates certainly that an event will not occur; while “1” indicates certainly that the event will occur.

Logically put, when sound is produced (or a note/tone is played) the binary value is 1, otherwise it is 0. That is:

  • 1 – sound is produced (or a note/tone is played)
  • 0 – silence (no note is played)

Therefore, the binary musical concept perceives music as a stream of 1s and 0s. This therefore, means that musical ideas and concepts are logical – hence computable.

Hybridized Interactive Algorithmic Composition Model

Actually, improvisation is a very challenging craft which requires the coordination of many factors at the same time such as chords and scales structures, form, rhythm, articulation, patterns and feeling. The art of improvising requires that a musician should have deep knowledge of all that and be skillful at both theoretical and practical levels.

Music improvisation depends on performing or composing piece of music, without any preparation or set text to follow. Therefore, it is clear that there are 216 (65,536) possible rhythms for just one bar/measure with the common meter (4/4). Furthermore, if the notes (which make up the rhythm) are pitched to form a melody line (from the notes of the Chromatic scale and octaves), then millions of musical themes could be produced – based on the probability and randomization of the notes. This becomes an enormous complex task for the human brain to comprehend.

The Hybridized Interactive Algorithmic Composition (HIAC) Model provides the possibility of taking the leverage of speed and accuracy of the computer to complement human creativity in music improvisation and composition. The Hybridized Interactive Algorithmic Composition model is a galvanization of the existing algorithmic composition models so as to minimize the weaknesses experienced when such models are used singlehandedly in music composition.

The Hybridized Interactive Algorithmic Composition model also supports the process of creative musical improvisation and composition in an interactive environment and the resulting needs of the creator which may be addressed through new software tools. To provide an enabling environment where the user/musician would be able to combine the skills of composition and improvisation based on composition algorithms.

Some of the benefits of this model include:

  • Enabling interactivity in algorithmic composition.
  • Creation of an enabling environment to teach beginners music improvisation and composition skills.
  • The possibility of creating memorizable and improvisable musical themes with just a click of the mouse.
  • The possibility of generating musical themes by people with no musical skills or knowledge.
  • The ability to complement the effort of the musicians in creating musical themes within the digital domain. This is where creativity is driven by Music Technology.
  • The possibility of discovering new music styles and genres.

Therefore, computer programming provides an inevitable gateway to music synthesis, generation and composition. This makes the production of soundtracks, jingles, melodies for TV, radio, film and special occasions very cost effective; while it serves as a boost to creativity in music composition and a means of developing and discovering new complex musical concepts.

Various models/software in the area of computer aided music composition support music composition either automatically or interactively. This article, however, gives a brief introduction to Hybridized Interactive Algorithmic Composition Model

In automatic computer aided music composition, the generation and synthesis of music is completely carried out by the computer (via the algorithm of the music software). In this case, the user of the software has no interference with the process of generating and synthesizing music. Though this process requires that the user have little or no knowledge of music, but it is not flexible. This is because the user cannot contribute musically to the process of music generation and synthesis.

On the other hand, interactive computer aided music composition is based on a specified balanced interactivity between the user (in this case, a composer – who has music knowledge) and the music creating software. This way, the user has the opportunity to compose music to his taste.

Also, most music composition software are based on the use of a single algorithmic music composition model to take care of all the stages of music composition (such as rhythm formation, variation, development, contrast, motif, pitching, pitch class segmentation and normalization, prime form creation and optimization, hierarchical mutation, transposition retrogradation, inversion and music score formulation).

However, the Hybridized Interactive Algorithmic Composition model provides the platform where human creativity in music improvisation and composition is complemented by the leverage of speed and accuracy of the computer. Also, the model is based on the hybridization of models (at different stages of composition) – in order to explore the advantages of such models. The models for algorithmic composition used include: Mathematical model, Grammar (Rule-Based) model, stochastic model, Knowledge-based model, and Evolutionary (Genetic Algorithm) model.

The development of this hybridized model was necessitated upon the fact that the existing algorithmic composition models can’t singlehandedly take care of all the stages of music composition. In essence, the hybridized model harnessed the strength of the existing models to overcome their weaknesses (when used individually).

It should be noted that the existing algorithmic composition models do not clearly distinguish the process of creation of both rhythm and melody. This makes music composition non-definitive as regards rhythm or melody creation. In the hybridized model, both rhythm and melody creation, generation and synthesis were however made explicit; such that music composition is made both interactive and flexible. Thus, a generated rhythm (for instance) could be applied to several melodic patterns to produce rhythm-based variations; on the other hand, a melodic line could be applied to a number of rhythmic patterns to produce melody-based variations.

This blog post, Introduction to Hybridized Interactive Algorithmic Composition Model is culled from: Hybridized Interactive Algorithmic Composition Model by E. J. Garba

 

Leave a Reply

Your email address will not be published. Required fields are marked *