Education and demonstration – Occupation
Reexamination Certificate
2002-07-29
2004-06-08
Rovnak, John Edmund (Department: 3714)
Education and demonstration
Occupation
C434S319000, C084S609000
Reexamination Certificate
active
06746246
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to apparatus for composing a song, and to a method of composing a song. It has particular relevance to adaptively modifying at least one original composed song to produce a more popular song. The song may, for example, be suitable for recording on a data carrier such as a compact disc, or for broadcast by a radio station. It will be understood that in the context of this application by song we mean any musical song or tune, with or without a vocal element, and with or without elements of melody or harmony, where the musical content of the song is at least partially comprised of predefined audio samples.
2. Description of Related Art
Many popular songs today are produced by stringing together predefined (usually pre-recorded) audio samples. This is especially true for dance music, where one or two short audio samples are played in sequence to form a part of a song. Samples are small phrases of sound, perhaps as short as a single drum beat but more usually lasting from one to maybe eight or sixteen measures at a particular tempo. These samples are most often copied from existing music produced by popular recording artists, such as a guitar phrase from a rock track, or a drum phrase from a reggae track. They are most often samples of audio produced by instruments but may also be samples of singing.
Often a song may be constructed which includes a track made up of a single sample (such as a four measure drum beat pattern lasting 8 seconds) repeated a large number of times during the duration of the song.
This style of music is now so widely accepted that the legalities of the use of the samples (such as obtaining copyright clearance) are readily understood. Furthermore, an entire industry has been established around the provision of sets of copyright free or “copyright cleared” samples which can be used by authors to produce new songs.
Until recently, composing a new song required the use of a sampler. This is a solid-state random-access audio recorder, trigerrable via MIDI from the keys of a keyboard. The samples are allocated to the keys and the user presses the keys in turn to play the samples in the preferred order from its memory. This requires a good degree of musicianship and strong keyboard skills to produce good results. There is now a variety of proprietary composing software available, which allows songs to be composed on a basic personal computer and played back in real time. Each of the available software packages differ in their detailed implementation but it is typical for the user to be presented with an “arrange” screen which displays the bars of a track onto which samples can be placed. The user can control the number of tracks, the tempo, and can choose which samples to place in each track. Once the samples have been dropped onto the screen the song can be played back to listen to the results. If the user is unhappy the samples can be moved around or replaced until a satisfactory result is obtained.
Although the composition of songs from samples is now very simple the user must still possess a great degree of skill and a good ear for music if a good song is to be produced. Samples cannot simply be placed at random in a track. Firstly, the user must judge when the tracks are to be silent or contain samples. Secondly, the individual samples must be chosen which must mix well with the other samples in the song.
At present, the author is not aware of any automated method of constructing good songs from samples. A computer cannot be readily trained to identify a good song from a bad song and so is unable to construct songs from samples.
BRIEF SUMMARY OF THE INVENTION
According to a first aspect the invention provides a method of automatically composing at least one musical song with a programmed processor comprising the steps of: defining and storing a set of songs which are each defined by a song template, the song template comprising at least one track comprising a sequence of measures and a plurality of audio samples allocated to the or each track, playing at least two of the songs to an audience; obtaining response signals from the audience indicating the popularity of each song with the audience; the programmed processor selecting at least one of the played songs in accordance with the popularity of that song with the audience; the programmed processor composing at least one song by automatically applying one or more rules to the template of the selected song, the rules modifying one or more characteristics of the template to produce a modified song; and the programmed processor replacing the less popular song with the composed modified song in the set of songs.
In a preferred arrangement, a song may be more likely to be selected for replacement the more unpopular a song is. The most unpopular song may always be selected. Otherwise, the song may be selected at random, perhaps with a weighting so that less popular songs are more likely to be selected, with the most popular song being the least likely to be selected. Other non-random selection strategies may be provided.
By utilising a measurement of the reaction of an audience to a set of initial songs and composing at least one song by automatically modifying at least one less popular song from the set the method can selectively compose a set of songs in such a way that unpopular songs are rejected over time. This will eventually leave a set of composed songs more popular than the initial set.
By automatically modifying the selected song we may mean that the rules are applied to the selected song without the need for human intervention to determine how the rules are applied.
The rules used to modify the selected song may include a rule whereby one or more of the characteristics of the song template for the selected song are replaced by one or more characteristics of at least one of the more popular songs from the set. This type of modification can be considered to be equivalent to a genetic recombination where the characteristics of a song template, such as the samples used, are equivalent to the genes of the song. Alternatively, the rules may comprise replacing one or more of the characteristic features of the template of a selected song at random with alternative characteristic features. For example, a sample may be replaced by a randomly selected alternative sample. This is the equivalent to a genetic mutation, which allows the song to evolve randomly over time. A combination of a mutation and a recombination may be applied to a selected song.
In a preferred method—in which samples or sets of samples or tracks are replaced with samples or sets of samples or tracks from a more popular song—the set of songs will evolve over time in an analogy to genetic evolution. They may be replaced with samples or sets of samples or tracks from more than one more popular song—just as animals are more likely to interbreed with fitter (more popular) mates to produce a more fit set of offspring.
In nature, forces such as the environment and predators determine which mutations and recombinations are successful. In the method of the invention the responses of the audience provide the equivalent force, which determines whether a song is modified or survives. Only the most popular songs will survive.
According to a second aspect the invention provides a song composition apparatus comprising: a memory containing a store of audio samples and a set of songs which are each defined by a song template, the song template comprising: at least one track comprising a sequence of measures and a plurality of audio samples allocated from the store of audio samples to the or each track, sound reproduction apparatus adapted to play at least two of the set of songs to an audience, one or more monitoring devices adapted to monitor the response of the audience to the played songs to provide crowd reaction signals representative of the popularity of each played song with the audience; a processor programmed to select on the basis of the crowd reaction signals at least one o
Hewlett--Packard Development Company, L.P.
Rovnak John Edmund
LandOfFree
Method and apparatus for composing a song does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for composing a song, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for composing a song will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3349915