Automatic composition of sound sequences using finite state automata

• US 7,902,447 B1
• Filed: 10/03/2006
• Issued: 03/08/2011
• Est. Priority Date: 10/03/2006
• Status: Active Grant

• Current Assignee: Sony Interactive Entertainment Inc. (Sony Group Corp.)
• Original Assignee: Sony Computer Entertainment Incorporated (Sony Group Corp.)
• Inventors: Abrego, Gustavo Hernandez
• Primary Examiner(s): Qin; Jianchun
• Application Number:
  US11/542,699
• Time in Patent Office: 1,617 Days
• Field of Search: 846/09
• US Class Current: 84/609
• CPC Class Codes: G10H 1/0025 Automatic or semi-automatic music composition, e.g. producing random music, applying rules from music theory or modifying a musical piece automatically producing a series of tones G10H1/26

Abstract

In one embodiment, a method for the automatic composition of music is disclosed. The method begins by receiving a plurality of input sound sequences containing sound frequencies with corresponding time duration. The method continues with converting the plurality of input sound sequences to a finite state automaton using a system that allows over-generation, followed by receiving exploration rules that constrain how the finite state automaton is to be traversed. The next step is creating a path marker data structure indexing a plurality of path markers, where each path marker contains a path marker history and a path marker registry. After the path marker data structure is created, the method continues by traversing the finite state automaton with a graph exploration procedure that uses the exploration rules and the plurality of path markers to determine paths across the finite state automaton. During the exploration the path marker history and the path marker registry of particular path markers are updated when traversing the finite state automaton. As the finite state automaton is traversed the method includes storing the paths across the finite state automaton to the path marker data structure to define recorded path markers, wherein the recorded path markers that are not found in the plurality of input sound sequences define new music compositions.

Claims

1. A method for the automatic composition of music, the method comprising:

receiving a plurality of input sound sequences containing sound frequencies with corresponding time duration;

converting the plurality of input sound sequences to a finite state automaton using a system that allows over-generation;

receiving exploration rules that constrain how the finite state automaton is to be traversed;

creating a path marker data structure indexing a plurality of path markers, where each path marker contains a path marker history and a path marker registry;

traversing the finite state automaton with a graph exploration procedure that uses the exploration rules and the plurality of path markers to determine path across the finite state automaton, such that the path marker history and the path marker registry of particular path markers are updated when traversing the finite state automaton; and

storing the paths across the finite state automaton to the path marker data structure to define recorded path markers;

wherein the recorded path markers that are not found in the plurality of input sound sequences define a new music composition.

2. A method for the automatic composition of music as recited in claim 1, wherein over-generation enables sharing of nodes of the finite state automaton.

3. A method for automatic music composition as recited in claim 2, wherein over-generation within the finite state automaton is facilitated by a user-defined history value that defines a number of previous nodes that must be identical before creating a shared node.

4. A method for automatic music composition as recited in claim 1 wherein the path marker registry is used to record links that are blocked.

5. A method for automatic music composition as recited in claim 1, wherein the exploration rules define how many times the graph exploration procedure can traverse a link before designating the link as blocked.

6. A method for automatic music composition as recited in claim 1, wherein the recorded path markers are recorded in the path marker structure.

7. A method for automatic music composition as recited in claim 1, wherein the exploration rules can be varied for different time periods in the plurality of input sound sequence.

8. A method for automatic music composition as recited in claim 1, wherein traversing the finite state automaton includes,

(a) creating a path marker at an origin node and recording the origin node as the contents of the path marker history for the origin node;

(b) following a link from the origin node to a destination node if the link is not blocked;

(c) recording the contents of the path marker history for the origin node and the destination node to a path marker history for the destination node unless the exploration rules are violated in which case it is recorded to a path marker registry for the origin node that the link between the origin node and the destination node is blocked;

(d) following any links from the origin node that have not been followed;

(e) if there is nothing recorded in the path marker registry for the origin node, deleting the path marker for origin node and renaming the destination node to the origin node;

(f) repeating steps (a)-(e) until reaching the end of the finite state automaton.

9. A method for generating new sound sequences based on input sounds as recited in claim 1, further comprising using a form of the recorded path markers defining a new sound sequence as input in a sound synthesizer to produce sounds.

10. A method for generating new sound sequences based on input sounds as recited in claim 1, further comprising converting the recorded path markers defining a new sound sequence into a written form.

11. A method for generating new sound sequences based on input sounds as recited in claim 1, further comprising converting the recorded path markers defining a new sound sequence into a form of music capable of being stored on a computer readable medium.

12. A computer readable media including program instructions for generating new sound sequences based on input sounds, comprising:

program instructions for receiving a plurality of input sound sequences containing sound frequencies and corresponding and time durations;

program instructions for converting the plurality of input sound sequences to a finite state automaton using a system that allows over-generation;

program instructions for traversing the finite state automaton using a graph exploration procedure that uses exploration rules and a plurality of path markers to determine paths across the finite state automaton; and

program instructions for storing the paths across the finite state automaton to a path marker data structure to define recorded path markers;

wherein the recorded path markers that are not found in the plurality of input sound sequences define a new sound sequences.

13. The computer readable media as recited in claim 12, wherein the path marker data structure indexes the plurality of path markers, where each path marker contains a path marker history and a path marker registry.

14. The computer readable media as recited in claim 12, wherein the exploration rules define how many times the graph exploration procedure can traverse a link before designating the link as blocked.

15. The computer readable media as recited in claim 12, wherein the program instructions for traversing the finite state automaton includes,

(a) program instructions for creating a path marker at an origin node and recording the origin node in a path marker history for the origin node;

(b) program instructions for following a link from the origin node to a destination node if the link is not blocked;

(c) program instructions for recording the contents of the path marker history for the origin node and the destination node to a path marker history for the destination node unless the exploration rules are violated in which case it is recorded to a path marker registry for the origin node that the link between the origin node and the destination node is blocked;

(d) program instructions for following any links from the origin node that have not been followed;

(e) program instructions to delete the origin node path marker and renaming the destination node to the origin node if there is nothing recorded in the origin node path marker registry;

(f) program instructions repeating steps (a)-(e) until reaching the end of the finite state automaton.

16. A method for generating new sound sequences based on input sounds, the method comprising:

receiving a plurality of input sound sequences containing sound frequencies with corresponding time duration;

converting the plurality of input sound sequences to a finite state automaton using a system that allows over-generation;

traversing the finite state automaton with a graph exploration procedure that uses exploration rules and a plurality of path markers to determine paths across the finite state automaton; and

storing the paths across the finite state automaton to a path marker data structure to define recorded path markers;

wherein the recorded path markers that are not found in the plurality of input sound sequences define a new sound sequence.

17. A method for generating new sound sequences based on input sounds as recited in claim 16, where over-generation enables sharing of nodes of the finite state automaton.

18. A method for generating new sound sequences base on input sounds as recited in claim 16, wherein the exploration rules constrain how the finite state automaton is be traversed.

19. A method for generating new sound sequences based on input sounds as recited in claim 16, wherein the path marker data structure indexes the plurality of path markers, where each path marker contains a path marker history and a path marker registry.

20. A method for generating new sound sequences based on input sounds as recited in claim 16, wherein the path marker history and the path marker registry of particular path markers are updated when traversing the finite state automaton.