Metrical grid inference for free rhythm musical input
View Patent ↗Computer-based methods infer a metrical grid from music that has been input without a predetermined time signature or tempo, enabling such free rhythm input to be annotated with the inferred grid, and stored and transcribed as a musical score. The methods use Bayesian modeling techniques, in which an optimal metrical grid is inferred by identifying the metrical grid that best explains the given sequence of notes by maximizing the posterior probability that it represents the note sequence. Prior musical input from a given user as well as explicit information about the musical style of the input may be used to improve the accuracy of the transcription.
1. A method for free rhythm musical input, the method comprising:
receiving musical input comprising a sequence of notes;
determining an optimal metrical grid corresponding to the sequence of notes, wherein the optimal metrical grid is defined as the metrical grid that optimally represents the sequence of notes, and wherein determining the optimal metrical grid is based in part on an a priori probability of the occurrence of the optimal metrical grid, wherein the a priori probability is independent of the sequence of notes;
annotating the musical input with the optimal metrical grid; and
storing a representation of the annotated musical input.
2. The method of claim 1 , wherein the metrical grid comprises a time signature and a tempo.
3. The method of claim 2 , further comprising generating a musical score representing the sequence of notes using the time signature and tempo.
4. The method of claim 2 , wherein the optimal metrical grid comprises a plurality of tempos including the first-mentioned tempo.
5. The method of claim 4 , wherein each of the plurality of tempos starts at the beginning a measure based on the time signature.
6. The method of claim 4 , wherein one or more of the plurality of tempos starts at a location other than the beginning of a measure based on the time signature.
7. The method of claim 4 , wherein an priori probability of a given one of the plurality of tempos representing a portion of the musical input is based in part on a characteristic of a portion of the musical input received prior to receiving the given portion.
8. The method of claim 4 , wherein the optimal metrical grid comprises a plurality of time signatures, including the first-mentioned time signature.
9. The method of claim 1 further comprising determining a start location for the optimal metrical grid.
10. The method of claim 1 , further comprising modifying the musical input by adjusting a temporal location of at least one of the sequence of notes to more closely conform to the optimal metrical grid.
11. The method of claim 1 , wherein the sequence of notes is represented as a sequence of MIDI events.
12. The method of claim 1 , wherein the musical input is received as audio input, and the sequence of notes is determined by analyzing the audio input.
13. The method of claim 1 , wherein determining the optimal metrical grid is involves Bayesian modeling of the probabilities of a range of possible metrical grids for representing the sequence of notes.
14. The method of claim 1 , wherein the a priori probability of the occurrence of the metrical grid is based in part on information about a musical style of the musical input.
15. The method of claim 1 , wherein the musical input is received from a given user, and the a priori probability of the occurrence of the metrical grid is based in part on a musical style of the given user.
16. A computer program product, comprising:
a non-transitory computer-readable medium with computer program instructions encoded thereon, wherein the computer program instructions, when processed by a computer, instruct the computer to perform a method for free rhythm musical input, the method comprising:
receiving musical input comprising a sequence of notes;
determining an optimal metrical grid corresponding to the sequence of notes, wherein the optimal metrical grid is defined as the metrical grid that optimally represents the sequence of notes, and wherein determining the optimal metrical grid is based in part on an a priori probability of the occurrence of the optimal metrical grid, wherein the a priori probability is independent of the sequence of notes;
annotating the musical input with the optimal metrical grid; and
storing a representation of the annotated musical input.
17. A system for free time musical input, the system comprising:
an input for receiving music data comprising a sequence of notes;
a memory for storing computer-readable instructions; and
a processor connected to the memory, wherein the processor, when executing the computer-readable instructions, causes the free time musical input system to:
determine an optimal metrical grid corresponding to the sequence of notes, wherein the optimal metrical grid is defined as the metrical grid that optimally represents the sequence of notes, and wherein determining the optimal metrical grid is based in part on an a priori probability of the occurrence of the optimal metrical grid, wherein the a priori probability is independent of the sequence of notes;
annotate the musical input with the optimal metrical grid; and
store a representation of the annotated musical input.