An information processing device includes: an input interface; and at least one processor, configured to perform the following: selecting an instrument, a musical tone of which is to be digitally synthesized based on corresponding musical tone data, via the input interface; acquiring a parameter value that has been set for the selected instrument; generating a random number based on a random function; and changing a pitch of the musical tone of the selected instrument based on the generated random number and the acquired parameter values.

Music data memory includes pieces of music within a group and other pieces of music outside the group. The next piece to be played is automatically determined by random table among pieces within the group. Favorite or newest piece is weighted to be more frequently played in the group. Piece in music data memory is automatically included into the group by random table. Newly downloaded piece into music data memory is included into the group by priority. Most frequently played piece is excluded from the group in place of newly included piece. Favorite or newest piece may be an exception of exclusion. Next piece is capable of being played in tempo similar to that of preceding piece by means of tempo-adjusted or piece replacement or repetition of the same piece for the purpose of continued baby cradling in synchronism with the same tempo of succeeding pieces.

A sound signal synthesis method includes generating first data representing a deterministic component of a sound signal based on second control data representing conditions of the sound signal, generating, using a first generation model, second data representing a stochastic component of the sound signal based on the first data and first control data representing conditions of the sound signal, and combining the deterministic component represented by the first data and the stochastic component represented by the second data and thereby generating the sound signal.

Techniques are disclosed relating to determining composition rules, based on existing music content, to automatically generate new music content. In some embodiments, a computer system accesses a set of music content and generates a set of composition rules based on analyzing combinations of multiple loops in the set of music content. In some embodiments, the system generates new music content by selecting loops from a set of loops and combining selected ones of the loops such that multiple ones of the loops overlap in time. In some embodiments, the selecting and combining loops is performed based on the set of composition rules and attributes of loops in the set of loops.

According to this disclosure, a method of synthesizing an audio stream sample using a processor is provided. The method comprises: generating a set of unnormalized log probabilities using a neural network, each unnormalized log probability associated with a possible value for the audio stream sample, sampling a Gumbel distribution for each of the unnormalized log probabilities, adding the samples from the Gumbel distribution to each of the respective unnormalized log probabilities to generate a set of modified log probabilities, each modified log probability associated with a possible value for the audio stream sample, and selecting the possible value of the audio stream sample associated with the largest modified log probability from the set of modified log probabilities as the audio stream sample.

Techniques are disclosed relating to determining composition rules, based on existing music content, to automatically generate new music content. In some embodiments, a computer system accesses a set of music content and generates a set of composition rules based on analyzing combinations of multiple loops in the set of music content. In some embodiments, the system generates new music content by selecting loops from a set of loops and combining selected ones of the loops such that multiple ones of the loops overlap in time. In some embodiments, the selecting and combining loops is performed based on the set of composition rules and attributes of loops in the set of loops.

A method of creating AI-composed music having a style that reflects and/or augments the personal style of a user includes selecting and/or composing one or more seed compositions; applying variation and/or mash-up methods to the seed compositions to create training data; training an AI using the training data; and causing the AI to compose novel musical compositions. The variation methods can include methods described in the inventor's previous patents. A novel method of creating mash-ups is described herein. Variation and mash-up methods can further be applied to the A compositions. The AI compositions, and/or variations and/or mash-ups thereof, can be added to the training data for re-training of the AI. The disclosed mash-up method includes parsing the seed compositions into sequences of elements, which can be of equal length, beat-matching the seed compositions to make corresponding elements of equal beat length, and combining the elements to form a mash-up.

Techniques are disclosed relating to determining composition rules, based on existing music content, to automatically generate new music content. In some embodiments, a computer system accesses a set of music content and generates a set of composition rules based on analyzing combinations of multiple loops in the set of music content. In some embodiments, the system generates new music content by selecting loops from a set of loops and combining selected ones of the loops such that multiple ones of the loops overlap in time. In some embodiments, the selecting and combining loops is performed based on the set of composition rules and attributes of loops in the set of loops.

An audio analysis method includes calculating, from an audio signal, a sound generation probability distribution which is a distribution of probabilities that sound representing the audio signal is generated at each position in a music piece, estimating, from the sound generation probability distribution, a sound generation position of the sound in the music piece, and calculating, from the sound generation probability distribution, an index of validity of the sound generation probability distribution.

Techniques are disclosed relating to determining composition rules, based on existing music content, to automatically generate new music content. In some embodiments, a computer system accesses a set of music content and generates a set of composition rules based on analyzing combinations of multiple loops in the set of music content. In some embodiments, the system generates new music content by selecting loops from a set of loops and combining selected ones of the loops such that multiple ones of the loops overlap in time. In some embodiments, the selecting and combining loops is performed based on the set of composition rules and attributes of loops in the set of loops.