Humanizing music sequence

A digital audio workstation comprising at least one processing core and memory comprising computer program code that when executed by the at least one processing core cause:

• obtaining, by the digital audio workstation, sparse models of human-played music for humanizing a rhythmic structure of a music sequence (S), said sparse models comprising information indicating timing (t′1, . . . , t′n) of sounds (s1, . . . , sn) of a humanized music sequence (S′);
• selecting, by the digital audio workstation, a model from the obtained sparse models based on at least one of the following:
• generating for each of the obtained sparse models a score;
• and selecting the model that has the highest score;
• or
• characterizing said obtained sparse models on a user interface of the digital audio workstation, and receiving user input on the user interface indicating the selected model;
• and further cause:
• adjusting, by the digital audio workstation, based on the selected model, timing information (ti) of the rhythmic structure of the music sequence (S) to comprise an offset (oi);
• and
• outputting, by the digital audio workstation, the music sequence (S) comprising one or more adjusted times (ti+oi).

A non-transitory computer-readable medium storing instructions that, when executed by at least one processor of a digital audio workstation, cause the digital audio workstation to perform a method comprising:

• obtaining, by the digital audio workstation, sparse models of human-played music for humanizing a rhythmic structure of a music sequence (S), said sparse models comprising information indicating timing (t′1, . . . , t′n) of sounds (s1, . . . , sn) of a humanized music sequence (S′);
• selecting, by the digital audio workstation, a model from the obtained sparse models based on at least one of the following:
• generating for each of the obtained sparse models a score;
• and selecting the model that has the highest score;
• or
• characterizing said obtained sparse models on a user interface of the digital audio workstation, and receiving user input on the user interface indicating the selected model;
• said method further comprising:
• adjusting, by the digital audio workstation, based on the selected model, timing information (ti) of the rhythmic structure of the music sequence (S) to comprise an offset (oi);
• and
• outputting, by the digital audio workstation, the music sequence (S) comprising one or more adjusted times (ti+oi).