Audio signal transforming by utilizing a computational cost function
First Claim
1. A method of audio signal transforming, the method comprising the steps of:
- receiving a sequence of time domain multiple digital audio samples representing sound;
computer-implemented processing the time domain multiple digital audio samples to derive a corresponding sequence of multiple audio pulses in the time domain, each said audio pulse being associated with a characteristic frequency of a single pulse corresponding to a pitch associated with the audio pulse, each said audio pulse having a length corresponding to one pitch period of the respective audio pulse;
generating for each pulse a sequence of repetitions of said audio pulse, said audio pulse being repeated according to its own characteristic frequency;
deriving frequency domain information associated with at least some of the sequences of repetitions of said audio pulses, each said sequences of repetitions of said audio pulse being represented as a vector of sinusoids based on the derived frequency, said vector of sinusoids corresponds to a sinusoidal series expansion of the specific audio pulse;
computer-implemented transforming the sound represented by the time domain digital audio samples by processing the audio pulses using the frequency domain information;
overlapping the audio pulses in a pitch-synchronous manner;
transforming each input pulse according to transformation parameters including timbre modifications and estimating a timbre envelope defined by harmonics;
transforming the timbre envelope using a mapping function; and
obtaining the transformed timbre envelope at each frequency by computing values of the input timbre envelope at the mapped frequency;
wherein in the speech transformation an overlapping window is used, a computational cost function which depends on a product of the number of pitch periods and the inverse of the minimum fundamental frequency within the window is determined.
1 Assignment
0 Petitions
Accused Products
Abstract
A sequence is received of time domain digital audio samples representing sound (e.g., a sound generated by a human voice or a musical instrument). The time domain digital audio samples are processed to derive a corresponding sequence of audio pulses in the time domain. Each of the audio pulses is associated with a characteristic frequency. Frequency domain information is derived about each of at least some of the audio pulses. The sound represented by the time domain digital audio samples is transformed by processing the audio pulses using the frequency domain information. The sound transformation utilizes overlapping windows and a computational cost function which depends on a product of the number of the pitch periods and the inverse of the minimum fundamental frequency within the window is determined.
-
Citations
25 Claims
-
1. A method of audio signal transforming, the method comprising the steps of:
-
receiving a sequence of time domain multiple digital audio samples representing sound; computer-implemented processing the time domain multiple digital audio samples to derive a corresponding sequence of multiple audio pulses in the time domain, each said audio pulse being associated with a characteristic frequency of a single pulse corresponding to a pitch associated with the audio pulse, each said audio pulse having a length corresponding to one pitch period of the respective audio pulse; generating for each pulse a sequence of repetitions of said audio pulse, said audio pulse being repeated according to its own characteristic frequency; deriving frequency domain information associated with at least some of the sequences of repetitions of said audio pulses, each said sequences of repetitions of said audio pulse being represented as a vector of sinusoids based on the derived frequency, said vector of sinusoids corresponds to a sinusoidal series expansion of the specific audio pulse; computer-implemented transforming the sound represented by the time domain digital audio samples by processing the audio pulses using the frequency domain information; overlapping the audio pulses in a pitch-synchronous manner; transforming each input pulse according to transformation parameters including timbre modifications and estimating a timbre envelope defined by harmonics; transforming the timbre envelope using a mapping function; and obtaining the transformed timbre envelope at each frequency by computing values of the input timbre envelope at the mapped frequency; wherein in the speech transformation an overlapping window is used, a computational cost function which depends on a product of the number of pitch periods and the inverse of the minimum fundamental frequency within the window is determined. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20)
-
-
21. A computer program product tangibly embodied in one or more non-transitory machine-readable media for audio signal transforming, the computer program product comprising instructions that are executable by one or more processing devices to:
-
receive a sequence of time domain digital audio samples representing sound; process the time domain digital audio samples to derive a corresponding sequence of audio pulses in the time domain, each of the audio pulses being associated with a characteristic frequency; derive frequency domain information about each of at least some of the audio pulses; and transform the sound represented by the time domain digital audio samples by processing the audio pulses using the frequency domain information; wherein in the sound transformation an overlapping window is used, a computational cost function which depends on a product of the number of pitch periods and the inverse of the minimum fundamental frequency within the window is determined. - View Dependent Claims (22, 23, 24)
-
-
25. A system for audio signal transforming, the system comprising:
-
a receiver to receive a sequence of time domain digital audio samples representing sound; and a processor to (a) process the time domain digital audio samples to derive a corresponding sequence of audio pulses in the time domain, each of the audio pulses being associated with a characteristic frequency; (b) derive frequency domain information about each of at least some of the audio pulses; and (c) transform the sound represented by the time domain digital audio samples by processing the audio pulses using the frequency domain information; wherein in the sound transformation an overlapping window is used, a computational cost function which depends on a product of the number of pitch periods and the inverse of the minimum fundamental frequency within the window is determined.
-
Specification