Method and device for correcting pitch and timbre of audio based on resampling acceleration calculation

Abstract

The invention discloses a method for calculating and correcting audio content pitch and timbre based on resampling acceleration, which comprises the following steps: acquiring a fundamental frequency sequence of an audio, and acquiring an original pitch sequence of the audio based on the fundamental frequency sequence; constructing a fundamental frequency sequence array needing to be adjusted based on the obtained pitch sequence; establishing a fundamental frequency array mapping relation array based on the audio total duration; calculating a time step length and transmitting the time step length into an audio sequence by a resampling acceleration calculation method based on a comparison result; and a step of obtaining the audio and correcting the audio to obtain a corrected audio. The invention also discloses a corresponding device for correcting the pitch of the audio.

Description

technical field [0001] The present invention relates to the field of audio signal processing, and in particular, to a method and device for correcting the pitch and timbre of audio content based on resampling. Background technique [0002] As early as 1998, Harold Hildebrand, an electrical engineer from Exxon Mobil Corporation in the United States, invented a technical patent for automatic pitch correction, and handed over the patent to Antares Audio Technologies. Auto-tune. Since then, it has been used in the recording industry and modern music culture. The effect presented by this patent not only changes the pitch information of the audio, but also renders the timbre to a certain extent, forming the classic "electronic timbre of the recording industry". ". [0003] With the advancement of technology, using the autocorrelation function to extract the fundamental frequency is no longer the best technical solution for extracting the fundamental frequency, and it is insuffic...

AI Technical Summary

Problems solved by technology

[0003] With the advancement of technology, the method of using autocorrelation function to extract fundamental frequency is no longer the best technical solution for extracting fundamental frequency, and there are deficiencies in accuracy and extraction speed; and Auto-Tune products have undergone 20 years of technical precipitation, but so far There are no products suitable for mobile terminals, and it is difficult to meet the needs of mobile terminal timbre correction and timbre change

In the mobile Internet era, domestic mobile Internet companies have not achieved this type of timbre effect. At this stage, domestic mobile Internet companies have two technical alternatives: 1) Realize pitch correction or electronic sound effects based on the traditional algorithm Psola, but due to Psola The limitations of the underlying technology of the algorithm cannot essentially solve the problem of audio frame jitter, and there are some differences with the classic "electronic sound" of the record industry; 2) Use deep learning to correct pitch, but this type of technology presents sound There are more timbre differences between the effect and the Auto-tune series products, and it is not the classic "electronic timbre" of the record industry

Images