A Method of Feedback Delay Network Timbre Optimization

Abstract

The invention relates to a feedback delay network timbre optimization method. The delay coefficients of the delay line of a feedback delay network are selected by the following steps: selecting and determining multiple delay coefficient combinations according to preset selection conditions; calculating multiple spectrum flatness values respectively corresponding to the multiple delay coefficient combinations; and comparing the multiple spectrum flatness values, and taking the delay coefficient combination corresponding to the maximum spectrum flatness value as the delay coefficient combination of the delay line of the feedback delay network. According to the invention, the multiple spectrum flatness values corresponding to the multiple delay coefficient combinations are compared, and the delay coefficient combination corresponding to the maximum spectrum flatness value is taken as the delay coefficient combination of the delay line of the feedback delay network, which makes the timbre of output signals highly natural.

Description

technical field [0001] The invention belongs to the technical field of audio signal processing, and in particular relates to a method for optimizing timbre of a feedback delay network. Background technique [0002] Rendering of audio signals based on headphones and speakers is often used to simulate room reverberation. The earphone-based rendering includes the same method as the speaker and the earphone-specific dual-channel virtual surround sound method. Its application scenarios are such as in a quiet room, through the new signal processing algorithm, the audio signal processed by reverberation can be played through the speaker, which has achieved the effect of simulating theaters and caves. Such audio signal processing can be performed by convolving the original signal with the reverberation of the desired room. However, in practical applications, since the room reverberation (that is, the impulse response) is usually a very high-order FIR filter, the algorithm complexi...

AI Technical Summary

Problems solved by technology

Poor choice of M will lead to distorted timbre of the final result. Although some people have proposed some selection conditions for M, after selecting M according to this condition, in practical applications, it is found that even if the above requirements are met, the final output audio signal will still have a certain timbre. Distortion, seriously affecting the sense of hearing

Images