Design And Realization Of Karaoke Audio Effect Algorithm Based On DSP

With the function of improving the timbre and the output sound quality, the audio effects processing system has been widely used in the professional areas such as radio and TV station. But by the constraints of economic cost and processing capacity, the audio effects processing system used in Karaoke at present exhibits the general problems of poor output effect and poor user experience. Therefore, it is of great economic value and social significance to design the economic and practical Karaoke audio effects processing system.First of all, this thesis presents the design requirements about sound field extension, equipment protection, gain range expansion and audio equalization from the points of improving the output space sense, suppressing howling, expanding the scope of system gain and enhancing the adjustment ability in spectrum of the system. Then, this thesis analyses and designs the correlation algorithms based on the related theory of the Karaoke audio effects to enhance the output audio effects and improve the user experience. At last, the algorithms are transplanted to the general digital signal processing hardware provided by the cooperative company, and then, the thesis completes a set of economic and efficient Karaoke audio effects processing system.The majority work in this thesis is listed as follows:(1)In terms of sound field extension, from the point of extending the echo delay time and enhancing the fidelity of reverberation, this thesis solves the problem of insufficient sense of space. By designing the algorithm of multi-delay fusion, allocating echo delay in equal interval and reducing the sampling rate in echo, this thesis improves the utilization of the data buffer and extends the echo delay. By using the source image method to design the delay of comb filter, and by proposing constraints of all-pass filter to optimize the parameter in late reverberation, the reverberation effect is enhanced realistically.(2) In terms of equipment protection, this thesis improves howling suppression algorithm for howling. The proposed combination algorithm based on phase shift and frequency shift reduces the calculation cost and maintains the howling suppression ability compared to conventional amplitude suppression scheme. For the limited capacity of amplification for the input signal which aims to protect the equipment of the output, this thesis designs the dynamic gain control algorithm to assign smaller gain values to larger input signals, which improves the overall gain range of the system.(3)In terms of audio equalization, the lowpass/highpass filter whose order and cut-off frequency can be adjusted and the lifting / attenuation filter whose center frequency, bandwidth and center frequency gain are adjustable are designed to increase the adjustment ability in spectrum of the system.(4)For the realization of algorithms described above, the conversion of data types are designed to reduce the demand for the data buffer. At the same time, the addition of temporary variables and the expansion of nested functions are used to take advantage of the pipelining technology in DSP, which improves the processing speed.The thesis tests the effect of the output, and the results show that with the enhanced spatial sense, the suppression of howling, the extended output gain range, and the increased adjustment adaptability to spectrum in the system, all the Karaoke effects are remarkable. For obvious output effect and the improved user experience, the Karaoke audio effect processing system have been approved by the professional tuners from the cooperative company, which reaches the anticipated objectives.