The Research And DSP Implementation Of Multi-channel Wide Dynamic Range Compression Algorithm

Crew is a special group who chronically work and live under the environment of the ship noise.With the increasing of contacting noise time and noise level,hearing loss will gradually affect language frequency,and thus cause noise deafness which may increase the probability of crew with noise hearing loss.The hearing aids are used to solve the inconvenience of general patients in their life while helping the crews who suffer noise deafness.At present,the digital hearing aids have a great deal of consumer groups,but unfortunately the probability of Domestically-made algorithm is quite low.This paper studies the wide dynamic range compression algorithm of digital hearing aids,wide dynamic range algorithm can be used in hearing protection and compensation when the crew sail in the environment of the marine noise,it has an important application value on improve crew’s hearing loss.The paper introduces the principle of multi-channel wide dynamic range compression algorithm,which has a programmable input/output port within each channel,can satisfy the hearing disorder to achieve corresponding gain compensation requirements at different frequencies.Additionally,in order to make the experience index of hearing aids system’s maximum delay no more than 10 ms.The paper proposes a real-time WDRC algorithm that based on time-varying filter.In accordance with the property of product by frequency domain is equivalent to convolution by time domain in digital signal processing theory,the algorithm utilizes frequency sampling method that can take gain value as the frequency response of the FIR filter to design the corresponding filter,and then realizing the signal processing in time domain.The experimental results show that the scheme does not produce the distortion such as excessive signal harmonic,simultaneously,the time delay of algorithm and system theory time delay are reduced by 50% and 33.33% respectively,which greatly improving the working performance of the system.This paper also analyzes the influence to the compressed signal in different initiating and relieving time parameters.To solve this problem,this paper puts forward to a multi-channel wide dynamic range compression algorithm of adaptive time parameter.By simulating the experiment and listening the test to the adaptive time parameters of multi-channel wide dynamic range compression algorithm,the results show that this improved algorithm can effectively improve the words comprehension of hearing aids.This article implements the real-time algorithm on the platform of On Semiconductor company’s Ezairo 7100 DSP.It makes the running efficiency of the whole algorithm increasedby 56% approximately through the quantitative treatment to gain formula in the process of operation.Finally,we use the professional equipment and test software,which is Frye FONIX8000 to test the hearing aids of this algorithm.It shows that distortion degree(top right)is within 1%(distortion rate is respectively 0.3%,0.4%,0.1%)at the characteristic frequency(500 Hz and 800 Hz and 1600 Hz).It is obvious that the result can reach the desired index and demonstrate this algorithm having certain actual application value.