Research Of Signal Distortion Compensation Algorithm In Audio/Video Device's Assessment System

With the development of modern technology, many kinds of digital video equipments are designed to be used in different fields. And promoted by people's increasing demands, the quality requirements for broadcast television system become higher. Consequently, a precise, reliable, convenient and fast video measuring system is necessity for video equipments'development, compatibility and certification test, quality evaluation, auto-product test and open-loop video system test.According to the international and national standards, this paper develops and realizes a remotely controlled auto audio and video measuring system, which can comprehensively evaluate the component video signals transmitted by the equipment under test.The system is composed of a test signal generator and an assessment terminal. To make the system compatible and convenient, video signals are sampled by high definition capture device at the evaluation terminal. In order to improve the system's precision and stability, the sample principle of high definition device is studied while the quantization error in sampling process and the distortions caused by sampling format are carefully analyzed. Ramp signal, dither signal and mean method are proposed to remove the quantization error. And for precision decline caused by sampling operation, a uniform approximation FIR filter is employed to realize the ideal interpolation. What's more, distortions which are caused by unsatisfactory system feature, such as overshoot, undershoot, band limit and group delay are well analyzed. Correspondingly, filter/equalizer methods are proposed to correct the distortions. For example, frequency-domain equalization is employed to remove overshoot or undershoot distortion; window design method with phase equalizer are applied to reduce distortion caused by band limit; IIR all-pass filter is used to correct the phase so as to eliminate the group delay distortion. Precision of the measuring result is greatly improved after compensation, and its stability and reliability also perform better.