Adaptive dithering of one dimensional signals

This research work presents new techniques on the dithering of one dimensional signals. The first part of this work was to determine whether dither is needed when reducing the bit depth. This was done with time series hypothesis tests in time and frequency domains, and they evaluate whether the total error of the undithered quantization is white noise. If this is the case, then no undesired harmonics are added in the quantization or re-quantization process. The second part of this work was to develop a state of the art dither which is signal dependent and attempts to have lower quantization noise than the classic techniques. Experiments showing the effectiveness of the methods with both synthetic and real audio signals are presented.