Two more comments on dither:
(1) Previously when I mentioned the below:
Btw, what is the measure of "effectivity" in this regard?
[a href=\"index.php?act=findpost&pid=206314\"][{POST_SNAPBACK}][/a]
Measured as the trade off between the output variance in intensity from the mean value of the output in a portion of the picture which was of constant intensity at the input averaged over all inputs, and the deviation of mean output from the input.
[a href=\"index.php?act=findpost&pid=206339\"][{POST_SNAPBACK}][/a]
The sum of the "variance" quantity (sometimes called
apparent noise) and the "deviation" quantity (sometimes called
tonal quality) defines the reconstruction error variance. Degradation represented by tonal quality is perceptually more annoying, and hence, a good quantization scheme should seek to decrease tonal quality degradation, at the expense of apparent noise, such that the total reconstruction error is minimized.
And, adding dither signal before quantization helps us to accomplish the above, i.e., lowering tonal quality degradation, for a more pleasing image.
(2) The dither signal does not always have to be random. It can be
deterministic under certain situations, and in those circumstances may actually be better than random dither.
