What do you call what's shown in the image below? It happens in low light situations where the bitrate is too low. Banding? Artifacts? Something else?
Try StreamFab Downloader and download from Netflix, Amazon, Youtube! Or Try DVDFab and copy Blu-rays! or rip iTunes movies!
+ Reply to Thread
Results 1 to 4 of 4
Thread
-
-
Well, it's harder to see when it actually IS dark, but bumping it up in brightness makes it OBVIOUS that it is banding (aka too few quantization levels). Note that my bumping it up (below) was done in 16bit to alleviate the loss in range (so the great majority of that banding already was there prior to bumping up).
It IS an artifact: An artifact just means an unintended detrimental feature addition that is a byproduct of the process. Banding is one type of artifact. Banding occurs when QUANTIZATION levels are too low (this happens often when the bitrate is too low, however, and when the dynamic range is too large for the step size - often occurring in the low light level portions more than the middle or high light level portions).
There are also DCT blocking artifacts in that image.
ScottLast edited by Cornucopia; 25th Dec 2014 at 03:24.
-
OK, great. Thanks a lot for your help.
I'm afraid I don't know what "QUANTIZATION levels" and "dynamic range is too large for the step size" means but that's OK. I just needed to know the proper term for this phenomenon for right now. -
If you call Low Light "0" and High Light "100", do you have 2 intermediate steps?, 10 steps, 20? 50? 100? 1000? The more steps you have in that same range, the finer & smoother the gradient is and less banding.
If you call Low "0" but change High to "1000" (because you can now show an even brighter brightness), even if you used to have 100 steps and thought that was fine enough, now those 100 steps have to cover a range 10x the size, so those steps are once again too few (like the 10-step size in the first example). The greater dynamic range you have, the more steps you will need to keep the step size fine enough and avoid banding.
This all assumes light levels are linearly captured, stored & viewed (and interpreted by our brains). That is not the case. It has been shown that humans need finer steps in the dark areas than in the light areas.
Scott