I know this has been asked before, but many of the descriptions and other posters replies seem to conterdict one another.

What I have read, this is Macroblocking;
http://archive2.avsforum.com/avs-vb/attachment.php?attachmentid=45318

This is False contouring;
http://www.avsforum.com/avs-vb/attachment.php?attachmentid=93565&d=1194034482

...which can also be know as 'banding' or "posterization";
http://www.cambridgeincolour.com/tutorials/posterization.htm

All of which is due to lower bit rates and higher compressions.
Is all of this correct?

From my amateur's perspective, the descriptions look okay, however, I'm not sure if lower bit rates/higher compression also leads to decreased "bit depth" in the chroma and luma channels. I believe it is insufficient bit depth in the chroma and luma channels that causes false contouring (steps are too coarse).

I believe it is insufficient bit depth in the chroma and luma channels that causes false contouring (steps are too coarse). Ok, I'll buy that. Anyone else??

I didn't read all the explanations fully but in general they seem accurate.

The thing to remember is that macro-blocking is almost always introduced at the encoding of the source. It's simply a function of not having enough bandwidth to handle the signal.

Posterization and false contouring are usually introduced by the signal chain of your display not having enough bit depth. Essentially what happens is that rounding error of each stage of the signal processing accumulates until you can see it. High bit depth processing solutions don't exhibit this.

These are not hard and fast rules though. You can introduce posterization at the encode stage and similarly your post processing can introduce macroblocking depending on what you are trying to do.

In short though, they are all artifacts of insufficient bit depth.

Thanks for the replies. I understand why these happen, I was mostly concerned about the correct 'term' to use for each type of issue.

Are the images and the terms correct? There appears to be more that one term for each artifact which only confuses things more.

Aside from macro blocking, the others are used pretty much interchangeably. I wouldn't sweat it.

Macroblocks are the square compression artifacts that can be seen occassionally (hopefully rarely, but on some content a lot). These usually occur where there isn't enough bitrate to keep up with the entropy of the content.

Contouring/banding occurs when there is not enough bitdepth, and smooth gradients no longer appear smooth but appear to make distinct changes in color because the step-size is now too large.

Aside from macro blocking, the others are used pretty much interchangeably. I wouldn't sweat it. It's just that these are getting interchanged between one another and making things more confusing then they already are.

Chris; I pretty much understand this, it is just trying to find and use the correct term.

The consensus appears to be in your favor. What I was specifically commenting on was "All of which is due to lower bit rates and higher compressions." in your OP. I believe false contouring is accurately illustrated, but it's caused by insufficient bit depth, not insufficient bit rate as a result of over-compression. Whether the latter can also cause the former, i don't know...

The consensus appears to be in your favor. What I was specifically commenting on was "All of which is due to lower bit rates and higher compressions." in your OP. I believe false contouring is accurately illustrated, but it's caused by insufficient bit depth, not insufficient bit rate as a result of over-compression. Whether the latter can also cause the former, i don't know...

That is correct. And yes, insufficient bitrate can also cause banding problems, depending on how the content was encoded. You could say still that in those cases the immediate cause is insufficient effective bit depth, but the source of that problem is insufficient bitrate as things get smashed to one level rather than a larger range of levels that are possible, but require more information to describe. So there, the levels are there, the bitdepth is there to avoid banding, it's just not being utilized because it's all been compressed down so much.

The consensus appears to be in your favor. Golly. :eek:

Ok, but how do you know where the problem is? If it is on your end or 'theirs', since this can happen almost anywhere along the pipeline? :confused:

I still don't really get why a component feed would 'hide' or "smooth out" any of this where as a HDMI shows it up from a DVD player concerning that macro enhanced bug from that 2nd link I provided.

Welcome to the world of transport--things can happen anywhere along the way, just like in computer networks, phone systems, cable TV, shipping...

There's an natural integration/filtering function happening when going from the digital to analog domain, perhaps this "smoothing" contributes to what you perceive with component? It also quite likely that it is implementation dependent...

Welcome to the world of transport Is that anything like "Beam me up Scotty"?

Yeah--"no intelligent life forms detected...commence project Genesis on my mark" :D