Would be my pleasure
.
The process usually starts with examining what shape the "master" is in. Believe it or not, even recent masters tend to have a lot of quality issues. Repair can be anything from re-digitizing to clean up/noise reduction. The master is then color corrected to comply with HD video standar (BT709).
The master is usually in higher resolution than Blu-ray. The video sample resolution can be 10 bits instead of 8. And video sampling rate 4:2:2 rather than 4:2:0. Downsamplign requires care: if you just truncate, you get banding as values jump from one to the other. A little example. Let's say we have a number system where you have two digits of resolution. Let's see what happens when we truncate to go to one digit of resolution:
Code:
Input value: 10 11 12 13 14 15 16 17 18 19 20
Output value: 1 1 1 1 1 2 2 2 2 2 2
You see what happened? Before conversion, we had a smooth ramp. After conversion, we have one value for a while, then all of a sudden jump to another value and stay there. The eye catches the sudden transition and you see that as banding edges/contouring. So contrary to popular belief, unless you know what you are doing, higher resolution mastering can be a bad thing, not good.
The right solution calls for adding noise in the form of "dithering." That noise blurs the boundaries so the eye can't detect the edge. Unfortunately noise is harder to compress but is still better than not doing it.
The next step is compression. Here the process is as follows: the video is automatically encoded in two passes. In the first pass, the encoder scans the movie looking for difficult and easy parts. It then starts to encode the movie, saving bits in the easy parts and applying them to the hard parts. This works 80 to 90% of the time. But in other times, the encoder misjudges what is easy and assigns too few bits to it. An operator then, sitting in front of a workstation, tweaks the data rate and other encoder parameters as available to dial out artifacts.
As you can imagine, labor is expensive for a skilled compressionist and schedules always tight. So mistakes and shortcuts can occur in above. When that happens, you get blocking artifacts. The codec divides the screen into squares and compresses them separately. If it overcompresses a block, the edges show up. And since there are four edges, they look like artificial blocks and hence the name.
Another manifestation especially in some codecs like MPEG-4 is over softening of the video. In order to mitigate the above artifact, the codec attempts to soften the block edges and in doing so, the video resolution is reduced for that region.
In extreme cases, you get both of the above artifacts.