Interesting article about 4:2:2 and 4:1:1 sampling

[gshelley61]

Here's a technical paper written a few years ago when DV recording formats were just starting to replace BetaSP as a broadcast video standard. The author goes into some detail about analog video resolution/bandwidth and digital video, and then outlines the real-world qualitative differences between 4:1:1, 4:2:2 and 4:2:0 color sampling...

http://www.chumpchange.com/parkplace/Video/DVPapers/dv-beta.htm

It might be worth mentioning the large multimedia company I work for uses Panasonic DVC equipment for news, advertising and promotional production - that format uses 4:1:1 sampling.

[Capmaster]

Thanks for the tip. Interesting stuff. My TBC has "4:2:2" labeled on the front and after having it for two years I never really understood what that meant. Now it makes sense: the luma/chroma sampling scheme. Not that I can use that information ....it's just nice to not have mysteries

[gshelley61]

Thanks. I guess there is some concern about 4:1:1 sampling being inferior to 4:2:2 (which theoretically it is), but the actual visible quality difference is apparently quite small. The writer points out that ultimately it is high quality lenses, large CCD's and high quality video circuitry that has the most effect on image quality.

[lordsmurf]

Yes, DV is not a wholly bad format, especially not for shooting new video.

But when you take 4:2:2 analog source, especially ones that are already lossy like VHS or broadcast, and then reduce it down to 4:1:1, it's not too hard to see the issues it has with handling colors (especially red and green). It's virtually impossible to restore video footage, as color errors are further enhanced by the 4:1:1, making restoration twice as hard is not impossible. For the best video correction, you must stay in 4:2:2 while processing 4:2:2 source. That article suggests the same things I've seen, with brightness being pixellated from time to time in 4:1:1. It also suggest something new, that some of this is related to the codecs ... which now makes me wonder about the quality of the Canopus ADVC line. It may not only suffer from DV spec, but from something in it's hardware codecs.

DV is for original shooting, not a conversion format to DVD. Squeezing 4:2:2 down to 4:1:1 and then down to 4:2:0 is like giving your video a double beating.

People love their DV conversions, and may disagree, but some of us have a keen eye and for these reasons, DV is just a middleman while we wait for the better stuff to come down in price and get more mainstream (4:2:2 or 4:2:0 shooting). If you insist of shooting consumer DV, get one with multi CCD.

The good lens theory also holds true. I've got a $4k Nikon D1 still camera. I use a $2k Nikkor lens on it. If I stuck a $100 Quantary or Sigma on it, I may as well shoot with a point-n-shoot or low-dollar SLR instead. By not having good glass, I'm already losing out.