Kevin Crockett
July 5th, 2006, 11:02 AM
I know that if you shoot using a format that samples at 4:2:0 that you're getting a full sample of luminance and a half sample of chrominance in the red channel.
My question is how much of the blue channel are you getting? I'm assuming it's not "0" because blue information is present in the video.
Please answer quickly because I'm wasting too much time thinking about this!
Matt Davis
July 5th, 2006, 01:02 PM
Please answer quickly because I'm wasting too much time thinking about this!
I am not an expert in this, but I have been lead to believe that 4:2:0 is a slightly more complicated version of 4:1:1 in that the chroma info is offset DOWN by 1 scan line, but basically has the same rez.
So the blue content follows the red content. Just ensure you follow 4:2:0 all the way through.
LURKER MATERIAL...
Bottom line: 4:2:0 (DVD, HDV and PAL DV) has the same problems as 4:1:1 (NTSC DV) - colour info is very low rez, so be very careful if you want chromakey/green screen stuff.
The very important thing to know is that if you shoot NTSC DV then go to DVD, the colour info suffers a further knock because of the transition from 4:1:1 to 4:2:0. Also, if you take HDV 1080i/60 (US), which is 4:2:0, down to NTSC DV (4:1:1) then encode to DVD (4:2:0) it's not nice.
I love PAL! DV, HDV and DVD all at 4:2:0, 25fps can be squeezed into 24fps, and 25fps bandwidth allows for more detail than 29.97fps.
Glenn Chan
July 5th, 2006, 01:02 PM
See the pictures at
http://www.adamwilt.com/pix-sampling.html
The formulas are something like:
Rec. 601: Luma (Y’) = 0.299 R’ + 0.587 G’ + 0.144 B’
Rec. 709: Luma (Y’) = 0.2126 R’ + 0.7152 G’ + 0.0722 B’
Cb = Y' - B'
Cr = Y' - R'
By doing some basic algebra, you can get G' (green) if you know Y', Cb, and Cr.
Nitty gritty details:
Y' : luma Some people call it luminance; however, this is not the same luminance as in color science see http://poynton.com/papers/YUV_and_luminance_harmful.html
Cb and Cr: chroma, or color difference components
chroma differs from the chrominance of color science
If you read carefully at Adam Wilt's site, note that there are different 4:2:0 schemes. These (2) other schemes sample the Cb and Cr components from different pixels.
So actually, 4:2:0 can be an ambiguous term.
Because 4:2:0 (the one used in PAL DV) has the Cb and Cr samples on different lines/fields, there's an interplay with interlacing since one of the components is sampled earlier than the other. see http://www.hometheaterhifi.com/volume_8_2/dvd-benchmark-special-report-chroma-bug-4-2001.html