Questions concerning Blu-ray video/colour

Welcome to HTF Drew.  This is a pretty specific question.  I'm going to move it to the Display forum to increase our chances in getting it answered.

hi

you are sort of on the right path.

First video is 16-235 and not 0-255 which refers to PC levels.

The color gamut (or triangle) is defined by whatever color space is being used. Rec 709 is the usual space for HDTV. Although we call this triangle a triangle it is actually more of a 3 dimensional pyramid as the Luma value of a color will dictate the size of the triangle. 4:2:0 refers to to sampling of the RGB encoding. This is done to conserve bandwidth and in most circumstances does not affect viewing though some argue that this creates artifacts. One person I recently talked to in post production said that their job was to "manage artifacts".

Hope this helped. Poynton has some great books on the subject.

Gregg

A more specific answer to the question regarding the numbers...

4:4:4 is uncompressed, fullband on all channels.  Uncommon outside of high-end production.

A lot of medium-end production is 4:2:2, which has full-band on the Luma, and half band on the Pr and Pb (or u and v, or whatever the naming scheme is -- the two color channels!)

DV and conventional DVD is generally 4:1:1, which has full-band Luma, and quarter-band on the two color channels.

Some DVD is 4:2:0.  This is akin to S-VHS, which combined the two color channels into a single composite color-band -- although not "composite" as conventional VHS, but still independent of the Luma channel.

I've asked a couple people and I haven't gotten an answer as to if 4:2:0 is the same, better, or worse than 4:1:1.  I think most people would prefer to do post with 4:1:1 instead of 4:2:0, but again, I haven't gotten a clear answer on that.

Now, onto the realm of Color Gamut.  I can't remember where I saw this description, but I'll try to recreate the illustrations in words.

Imagine a 3-D chart, with Red, Green, and Blue on the X, Y, and Z axies.  (Help, what's the plural of Axis?)  If you take a 24-bit color gamut and plot it's area sorry, volume on this chart, you get a cube.

Now, draw a new 3-D chart.  X and Y are the U and V channels, and the Z element is Luma, or Y.  (Z, in this chart, is going "up", while X goes left right, and Y goes into the back of the screen.)  The Yuv color-space is now a diamond, forming along the vertical, and extending to the center of the X and Y axies. 

Is it the same as the RGB, but described in a different manner?  Sort of.  Maybe.  The math generally says it is, but the math isn't taking into account the actual limitations of the standards.  Such as, for example, the 16-235 on the luma channel.  It's yet another form of compression, which, as Gregg properly indicates is important from the perspective of managing artifacts.

Leo

My understanding, and this may not be accurate, is that, mathematically, there should be no difference between RGB and Component.  If this is actually accurate or not, I don't know.

On to point #2.  My understanding is that DVD-video can be either 4:2:0 OR 4:1:1.  And in a grand total of 30 seconds of research, I didn't find anything that indicated that BluRay couldn't be 4:1:1 or even 4:2:2.  Most of the actual codecs do have extensions that support 4:2:2 and even 4:4:4.  But, moving on, a lot of the actual "masters" aren't really video files.  I know during production, interchange between post-houses and effects houses and things like that is often done with huge collections of numbered, uncompressed still image files -- be they TIFF, DXN? (Or is it DNX?) BMP, or perhaps even PSD, depending.  During production, from what I understand, 14-bit files (aka 42-bit, or 56-bit with alpha-channel, which might be designated 4:4:4:4!) aren't terribly uncommon.

I would hope that any "deep color" expansion would be a jump to at least 10bit color-depth. 

Leo