Tip:
Highlight text to annotate it
X
In this movie, we'll wrap up our expirations of terms using color correction
and grading with some under the hood technical information. We're not going to
cover everything, but we're going to cover some of the basics of this technical nomenclature.
So what do we need to define? We need to define RGB. We need to define Y'CbCr.
We need to discuss the differences between RGB and Y'CbCr. We need to talk
about Chroma subsampling, and we need to talk about Bit depth.
So what's RGB Color encoding all about? RGB is an additive color model and it's
how we see the world. In the RGB model, Luma or lightness and Chroma
are not separate. A manipulation of one affects the other. Only recently have
some video systems been capable of handling RGB video.
What's Y'CbCr encoding all about? Well, it's the way that most component
digital video is encoded, and sometimes the terms YUV is used as a generic,
even though technically this is wrong. Y-prime represents Luma and CbCr
represents the Chroma difference channels, the colored portion of the video signal.
This model allows for separate manipulation of Luma and Chroma. This model
allows for greater compression of Chroma, because Luma and Chroma are separate.
So let's compare RGB to Y'CbCr. In RGB encoding, Chroma and Luma are linked.
In Y'CbCr encoding, Chroma and Luma are separate. Neither color model is correct,
but RGB manipulations are how many computer applications work, and different
ways of thinking are required for each model, because in RGB, Luma and Chroma
are linked, and in Y'CbCr they are separate. So what do I mean by Chroma subsampling? This
is a term that we hear quite often. Basically Chroma subsampling refers
to the reduction of Chroma in the overall video signal. Because our eyes are
more sensitive to Luma than color, Chroma subsampling reduces Chroma in a video
signal. Chroma subsampling ratios are often represented
by ratios like 4:2:2 or 4:1:1. The first number represents Luma, the second
two numbers represent the Chroma difference channels, or the color portion
of the signal. The first number, Luma, is generally 4, while
the Chroma difference channels are a half to even a quarter of the first
number. Chroma subsampling can also serve as compression,
because we're reducing the amount of Chroma in the signal, thus the overall
data rate of the video signal is less. For color correction and grading purposes,
the more Chroma that you can get, the better. Some newer formats like Sony's
HDCAM-SR can actually record 4:4:4, so there is no Chroma subsampling taking place,
and this provides a pristine, perfect image for color correction and grading.
Bit depth is another concept that we hear about all the time. Bit depth refers
to the range of colors that are available in any given color model. Video is
commonly 8 or 10 bit, although some programs can process greater bit depths.
An 8-bit RGB image has 256 possible values per pixel, while a 10-bit image has
1,024. Due to the way that digital video is encoded, most of the time we can
only access 220 Luma values for 8- bit and 876 values for 10-bit video.
Some newer codecs can actually support RGB encoding, so we get the full range
of values. 0-1,023 in 10-bit, counting the 0.
Basically, the more values the better for color correction. Whenever possible,
work with 10-bit video. Up next, we'll talk about another very important
concept in our discussion about the basics of color correction, physical
setup of your grading environment.