-
Terminology
Ok, so this may (and probably will) be a stupid question, but there are obviously a lot of discussions concerning various subjects, but a lot seem to encompass some things I sort of understand, but can either find descriptions on the internet for complete novices (ie ‘4444 is better than 422 because it is better quality’), or for people who don’t seem to need it explaining (‘the matrix occurs when 5641615 parts of code line 16 equate to the vertex navigator of the codec which obviously equals the YUV alignment of spectral 42’).
So if possible, I’d love in KIND of layman’s terms for someone to explain:
8 bit video
10 bit video
422, 422 HQ and 4444 (I appreciate they’re codecs for good useable compression in FCP but I’m not sure about the differences :S)
YUV and RGBI know (from searches) other posts touch on explanations but there’s nothing particularly comprehensive that I could see :S
Cheers!
Tom😀
Tom Morter-Laing
Certified Apple Product Proffessional, 2010
Production Assistant, Grace Productions
Degree; TV Production -
4 Replies
-
Many questions Thomas.
[Thomas Morter-Laing] “8 bit video 10 bit video “
10 vs 8b basically mean more precision.
Imagine that you have to measure the size of an object and you have two rules, one in centimeter and another one in millimeters. With the rule in millimeters your measure will be more precise.
So think of the 8b a rule that have only marked the centimeters and 10b a rule that every centimeter have 3 in between marks.
That’s all.[Thomas Morter-Laing] “422, 422 HQ and 4444 “
Forget about that HQ.
That means just more data rate for the Prores 422 HQ than for the standard Prores 422.Everything you need to know about chroma downsampling is in this article of the master Nattress;
https://www.kenstone.net/fcp_homepage/chroma_investigation_nattress.html
Although the industry uses this term, 444 doesn’t makes sense when talking about RGB. In video, RGB is always 444.
rafael -
[Thomas Morter-Laing] “Ok, so this may (and probably will) be a stupid question,”
Google maybe? these are really really basic questions.
“8 bit video -10 bit video “
First “Video”- is a moving image captured by a imager and recording system in the Y’Cb’Cr color space( which is called YUV in the Standard definition formats.8bit or 10bit – the levels of grey that can be recorded 8bit is 2 to the 8 power or 256 levels of grey. 10bit is 2 to the 10th power or 1024 levels of grey that can be recorder in each channel the next is 12bit at 4096-levels.
“422, 422 HQ and 4444 “
https://en.wikipedia.org/wiki/Chroma_subsampling
422 and 4444 in this instance are defining color and sampling of the image when recorded.
4:2:2 defines color recording somewhat the same as your eye sees, where for every 4 luma samples there at 2 +2 samples of color ( or half the color info for the same number of luma samples)
444+4 – is one R /G/B/ sample for every Luma Sample- the extra 4 covers the alpha channel or mask info.
In the Apple ProRes world the terms 422, 422HQ and 4444 are merely names given to the various flavors of the ProRes codec with increasing levels of quality with decreasing levels of compression. the 422 codec is about 145mpps, the 422HQ being about 220mbps and the 4444 version being about 330mbps compressions.
“YUV and RGB “
As I said YUV is the color space for video – and it is correctly called Y’Cb’Cr’ and it functions like CIE LAB color does in print- it allowed a B&W signal to be extracted from the color broadcast long before the advent of cable and color TV.
RGB (red green and blue)- the colorspace that is defined by transmitted light and what most FILM records as.
gary adcock
Studio37Post and Production Workflow Consultant
Production and Post Stereographer
Technology Development
Quality Assurance AssistanceChicago, IL
https://blogs.creativecow.net/24640
-
the first number (4) is the multiplier applied to the sampling frequency of the luminence channel
the second two numbers are the multipliers applied to the sampling frequency for each of the color difference channels
HQ is short for high quality but is really a misnomer and out to be used for material 2K and higherWhen a fourth number is present it refers to the alpha channel and is once again the multiplier applied to the sampling frequency of that channel.
The effective sampling rate for HDTV is 74.5 MHz
Some heavy reading yet concise
Doug Beal
Editor / Engineer
Rock Creative Images
Nashville TN -
[Doug Beal] “the first number (4) is the multiplier applied to the sampling frequency of the luminence channel
the second two numbers are the multipliers applied to the sampling frequency for each of the color difference channels “The first 4 means than in a 4 pixels block, all the pixels have their Y’ value.
If the codec is 444, means that apart from the Luma, every pixel have his proper C’b and C’r values.
422 means that every pixel have his Y’, but only one every two pixels have C’b and C’r.
Same for 411: Y’ in every pixel, and C’r and C’b only one every 4 pixels.
420 comes from the PAL world, and the pixels block instead of being part from an unique line, takes two pixels from one line and two from the adjacent. But in the end is 4 Y’ samples and 1 C’b plus 1 C’r from different lines.
Rafael
Reply to this Discussion! Login or Sign Up