Can somebody explain color space to me?

D

Dirty_Ass

Well-known member
Any links, info, basic knowledge for me. It's extremely overwhelming, and I dont even know where to start with it. Any info would be appreciated
 
Well do you mean color sampling (4:1:1, 4:2:2, 4:4:4 etc) or color space such as LAB, RGB, CMY etc?
 
Both. I've heard all sorts of talk about 4:4:4 and I thought it was some ****ed up sort of pulldown. Color sampling and color space are ****ing with my head. Pardon my noobness, but any info would be appreciated
 
The 4:4:4 etc. are digital color sampling resolutions. The first number is the brightness which is always 4 (Im pretty sure of that), the 2 numbers that follow are 2 color channels. The higher the numbers the better the resolution. So 4:4:4 means the 2 color channels are equal to the brightness resolution. 4:2:2 means that the 2 color channels are half the resolution. 4:1:1 means they're at 1/4 the res. 4:4:4 is high end - Cinema etc. I'm pretty sure the DVX is 4:1:1. That's my basic understanding of it, I could be a bit off cuz I'm not a very technical person.

As for CMYK, RGB, LAB etc.
CMYK is a printing term for color (cyan, magenta, yellow, black) These are the 4 colors that printers use to make all the colors you see in magazine ads, product packaging etc. Those 4 colors block certain light rays to give you a full spectrum of color.
RGB that's what video, web pages, etc. is made up of (red, green, blue) These colors are mixed through pixels to give a full spectrum of color. For instance: R:255 G:255: B:255 will create white. R:0 G:0 B:0 will create black. R:255 G:132: B:0 will create orange and so forth.
LAB color is similar to RBG just uses a different formula L is black & white, A is the spectrum between Green and Magenta, and B is the spectrum between blue and yellow. Mixing these values between 1 and 100 creates a full spectrum of color.

Well that's my basic understanding on that. Hope it helps.
-Steve
 
When it comes to video color space and RGB, there are many different types beyond just this. You have Log, linear, and vid space's. Each have there own use. Log files are generally what Cineon (film scans) files are stored in. These will look very dark, almost black, on your monitor without using some kind of Lookup table. Linear has to do with float point and can handle overbright pixels, i.e. HDR images. I understand this to a degree, but I find it much harder to explain.

Go check out Stu Maschwitz's blog on color spaces, lin, log, etc...lots of good into there.

http://prolost.blogspot.com/
 
Here is an excellent tech article on YUV colorspace http://en.wikipedia.org/wiki/YUV

However, your question seems to be more about sampling within the color space.
4:1:1, 4:2:2, 4:2:0, and 4:4:4 are all different samplings of YUV colorspace.

What it all boils down to is color quality vs the bandwidth required to carry the data stream. The first number says that every pixel in the camera provides luminance information to the stream. The second and third numbers are about how the color info is handled (see http://en.wikipedia.org/wiki/Chroma_subsampling)
 
svf1972 said:
As for CMYK, RGB, LAB etc.
CMYK is a printing term for color (cyan, magenta, yellow, black) These are the 4 colors that printers use to make all the colors you see in magazine ads, product packaging etc.
Click to expand...


Partially right. CMY (the K is only added in print) are the primary colors of the subtractive system of color. It's not just used in print. For example, color film from top to bottom goes Y M C. That is, the subtractive system utilizes tools to "remove" colors from light. So a filter, a gel, film, etc are all subtractive.

RGB are the primary colors of the additive system of light. That is, when you shine two lights together, you add their colors together. This color model is used more in theatrical lighting, not in film practice. It is also the standard used for graphics designed for the web or computer monitor viewing.

basically, with subtractive you start out with white light and take away light to make colors. with additive you start out with blackness and add light to it make colors.

Thus, if you put a Blue gel on white light, essentially what you are doing is removing all colors but blue, especially its compliment, yellow, and only allowing blue to pass through. By removing all those other colors, you are also decreasing the amount of light coming from the source. Thus gels and filters reduce light.

And remember, you add a compliment to a color to reduce its value. So if you wanted less blue, you add yellow. etc.

On a side note. I learned from a very talented man named Bruce Block, that computer monitors and televisions actually use NEITHER system. We all assume RGB, but what they actually use is little lights (called pixels or nodes) that are too small for the human eye to discern at a given distance. Those lights don't add or subtract color, they SWITCH (not mix together) from red to green to blue at a given refresh rate. Thus, from a given distance, when we see a matrix of these pixels, our eyes do the adding together to create the colors.

I'm sure this confuses you even more. The point is: CYMK = print is a misgnomer. CMY is very much used in film and video as well. Everytime you put a gel up, you are subtracting color!
 
zoostory said:
On a side note. I learned from a very talented man named Bruce Block, that computer monitors and televisions actually use NEITHER system. We all assume RGB, but what they actually use is little lights (called pixels or nodes) that are too small for the human eye to discern at a given distance. Those lights don't add or subtract color, they SWITCH (not mix together) from red to green to blue at a given refresh rate. Thus, from a given distance, when we see a matrix of these pixels, our eyes do the adding together to create the colors.
Click to expand...

Hmmm... well I'm sorry to say your knowledgable friend is about 80% BS and 20% misleading.

No It isn't a 'true' RGB additive color space because the R the G and the B aren't being added into the same physical pixel location.

Yes it is an additive RGB color space because your eyes suck, and they can't tell that the R, G and B dots aren't in the same location.

No individual phosphors are not changing colors.

Yes, your monitor does refresh multiple times per second, but no, it only takes one pass to create all of the color. It refreshes multiple times per second often to repeat the exact same color (Phosphors are only energized briefly, except for on cheap LCDs, then you get smearing.)

I recommend a hearty re-education here:
http://en.wikipedia.org/wiki/Cathode_ray_tube
and here:
http://en.wikipedia.org/wiki/Liquid_crystal_display
 
im.thatoneguy said:
Hmmm... well I'm sorry to say your knowledgable friend is
I recommend a hearty re-education here:
http://en.wikipedia.org/wiki/Cathode_ray_tube
and here:
http://en.wikipedia.org/wiki/Liquid_crystal_display
Click to expand...


That one guy. I read both articles. And, to make matters simpler, I will stick to CRT. The article actually ramified what I was saying. Or was trying to say, pardon my lack of clarity.

The bottom line is, within one pixel, or "phosphor" there is no additive color mixing. Each phosphor can only emit red, green, or blue at any given time. Not an addive mixture of the two.

To quote the article you turned me to:

"Color tubes use three different materials which specifically emit red, green, and blue light, closely packed together in strips."

Thus, as you correctly stated, our eyes blend together these "strips" and tell our brain that a color is being created. But if you look under a microscope (again, here I will reference an image from the article you turned me on to... take a look at "5")

CRT_color.png


You will find that no one phosphor is anything but red green or blue.

The fact that it is our eyes doing the blending, NOT the lights themselves, is what leads this not to be an additive system, but something else. I wish I could remember the technical name for this system, I will have to look it up. I think we agree on everything, only I am saying that by our eyes being tricked instead of actual light mixing is NOT called the additive system. All this is just terminology, and I think we are on the same page in regards to how light and color actually work. It's all negligible in the end, as none of this really helps this guy figure out what color space is :)

All that being said. My "friend" Bruce Block is actually more my former professor Bruce Block, here at USC. He teaches and has taught a class in Visual Expression for over twenty years. Not just here, but at AFI and all the major studios, and at processing labs. You can get his book here:

http://www.amazon.com/gp/product/0240804678/sr=8-1/qid=1143776776/ref=pd_bbs_1/002-7592314-0202401?%5Fencoding=UTF8

He's an expert in the field and could probably explain all this way better than me. I highly recommend this book to everyone. It is a language for approaching the story through visuals, and it changed the way I look at cinema.
 
Last edited:
I don't want to hijack this thread or anything but I'm not really understanding why someone would really need to know this, other than it being an interesting subject relative to video. I assume that it is important if a professor at USC (and probably other film schools) is teaching this. Is it really practical to have an understanding of light space?
 
Dustin Wadsworth said:
I don't want to hijack this thread or anything but I'm not really understanding why someone would really need to know this, other than it being an interesting subject relative to video. I assume that it is important if a professor at USC (and probably other film schools) is teaching this. Is it really practical to have an understanding of light space?
Click to expand...

Well, in this digital era of filmmaking this holds some importance, especially for post production. For some indie filmmakers it's not a big deal, but for anyone who wants to stay on top of the technological curve, it's good knowledge.

Knowing what color space you're working in helps especailly in visual effects/DI to create certain looks. Working in higher color spaces gives you better results. I try to comp all my shots in float point (32-bit Linear color space) because the results are more natural and I can work in photographic terms, i.e. make a part of the shot one stop higher, etc. It also helps with banding and other artifacts while CC'ing. When it comes to footage if you have YUV 4:4:4 Log Cineon files you will need to know that to apply a LUT to view your footage on a computer monitor, otherwise it's practically all black. Also, if you're shooting 4:1:1 you will want to understand the limits of that color sampling, etc.

So, it helps to know a little, but it gets very technical and you definately don't need to know it all. Digital is here to stay and having these tools and knowledge at your disposal will only help.
 
If you plan to do any sort of higher end color correction and grading or visual effects this stuff is imperative to understand. And it's just plain fascinating =)
 
Dustin Wadsworth said:
I don't want to hijack this thread or anything but I'm not really understanding why someone would really need to know this, other than it being an interesting subject relative to video. I assume that it is important if a professor at USC (and probably other film schools) is teaching this. Is it really practical to have an understanding of light space?
Click to expand...

it is only necessary to know enough that you can sound like you understand.

"Listen, if you plug the CMYK cable into the XLR jack and cut out the carbagen dioxygens, you're gonna boost your RGB's up over 10db! How are you gonna work with that? Huh? How?"
 
This is the best definition I've seen.
It's what got me to totally understand what 4:4:4, 4:2:2 AND 4:1:1 really means.
It will make you understand why DV video (4:1:1) is not the best for green screening (chroma key) footage.
http://www.macbreak.com/

It's a video podcast and is hosted by visual effects artist Alex Lindsay. He worked at ILM doing GRFX on Star Wars episode 1 and has moved on to teaching visual effects.
The podcast (vidcast) on color space is called: "The Road to 1080p (part 2)". Not part 1 but part 2 is what you want.
This vidcast was so informational to me I can practically teach what 4:4:4, 4:2:2 and 4:1:1 color space means.

When you get to Macbreak.com click the iTunes link to get to their podcast series. They have about five I think but start with the one I mentioned above. It's a real eye opener
 
You must log in or register to reply here.
Back
Top