0

Aproximayly 1/3 volume of a cube is at this picture from wikipedia

enter image description here

So does YCbCr JPEG have aproximayly 2^(3*8)/3 = 5 592 405 colors?

Also coefficients KR+KG+KB=1 are used for converting YCbCr to RGB. Does each one dicrete of Cb or Cr change one of the R or G od B by one discrete? How many colors will remain after the transformation from 3x8bits YCbCr to 3x8bits RGB (that is common conditions)?

Improve this question
4
  • Hi. Welcome to GDSE. This isn't really graphic design question, so I'm voting to close it as off-topic. Sorry about that. Perhaps try Signal Processing Stack Exchange, or maybe Stack Overflow.
    – Billy Kerr
    Commented Feb 20 at 13:28
  • Is this graphicdesign.stackexchange.com/questions/138943/… not off-topic?
    – Imyaf
    Commented Feb 20 at 13:32
  • 1
    CMYK and RGB questions are on topic here because these colour modes are used all the time in graphic design applications. YCbCr is not. It's not even an option in most graphic design software. It's usually something done under the hood (by the software) when exporting JPEGs. When we import JPEGs, the software decompresses it, and converts it to RGB. I think you would probably get a better response on a stack exchange site that deals with such image processing.
    – Billy Kerr
    Commented Feb 20 at 17:26
  • I don't think there is a simple answer to your question. Numerical values don't necessarily correspond to a single colour. That's why the same RGB or CMYK values look different depending on the profile used.
    – user183813
    Commented Feb 20 at 17:46

1 Answer 1

Reset to default
0

To reduce math precision problems Y can vary 16...235, Cb and Cr can vary 16...240. That's said in the Wikipedia article where you caught your image from.

The remaining number of available color codes is 220 x 225 x 225 = 11137500.

That's theory. JPEG compression causes errors. A mathematician is needed to check if the actually used JPEG coding and decoding implementation reduces the number of available color codes or generates some new Y, Cb or Cr values which are not inside the original range and the display system shows them as new colors. Such new colors, of course, are erratic, but they might be seen. Finally the conversion to a RGB display can make more or less available codes useless.

I'm afraid very few of us know enough of the actually used implementations. Let's hope a valid answer which covers some commonly used real systems will be written soon. While waiting one ask also people who work with image signal processing.

Improve this answer
2
  • Aproximayly 1/3 volume of a cube is at the picture from wikipedia. For example, If Y = 0 or Y = 1.0 (255 of 255 or 235 of 235), than Cb and Cr can be only exact 0 (only one discrete from 256 or 225). So, amount is not 219 x 225 x 225 or 256^3, must be more less.
    – Imyaf
    Commented Feb 21 at 18:56
  • I had miscalculated the number of available Y levels. Now we disagree even more. You seem to have done the following reasoning: "The image seems to have only a third of a full symmetric cube, That's why 3x8 bit cannot have 256 x 256 x 256 combinations, but only a third of it". I wouldn't use any image as a foundation of math reasoning if I didn't know the math facts behind just that image. The linked article gives no facts of the geometric dimensions of the illustration.
    – Eric Willis
    Commented Feb 21 at 20:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.