No announcement yet.

Krita 5.0 Beta Released With Better Performance, UI Polishing

  • Filter
  • Time
  • Show
Clear All
new posts

  • #71
    Originally posted by RomuloP View Post
    This could not be more wrong in fact, color spaces are infinite coordinate spaces, with coordinate transformations you can map any point from any color space to any other using XYZ color space
    Except this is half true. Yes when you map a individual color this works. When you map 2 or more this can cease to work. Mathematically says this is a problem.
    Safe color mapping for each spot color you need channel. So CMYK printer only has 4 channels so can only take 4 spot colors include K channel.

    What is the problem here you have failed to take in the difference between additive colours and subtractive ones. CIE 1931 XYZ color space is what you would be referring to by XYZ color space. XYZ color space is additive. CMYK is subtractive. This is one of the traps there is no black in the XYZ color space.

    Another fun point photoRGB has points outside the XYZ color space and that is need.

    The XYZ color space is in fact wrong. Humans with the condition commonly called tetrachromacy can see color that are outsize the XYZ color space by a large margin. Depending on the lighting as well printable colors outside the XYZ color space can also be visible normal humans without tetrachromacy.

    The color range of the XYZ color space was decided by a total of 10 people unfortunately those 10 people did not have best vision heck not even average for color detection for a human. What partly screwed them is their sex all the people who deterred the XYZ color space were male. Unfortunately odd part of humans is that tetrachromacy and other advance human vision traits only appear in people with two X chromosomes so females or Klinefelter syndrome males(these are insanely rare). Yes those who made the XYZ color space had average color detection for a male human but over half the humans are not male and its the female humans with the better vision. Yes the man problem where lady asks him to pick between 2 colors and not be able to tell the difference can 100 percent be the vision difference between male and female humans. So yes XYZ color space is another historic screw up. Yes people who don't know this subject always say what you said that you can use XYZ color space to map from one color space to another that would be possible of XYZ color space was correct in the first place.

    Originally posted by RomuloP View Post
    What is true is that many printers today will be able to do a better job converting from a specific RGB profile than a CMYK one simply because some printers are adding spot color or even blue, red or anything else, and if you can avoid CMYK completely, this turn to be great so nobody cares to work on mapping the printer profile to a proper CMYK profile for color proffing..
    One of the problems here your complete CMYK maths are wrong. CMYK software maths are not design to take account for the additive parts of the pigment(the inkj) instead only design to take account for the subtractive. This is why almost every time someone says they have their correct color profile for their printer for CMYK when you in fact test it they are in fact off.

    The problem here is not being able to use a spot ink profile per channel. This measures two different things the colours absorbed by a ink and the colours reflected by the ink that give two different RGB values being additive and subtactive. CMYK even with the color proffing is a hack its not designed for printing. The CMYK designed for print is CMYK spot printing that use per channel profiles and this is a different beast.

    The reality here is software CMYK and XYZ color space are screw ups and are done incorrectly. We really do need spot channel printing for printers where you can put the ink properties on each channel and have what ever number of channels the printer has. Most inks are both additive and subtractive. Less common ones are only additive or only subtractive they are not what is in your general printers. CMYK in software is based around a theoretical ink set of pure subtractive that does not exist in the common market space that people are incorrectly attempt to fix for by apply printer profiles designed on the theory of subtractive of course this is not working right.

    The levels of screw up here are insane. Yes people have got really stuck in their way with CMYK for printing are are not willing to step back see it wrong and see that they need spot printing methods so they can in fact configure the channels correctly for the inks. Yes a lot of print shops issues with having to do many samples with customers comes in a lot of time due to the use of CMYK and the problem how the calibration is off.


    • #72
      Originally posted by oiaohm View Post
      Except this is half true. Yes when you map a individual color this works. When you map 2 or more this can cease to work. Mathematically says this is a problem.
      No it is not, mathematically this is perfectly fine. I recommend you give a look at Nine Degrees Below Photography programmers guide to XYZ, RGB, ICC, xyY, and TRCs as a starting point, specially chapter Colors that weren't measured and Why so many RGB working spaces? Again, mostly, your complains and suppositions really don't matter, if you want yo assume what I was doing was spot color, you are wrong, we made a proof with a chemical proof generic profile and their instructions on the amount of black and cyan and result got perfect.


      • #73
        Originally posted by RomuloP View Post
        [...] for painting [...]

        Actually, we can see how it may be done:

        There's more in: