Gamut Mapping and Digital Color Management

Transcription

1 Gamut Mapping and Digital Color Management EHINC 2005 EHINC 2005, Lille 1

2 Overview Digital color management Color management functionalities Calibration Characterization Using color transforms Quality control ICC Profile quality issues 2

3 Gamut Mapping and Digital Color Management Digital color management 3

4 Digital color management Why color management? Different types of color devices: input - monitor - output 4

5 Digital color management The color management solution TRANSFORM TRANSFORM TRANSFORM 5

6 Gamut Mapping and Digital Color Management Color management functionalities 6

7 Color management functionalities Calibration Characterization Device modeling Gamut mapping Building color tables Using color transforms Quality control 7

8 Calibration Checking and readjusting color devices to predefined settings and specifications according to manufacturer s recommendations. Calibration = Measurement + Control 8

9 Calibrating color devices Input Monitor Output Automatic internal scanner calibration 9

10 Calibration Defining tonal responses Condition on single ink processes Quantity is visual Tonal response is linear Fix endpoints of the ink processes Ink limitations for C, M, Y and K Multi-density inks Increase apparent resolution Measure ink combinations Choose an Ink Mixing Path 10

11 Defining tonal responses Measured quantities are visual Tonal response is linear Visually stable Smooth color gradations Multi-density inks: dot gain? Equal ink steps are equal visual steps => better use of levels in profiling => uniform characterization target Use CIELAB L* for C, M and K C* for Y 11

12 Defining tonal responses Fix endpoints of the ink processes 100 C* (for yellow) Differs! Fixed in L *or C* NOT in Ink % Ink % 12

13 Defining tonal responses Ink limitations What is needed? What is possible? Large enough gamut Avoid bleeding, coalescence Drying characteristics Difficult trade-off Must be done by end user Determination of bleeding & gamut based on measurements 13

14 Defining tonal responses Ink limitations (cont d) Bleeding: Correlate characteristics 14

15 Defining tonal responses Ink limitations (cont d) Ink limitation wizard: Trade-off: bleeding <=> gamut 15

16 Multi-density inks Increase apparent resolution (Cl,Ch,Ml,Mh,Y,K) More complex 0% Multi-density inks: More opportunities Practical solution: Ink Mixing Balance: visibility dots and ink use 0% Light cyan 100% Heavy cyan 100% 16

17 Multi-density inks Measure ink combinations Lightness Variations Hue variations L* ht g i L b* % 100 n cya 0% 100 y v a e % H n cya % a*

18 Multi-density inks Choose an Ink Mixing Path 90 Light C Ink % CL and CH Light 30 Heavy 20 Heavy C Ink % global C

19 Characterization Measuring the color behavior of a color device. During characterization a relationship is built between the colors produced by a device and a device independent color space Building color transforms => Profile Device dependent space device independent space 19

20 Characterization Input devices Scanners Digital cameras Transparent (IT8.7/1) or Reflective (IT8.7/2) photographic scanner target (Agfa/Fuji/Kodak) Internal image rendering, often to srgb or wide gamut RGB space RIMM / ROMM spaces 20

21 Characterization Scanners Scan Digital Reference Data Color measurements Calculate Profile 21

22 Characterization Monitors Measure Calculate Profile Test pattern 22

23 Characterization Output devices Measure Print Digital Reference Data Calculate Profile IS12642 printer target 23

24 Building color transforms Device modeling Gamut mapping Building color tables CMS 24

25 Device modeling : Scanners Device dependent color space: linear RGB space R = I ( λ ) R ( λ ) S R ( λ ) dλ G = I ( λ ) R ( λ ) S G ( λ ) dλ B = I ( λ ) R ( λ ) S B ( λ ) dλ with I(λ) the illuminant in the scanner R(λ) the reflectance curve of the scanned object SR(λ) the spectral sensitivity of the red scanner channel 25

26 Device modeling : Scanners Scanner model Relation RGB and XYZ not unique due to scanner metamerism Empirical technique for scanner modeling based on polynomials e.g. X = arr + agg + abb + argrg + agbgb + abrbr + arrr2 + aggg2 + abbb2 + Y = brr + bgg + bbb + brgrg + bgbgb + bbrbr + brrr2 + bggg2 + bbbb2 + Z = crr + cgg + cbb + crgrg + cgbgb + cbrbr + crrr2 + cggg2 + cbbb2 + 26

27 Device modeling : Monitors Device dependent color space: RGB space Modeling Simple model : Gain Offset Gamma model (GOG) based on chromaticity coordinates of the phosphors gamma values for the phosphors white point black point works fine if phosphors are stimulated independently Complex model : Polynomial fitting 27

28 Device modeling : Monitors GOG model Luminance γr R R = ar d + br 255 γg G G = ag d + bg 255 γb B B = ab d + bb 255 X m11 Y = m21 Z m 31 m12 m22 m32 m13 R m23. G m33 B γ >1 Voltage with (Rd,Gd,Bd) the monitor RGB values (R,G,B) the tristimulus values relating to the phosphors (X,Y,Z) the XYZ tristimulus values The mij values are determined by the phosphor chromaticity values and the white point of te display 28

29 Device modeling : Printers Device dependent color space: CMY,CMYK,RGB,... Printer model Function colorant space color space Domain colorant cube Range gamut 29

30 Device modeling : Printers Modeling Subtractive color mixing Lambert-Beer law e.g. photography Printing devices with halftoning Neugebauer process combination of additive and subtractive color mixing e.g. offset printing 30

31 Device modeling : Printers Neugebauer model e.g. X of the XYZ space for a CMY process: X = 1 c 1 m 1 y X w + c 1 m 1 y X c + 1 c m 1 y X m + 1 c 1 m yx y + cm 1 y Xcm + c 1 m yx cy + 1 c myx my + cmyx k with c,m,y, the CMY colorant values More complex behavior Polynomial fitting 31

32 Gamut representations xy-chromaticity diagram provides limited gamut information incorrect for scanners sufficient for monitor poor gamut info for printers Needed color values primary and secondary colors 32

33 Gamut representations Nrofcolors per lightness / hue section Pro: limited data to represent a gamut more accurately good gamut representation if both systems have a similar color mixing behavior e.g. comparison of different screening techniques gamut due to different total amount of ink Contra: insufficient information to compare gamuts of different types of color reproduction devices accurately 33

34 Gamut representations Offset process with different total ink amounts Nrofcolors per lightness section Offset 400 Nrofcolors 8000 Offset Offset Offset 150 Offset Offset Lightness 34

35 Gamut representations Offset process with different total ink amounts Nrofcolors Nrofcolors per hue section Offset 400 Offset 250 Offset 200 Offset 150 Offset 100 Offset Hue 35

36 Gamut representations Gamut cross sections Most important cross sections constant lightness constant hue Different approaches Heuristic techniques Analytical techniques: measuring many printed color patches detection of boundaries; e.g. with convex hull construction of a model for the color device calculation of gamut boundaries based on this model Required for accurate gamut mapping 36

37 Gamut representations Gamut cross sections Offset processs / srgb L* = 50 Offset process / srgb hue = 0o 37

38 Gamut representations Offset processs lightness intersection at primary and secondary colors Red Yellow Green Cyan Blue Magenta 38

39 Gamut mapping Different gamuts due to different color technology or device settings Input device Output device Proofing device 39

40 Gamut mapping Rendering intents = mapping strategies Photographic (perceptual) Colorimetric (absolute/relative) Computer graphics (saturation match) Proofing / Simulation Source profile Destination profile Proofer profile 40

41 Gamut mapping Colorimetric mapping Perceptual mapping L* L* Original gamut 0 C* 0 Reproduction gamut C* Gamut mapping 41

42 Building color tables Regular grid in device dependent space or device independent space Number of sampling values Sampling points Interpolation Multi-linear interpolation Tetrahedral interpolation 42

43 Using color transforms Color Management Module (CMM) in OS => to be used by applications TRANSFORM TRANSFORM CMS TRANSFORM 43

44 Using color transforms Link keys Rendering intent Black point compensation Smart CMM Link exceptions Black: to render text 400 % CMYK Keep white Primary and secondary colors 44

45 Using color transforms Link key: Black point compensation L* L* BPC C* C* 45

46 Using color transforms Link key: Smart CMM Errors mainly at gamut boundary L* Interpolation error L*-gamut cross section Proofing device b* a* Gamut proofing device Gamut mapping Offset 46

47 Quality control Goal Monitor consistency of print Identify problems Propose solutions (e.g. calibration) Method Measure control strip 47

48 Quality control Result: Consistency scores Global and per color Linked to tolerances 48

49 Quality control Solution: Rule based system Based on expertise & common sense 49

50 Quality Management System QMS Quality control of the workflow 50

51 Gamut Mapping and Digital Color Management ICC: an open CMS 51

52 ICC International Color Consortium Goal: Create, promote and encourage the standardization and evolution of an open vendor-neutral, cross-platform color management system architecture and components 52

53 ICC: Profile format Concept : Color processing model Converting device color data in and out of a reference Profile Connection Space (PCS) Combination of 3x3 matrix, 1D and multidimensional LUTS Smart profiles, dumb CMM Structure Profile structure : Header Tag description Tags 53

54 ICC: Profile Classes Profile classes Input device Display device Output device Color space (device to color space) Device link (device1 to device2) Abstract profile (PCS to PCS) Named color (e.g. Pantones) 54

55 ICC: Color spaces Different spaces are supported XYZ, CIELAB (PCS) Gray RGB, HLS, HSV CMYK, CMY Profile Connection Space based on CIE 1931 Standard Observer D50 illuminant, 500 lux 0/45 or 45/0 Measurement Geometry Black and white point, reflective medium 55

56 ICC : Dumb profiles, smart CMM Profiles Measurements Viewing conditions Parameters (preferences,...) CMM Modeling devices Gamut mapping Appearance matching User preferences 56

57 Reference medium gamut RMG Union gamut several output processes Perceptual intent Primaries to be defined Defined in maximum chroma for L*-C* combinations RMG: Lightness 50 intersection b* a* 57

58 Reference medium gamut RMG: hue intersection Lightness degree degree Chroma 58

59 Reference medium gamut Hue Lightness

60 Without ICC 60

61 With ICC 61

62 Gamut Mapping and Digital Color Management Profile quality issues 62

63 Profile quality issues Yellow mapping often problematic Mapping has to be adapted Sampling points have to be well-chosen 63

64 Profile quality issues Lightness intersections offset- proofer Red Yellow Green Proofer Cyan Blue Magenta Offset 64

65 Profile quality issues Black dots are often too disturbing Extra gamut limitation required 65

66 Profile quality issues Additional black inks to reduce visibility of dots Min K Max K 66

67 Profile quality issues Results different ink usage significant ink reduction 7C 6C Cyan min K 6 inks Cyan Magenta Magenta Yellow Yellow Heavy Black Black Light Black Light Cyan Light Cyan Light Magenta Light Magenta max K 7 inks 67

68 References The Reproduction of Color, R. W. G. Hunt Color Science, Concepts and Methods, Quantitative Data and Formulae, Wyszecki and Stiles Color technology for electronic imaging devices, Henry Kang. Digital color halftoning, Henry Kang. Digital Color Management: Encoding Solutions, Giorgianni and Madden. Colour Engineering, edited by Phil Green and Lindsay MacDonald. Specification ICC.1: (Profile version ), Image technology colour management - Architecture, profile format, and data structure. 68