SlideShare a Scribd company logo
RECIPE PREDICTION
IN
PIGMENTED APPLICATION
1
UDAY KALE
DATACOLOR
MUMBAI
Company History
 Since 2009 Datacolor is a stand alone, pure play, publically
traded company on the Swiss stock exchange.
 Before 2009, Datacolor was part of the Swiss based Eichhof
Holding.
 The sale of 2 of the 3 companies owned by the Holding allowed
Datacolor to be well positioned financially to pursue its growth
strategy.
2
A Global Business
United States
United States
China (Suzhou)
Switzerland
United Kingdom
France
India
Taiwan
Italy
Germany
Belgium
Hong Kong
China (Shanghai)
Corporate functions
Corporate functions
Sales and Service office
 Representation – 65
countries
 Employees – 300 in 25
counties 3
Brands that rely on Datacolor
The names, trademarks, service marks, logos and icons of the above companies are properties of the respective companies.
4
Datacolor Product Portfolio
Kubelka - Munk
6
Behaviour of light
Incident light
Specular %R
Diffuse %R (=color)
Internal %R
Absorbed light
Complex Subtractive Mixing
Turbid-Media Theory
• Kubelka-Munk
• K/S is proportional to concentration
• In the opaque case: K/S = ((1-r)*2)/2R
K/Smix = (c1K1 + c2K2 + c3K3 ...) / (c1S1 + c2S2 +
c3S3 ...)
• The full treatment:
• Others
Both Light Absorption and Light
Scattering are considered.
Kubelka Munk Relation
 When light passes through a film ( Plastic or Paint) it gets absorbed,
Reflected, Scattered or Transmitted.
 The passes of light through such film is very complex and the
mathematical description for the phenomenon is known as “ Turbid
Medium Theory”.
 Kubelka and Monk simplified the Turbid Media Theory.
 When the light passes through the pigment layer, the changes take
place in downward and upward direction (Two fluxes one in downward
and other in upward direction)
 This can be calculated using simple mathematical relation from their
scattering and absorption coefficients. ( K and S)
 Kubelka Monk equation co relates reflectance of an object with
Scattering and Absorption coefficients.
K/S = ( 1 –R )2 / 2R
Kubelka Munk Relation
 The most important feature of K M Relation is it is additive. i.e. K/S is
additive while Reflectance is non additive.
 K/S function is linearly proportional with concentration of colorants.
K/S = α c where α is a constant of proportionality.
 Pigments disperse in polymers have both the properties of absorption
and scattering hence it is necessary to calculate both the coefficients.
 Such application is termed as Two constant Application.
 In case of Textile dyes particles are in dissolved state hence not
capable of scattering. In this case the scattering is contributed by
substrate only.
 We need to calculate only absorption coefficient ( K ) for the colorant
 Such a application is called one constant application.
Kubelka - Munk
10
Kubelka-Munk Function
K/S =
2R
(1-R)2
Matching
11
Matching colors makes it
possible to determine which
colorants are needed and
which are their optimal
quantities to reproduce the
color.
Matching - General
12
To be able to do a formulation,
we need to define the following things mathematical:
colorants standard
Mathematical
relation
Matching - General
13
+ =
39.20% + 52.55% = 32.87%
Reflection doesn’t fit the additive principle
Kubelka - Munk
Solution for the
problem:
KUBELKA - MUNK
Function
Kubelka - Munk
Application of the Kubelka – Munk function
15
K/S = 0.6855
K/S = 0.2142
K/S = 0.4713
K/S mixture = 0.6855
The K/S function is additive.
Matching - General
16
Formulate the
color of the
standard is the
same as
reproduce the
spectral curve
by using a
combination of
colorants
Colorant set
Define
the behaviour of the pigments
in a colorant set
so we can match the target.
17
Matching
The quality of the matches is
directly related to the quality
of your primary samples!
18
Colorant set
 One constant function (transparent)
 Two constant function
• Relative (opaque)
• Absolute (translucent)
19
One constant function
One constant function
20
Transparent colorant set
One constant function
Used to match:
 Transparent or translucent samples that not include a white
pigment. If white is used, it is used in a fixed amount.
 Example: woodstains, transparent inks, textiles,...
21
One constant function
22
Substrate
Color layer
Due to the nature of the colorant, the
colorant absorbs but don’t scatter light.
Two constant function
Two constant function
23
Opaque and translucent
colorant set
Two constant function
24
Substrate
Due to the nature of the colorant, the colorant
selectively absorbs and scatters light.
Two constant function
Separated K & S values
are calculated
for the colorant
25
Two constant relative
 Used for opaque applications
 Substrate is not important
 Film thickness is not important
26
Two constant absolute
 Used for opaque and non opaque samples
 Substrate is important
 Film thickness is important
27
Saunderson Correction
The Kubelka-Munk color model doesn’t take into account the reflection losses
at the sample boundaries.
There are two losses that Saunderson takes into account:
• Internal loss (internal %R)
• External loss (surface %R – diffuse and specular)
28
Saunderson Correction
29
Behaviour of light
Incident light
Surface Specular %R
Diffuse %R (=color)
Internal %R
Surface Diffuse %R
Saunderson Correction
Note:
When we create a colorant set, we modify the measured
reflectance before calculating K and S values of the colorants.
30
Rc =
Rm - ke
1 – ke – ki * (1 – Rm)
Matching
31
Substrate isn’t important
Film thickness isn’t important
Will perform an opaque match
Fix the pigment loading if wanted
Opaque Matching
Matching
32
Fixed pigment loading
Opaque Matching
Matching
33
Opaque Matching
Defined pigment
load is reached
Matching
34
Substrate is important
Film thickness is important
You can do a CR match
You can do a fixed % load match
Translucent Matching
Matching
35
Translucent - CR Match
No pigment load
defined
Defined CR
Matching
36
Requested CR
is achieved
Free pigment
loading
Translucent - CR Match
Matching
37
No opacity defined
Fixed pigment
loading
Translucent - Fixed % Load match
Matching
38
Translucent - Fixed % Load match
Calculated opacity
Requested
pigment loading
is reached
Matching
39
Matching programs may calculate many recipes for a new shade
For each recipe the computer gives:
Color difference
Metamerism
Price
...
The colorist then selects the most suitable recipe considering:
Fastness properties
Compatibilty
...
Matching
Why don’t my predictions
come out right first time?
40
Matching
1. Theory assumes colorants behave the same in
combination as individually in the database. Does not
account for interaction.
2. Pigments can differ form the ones used to create the
colorant set
3. Reproduciability
4. Errors in sample preparation
5. Appearance
41
Error in database Error in lab sampling
DE* = 0.5
DE* = 0.5
New Shade Recipe Prediction Lab Sampling
What is Smartmatch?
 Smartmatch quantifies the interaction between individual
pigments and substrates (Axis Smartmatch)
 Smartmatch is a self learning matchprediction system
based on practical experience (Palette Smartmatch)
 Smartmatch is a function to correct matching theory
(single constant, two constant or multi constant theory
(combined method))
Smart Calibrator selects
the best optical model
(combined vs pairs)
Smart Calibrator
Performance factors
 Eveluate how the colorants perform in the current batch
 Based on a comparison between the amount of colorant you put
into the formula and the amount of colorant the system ‘sees’ in
the formula.
44
Performance factors
The colorant can perform 3 ways:
• PF = 1.0
Colorant is performing exactly as expected
• PF > 1.0
Colorant is performing stronger than expected
• PF < 1.0
Colorant is performing weaker than expected
45
Gloss Compensation
The gloss problem
46
The lack of agreement between visual
and instrumental evaluations (with
integrating sphere spectro’s) of color
samples that have different gloss
Gloss Compensation
When to use gloss compensation?
Batch (or product) gloss
is differentthan
the gloss of the standard.
Biggest effect for dark colors.
47
Gloss compensation
offers the ability to more accurately match, correct, and control color,
even as the gloss of the standards and the product differ
 New Feature
 Match and correct to an
offset value of the original
target in (CIE L*a*b*, D65)
 Utilize common target for
color appearance matching of
samples with various gloss
and texture.
CIE L*a*b* Offset
Uday Kale - Recipe Prediction for Non Textile Application - Datacolor_2.pdf

More Related Content

PDF
LCD Characterization Report
PDF
Textile application of the color sensitivity of a dye mixture
PPT
Paper Con 2011 Presentation Continuous Coloration Of Coatings For Coated Pa...
PDF
07 dimensionality reduction
PPT
8. computer colour matching steps to predict concentration
PPTX
What Color is Solid State Lighting - Panel Discussion
PDF
Linear Programming Module- A Conceptual Framework
PPTX
ROLE OF COMPUTER IN TEXTILE
LCD Characterization Report
Textile application of the color sensitivity of a dye mixture
Paper Con 2011 Presentation Continuous Coloration Of Coatings For Coated Pa...
07 dimensionality reduction
8. computer colour matching steps to predict concentration
What Color is Solid State Lighting - Panel Discussion
Linear Programming Module- A Conceptual Framework
ROLE OF COMPUTER IN TEXTILE

Similar to Uday Kale - Recipe Prediction for Non Textile Application - Datacolor_2.pdf (17)

PPTX
Understanding Black Box Models with Shapley Values
PPTX
Using SHAP to Understand Black Box Models
PPT
LinearProgramming-Graphicalnethod.ppt
PDF
IRJET- Crowd Density Estimation using Image Processing
PDF
Ph.D. Thesis Presentation: A Study of Priors and Algorithms for Signal Recove...
PPT
linear programming model formulation and graphical solution
PDF
linearprogramming.pdf
PPTX
Linear programming - Model formulation, Graphical Method
PPTX
QA CHAPTER II.pptx
PDF
KnowledgeFromDataAtScaleProject
PPTX
Image processing sw & hw
PPT
Colour technology
PPT
PDF
Photohsop movie 06
PPTX
Balaji-opt-lecture3-sp13.pptx
PPTX
DSP presentation_latest
PDF
Using A Application For A Desktop Application
Understanding Black Box Models with Shapley Values
Using SHAP to Understand Black Box Models
LinearProgramming-Graphicalnethod.ppt
IRJET- Crowd Density Estimation using Image Processing
Ph.D. Thesis Presentation: A Study of Priors and Algorithms for Signal Recove...
linear programming model formulation and graphical solution
linearprogramming.pdf
Linear programming - Model formulation, Graphical Method
QA CHAPTER II.pptx
KnowledgeFromDataAtScaleProject
Image processing sw & hw
Colour technology
Photohsop movie 06
Balaji-opt-lecture3-sp13.pptx
DSP presentation_latest
Using A Application For A Desktop Application
Ad

Recently uploaded (20)

PPT
Miokarditis (Inflamasi pada Otot Jantung)
PDF
168300704-gasification-ppt.pdfhghhhsjsjhsuxush
PPTX
IB Computer Science - Internal Assessment.pptx
PDF
Galatica Smart Energy Infrastructure Startup Pitch Deck
PPTX
Market Analysis -202507- Wind-Solar+Hybrid+Street+Lights+for+the+North+Amer...
PPTX
DISORDERS OF THE LIVER, GALLBLADDER AND PANCREASE (1).pptx
PDF
annual-report-2024-2025 original latest.
PPTX
Data_Analytics_and_PowerBI_Presentation.pptx
PPTX
IBA_Chapter_11_Slides_Final_Accessible.pptx
PPTX
STUDY DESIGN details- Lt Col Maksud (21).pptx
PPTX
Introduction to Knowledge Engineering Part 1
PPT
Quality review (1)_presentation of this 21
PPTX
The THESIS FINAL-DEFENSE-PRESENTATION.pptx
PPTX
MODULE 8 - DISASTER risk PREPAREDNESS.pptx
PPTX
Business Ppt On Nestle.pptx huunnnhhgfvu
PPTX
Microsoft-Fabric-Unifying-Analytics-for-the-Modern-Enterprise Solution.pptx
PPTX
Acceptance and paychological effects of mandatory extra coach I classes.pptx
PPTX
Introduction to Basics of Ethical Hacking and Penetration Testing -Unit No. 1...
PDF
Introduction to the R Programming Language
Miokarditis (Inflamasi pada Otot Jantung)
168300704-gasification-ppt.pdfhghhhsjsjhsuxush
IB Computer Science - Internal Assessment.pptx
Galatica Smart Energy Infrastructure Startup Pitch Deck
Market Analysis -202507- Wind-Solar+Hybrid+Street+Lights+for+the+North+Amer...
DISORDERS OF THE LIVER, GALLBLADDER AND PANCREASE (1).pptx
annual-report-2024-2025 original latest.
Data_Analytics_and_PowerBI_Presentation.pptx
IBA_Chapter_11_Slides_Final_Accessible.pptx
STUDY DESIGN details- Lt Col Maksud (21).pptx
Introduction to Knowledge Engineering Part 1
Quality review (1)_presentation of this 21
The THESIS FINAL-DEFENSE-PRESENTATION.pptx
MODULE 8 - DISASTER risk PREPAREDNESS.pptx
Business Ppt On Nestle.pptx huunnnhhgfvu
Microsoft-Fabric-Unifying-Analytics-for-the-Modern-Enterprise Solution.pptx
Acceptance and paychological effects of mandatory extra coach I classes.pptx
Introduction to Basics of Ethical Hacking and Penetration Testing -Unit No. 1...
Introduction to the R Programming Language
Ad

Uday Kale - Recipe Prediction for Non Textile Application - Datacolor_2.pdf

  • 2. Company History  Since 2009 Datacolor is a stand alone, pure play, publically traded company on the Swiss stock exchange.  Before 2009, Datacolor was part of the Swiss based Eichhof Holding.  The sale of 2 of the 3 companies owned by the Holding allowed Datacolor to be well positioned financially to pursue its growth strategy. 2
  • 3. A Global Business United States United States China (Suzhou) Switzerland United Kingdom France India Taiwan Italy Germany Belgium Hong Kong China (Shanghai) Corporate functions Corporate functions Sales and Service office  Representation – 65 countries  Employees – 300 in 25 counties 3
  • 4. Brands that rely on Datacolor The names, trademarks, service marks, logos and icons of the above companies are properties of the respective companies. 4
  • 6. Kubelka - Munk 6 Behaviour of light Incident light Specular %R Diffuse %R (=color) Internal %R Absorbed light
  • 7. Complex Subtractive Mixing Turbid-Media Theory • Kubelka-Munk • K/S is proportional to concentration • In the opaque case: K/S = ((1-r)*2)/2R K/Smix = (c1K1 + c2K2 + c3K3 ...) / (c1S1 + c2S2 + c3S3 ...) • The full treatment: • Others Both Light Absorption and Light Scattering are considered.
  • 8. Kubelka Munk Relation  When light passes through a film ( Plastic or Paint) it gets absorbed, Reflected, Scattered or Transmitted.  The passes of light through such film is very complex and the mathematical description for the phenomenon is known as “ Turbid Medium Theory”.  Kubelka and Monk simplified the Turbid Media Theory.  When the light passes through the pigment layer, the changes take place in downward and upward direction (Two fluxes one in downward and other in upward direction)  This can be calculated using simple mathematical relation from their scattering and absorption coefficients. ( K and S)  Kubelka Monk equation co relates reflectance of an object with Scattering and Absorption coefficients. K/S = ( 1 –R )2 / 2R
  • 9. Kubelka Munk Relation  The most important feature of K M Relation is it is additive. i.e. K/S is additive while Reflectance is non additive.  K/S function is linearly proportional with concentration of colorants. K/S = α c where α is a constant of proportionality.  Pigments disperse in polymers have both the properties of absorption and scattering hence it is necessary to calculate both the coefficients.  Such application is termed as Two constant Application.  In case of Textile dyes particles are in dissolved state hence not capable of scattering. In this case the scattering is contributed by substrate only.  We need to calculate only absorption coefficient ( K ) for the colorant  Such a application is called one constant application.
  • 10. Kubelka - Munk 10 Kubelka-Munk Function K/S = 2R (1-R)2
  • 11. Matching 11 Matching colors makes it possible to determine which colorants are needed and which are their optimal quantities to reproduce the color.
  • 12. Matching - General 12 To be able to do a formulation, we need to define the following things mathematical: colorants standard Mathematical relation
  • 13. Matching - General 13 + = 39.20% + 52.55% = 32.87% Reflection doesn’t fit the additive principle
  • 14. Kubelka - Munk Solution for the problem: KUBELKA - MUNK Function
  • 15. Kubelka - Munk Application of the Kubelka – Munk function 15 K/S = 0.6855 K/S = 0.2142 K/S = 0.4713 K/S mixture = 0.6855 The K/S function is additive.
  • 16. Matching - General 16 Formulate the color of the standard is the same as reproduce the spectral curve by using a combination of colorants
  • 17. Colorant set Define the behaviour of the pigments in a colorant set so we can match the target. 17
  • 18. Matching The quality of the matches is directly related to the quality of your primary samples! 18
  • 19. Colorant set  One constant function (transparent)  Two constant function • Relative (opaque) • Absolute (translucent) 19
  • 20. One constant function One constant function 20 Transparent colorant set
  • 21. One constant function Used to match:  Transparent or translucent samples that not include a white pigment. If white is used, it is used in a fixed amount.  Example: woodstains, transparent inks, textiles,... 21
  • 22. One constant function 22 Substrate Color layer Due to the nature of the colorant, the colorant absorbs but don’t scatter light.
  • 23. Two constant function Two constant function 23 Opaque and translucent colorant set
  • 24. Two constant function 24 Substrate Due to the nature of the colorant, the colorant selectively absorbs and scatters light.
  • 25. Two constant function Separated K & S values are calculated for the colorant 25
  • 26. Two constant relative  Used for opaque applications  Substrate is not important  Film thickness is not important 26
  • 27. Two constant absolute  Used for opaque and non opaque samples  Substrate is important  Film thickness is important 27
  • 28. Saunderson Correction The Kubelka-Munk color model doesn’t take into account the reflection losses at the sample boundaries. There are two losses that Saunderson takes into account: • Internal loss (internal %R) • External loss (surface %R – diffuse and specular) 28
  • 29. Saunderson Correction 29 Behaviour of light Incident light Surface Specular %R Diffuse %R (=color) Internal %R Surface Diffuse %R
  • 30. Saunderson Correction Note: When we create a colorant set, we modify the measured reflectance before calculating K and S values of the colorants. 30 Rc = Rm - ke 1 – ke – ki * (1 – Rm)
  • 31. Matching 31 Substrate isn’t important Film thickness isn’t important Will perform an opaque match Fix the pigment loading if wanted Opaque Matching
  • 34. Matching 34 Substrate is important Film thickness is important You can do a CR match You can do a fixed % load match Translucent Matching
  • 35. Matching 35 Translucent - CR Match No pigment load defined Defined CR
  • 36. Matching 36 Requested CR is achieved Free pigment loading Translucent - CR Match
  • 37. Matching 37 No opacity defined Fixed pigment loading Translucent - Fixed % Load match
  • 38. Matching 38 Translucent - Fixed % Load match Calculated opacity Requested pigment loading is reached
  • 39. Matching 39 Matching programs may calculate many recipes for a new shade For each recipe the computer gives: Color difference Metamerism Price ... The colorist then selects the most suitable recipe considering: Fastness properties Compatibilty ...
  • 40. Matching Why don’t my predictions come out right first time? 40
  • 41. Matching 1. Theory assumes colorants behave the same in combination as individually in the database. Does not account for interaction. 2. Pigments can differ form the ones used to create the colorant set 3. Reproduciability 4. Errors in sample preparation 5. Appearance 41 Error in database Error in lab sampling DE* = 0.5 DE* = 0.5 New Shade Recipe Prediction Lab Sampling
  • 42. What is Smartmatch?  Smartmatch quantifies the interaction between individual pigments and substrates (Axis Smartmatch)  Smartmatch is a self learning matchprediction system based on practical experience (Palette Smartmatch)  Smartmatch is a function to correct matching theory (single constant, two constant or multi constant theory (combined method))
  • 43. Smart Calibrator selects the best optical model (combined vs pairs) Smart Calibrator
  • 44. Performance factors  Eveluate how the colorants perform in the current batch  Based on a comparison between the amount of colorant you put into the formula and the amount of colorant the system ‘sees’ in the formula. 44
  • 45. Performance factors The colorant can perform 3 ways: • PF = 1.0 Colorant is performing exactly as expected • PF > 1.0 Colorant is performing stronger than expected • PF < 1.0 Colorant is performing weaker than expected 45
  • 46. Gloss Compensation The gloss problem 46 The lack of agreement between visual and instrumental evaluations (with integrating sphere spectro’s) of color samples that have different gloss
  • 47. Gloss Compensation When to use gloss compensation? Batch (or product) gloss is differentthan the gloss of the standard. Biggest effect for dark colors. 47 Gloss compensation offers the ability to more accurately match, correct, and control color, even as the gloss of the standards and the product differ
  • 48.  New Feature  Match and correct to an offset value of the original target in (CIE L*a*b*, D65)  Utilize common target for color appearance matching of samples with various gloss and texture. CIE L*a*b* Offset