IMCO: Measuring Color

Recap

• CIE Illuminant D65
• Reflectance
• etc.

Video time

Konica Minolta Sensing

How can we measure colors ?

• Spectrophotometer
  • Response through the entire visible spectrum
  • Relatively small areas (few
    $c{m}^2$) - Resolution is 1 point
  • "Falt surfaces"
• RGB Camera
  • Response in 3 wavelengths (Red, Green Blue)
  • Large areas - High Spatial Resolution (
    $<50$MPixels)
  • Any kind of surfaces
• Hyperspectral camera
  • Response throughout the entier visible spectrum (and more)
  • Large areas - Low Spatial Resolution (
    $\le 2$MPixels)
  • Any kind of surfaces

When NOT to Measure color

• Using instrument to measure color and compute differences objectively is not always needed
• For example: A company has a corporate color (possible
  ${}^{TM}$)
  • Tour de France: Pantone 123C
  • Veuve Cliequot: Pantone 137C
  • Louboutin: Pantone 18.1663TP
• Products carrying the color are sold; however they are manufactured by different providers

Judging by visual assessment

• Need consistent lightning
• Need consistent viewing
• Need to Check for Metamers

Use a light booth !

• Sufficient when there are few standard samples to be matched
• Sufficient when tolerance is judged visually by color experts
• Requires all manufacturers to have a physical copy of the standard, and to have the same hardware
• Because there are no measurements, we don't know to adjust color workflow in case we need to match a color

Measuring with Spectrophotometers

Remember Light interaction

Spectrometer can measure reflectance and transmittance (specular and/or diffuse)

Time for another video

What is a Spectrophotometer?

Light Reflection vs Material

• Matte
  • Light is reflected in all directions equally
• Semiglossy
  • Light is reflected in all direction but a small part is reflected orthogonal to the incident angle
• Glossy
  • Light is reflected in all directions but a big part is reflected orthogonal to the incident angle

Spectrophotometers: In a Nutshell

• Spectral reflectance
  • The ratio of reflected light (
    $r$) to the incident light (
    $i$) under specific geometric conditions

$$R_{\lambda }=\frac{{\varphi }_{\lambda }^r}{{\varphi }_{\lambda }^i}$$

• Spectral transmittance
• • The ratio of transmitted light (
    $t$) to the incident light (
    $i$) under specific geometric conditions

$$T_{\lambda }=\frac{{\varphi }_{\lambda }^t}{{\varphi }_{\lambda }^i}$$

• All measuring instrument need to be calibrated
  • using White Tile made from Spectralon

Spectrophotometers: reflectances ?

Interlude: fluoresence

Fluroescence can create colors we don't see

• Use an instrument called a Bispectrometer to measure it

Donaldson matrix obtained from a green sample emitting a more satured green light

Colorimeter vs Spectrophotometers

• Colorimeters are used generally to calibrate screens
• They mimic the way our eyes perceive color

They measure reflectance in 3 wavelengths (R, G, B)
They do not provide a spectral response

Spectrophotometers

Types

• Bidirectionnal
  • Non-structured and flat surfaces (paper, plastics)
• Sphere
  • Structured and glossy surfaces (textiles, metallic)

SPIN vs SPEX

• SPIN Specular Included (gloss is accounted for)
  • Color is measured independent of the sample's gloss or surface texture
• SPEX Specular Excluded

Specifications

Example: Automotive interior plaque (items produced using different materials)

• SPIN: looks at the material independant of surface texture
• SPEX: values which depend on gloss and surface conditions

Different spectro models

Specifications

Choose depending on what you need

	X-RIte i1Pro 2	X-RITE Ci62	Barbieri LFP qb
Measurment geometry	${45}^o$ a: $0$ (ring illumination)	$di:8^o$	${45}^o$c: $0$ (circumferential)
Light source	Gas filled tungsten lamp and UV LED	Gas-filled tungsten lamp	3 point circle, 7-LED chip

Geometry

Reflectance of a semi-glossy object

• $di:8^o$

A high gloss sample with the same pigmentation is visually judged darker by the eye when compared to a matte sample

• ${45}^o:0$: measure that color difference
• $di:8^o$ measure the same color in both cases
• ${45}^o:0^o$ simulates normal behavior
  • e.g. when we read a magazine

Aperture

	X-RIte i1Pro 2	X-RITE Ci62	Barbieri LFP qb
Measurment aperture	$4.5mm$	$4$ or $8mm$	$2$, $6$ and $8mm$

Small aperture

• Measures quickly
• may miss relevant info

Large aperture

• more accurate
• measurement takes longer
• needs larger sample

Conditions

	X-RIte i1Pro 2	X-RITE Ci62	Barbieri LFP qb
Measurment conditions	M0, M1, M2	N/A	M0, M1, M2, M3

• M0
  • legacy measurement (tungsten lamp, no standardization of UV content in illuminat, UV strength changes through time)
• M1
  • Spectral distribution of illuminant
• M2
  • UV is excluded
• M3
  • Polarized light

Measurement conditions impact the color

Spectral range

	X-RIte i1Pro 2	X-RITE Ci62	Barbieri LFP qb
Spectral range	$380-730nm$	$700-400nm$	$380-750nm$

Repeatability

	X-RIte i1Pro 2	X-RITE Ci62	Barbieri LFP qb
Short term repeatability	$0.1$ $\mathrm{\Delta }{E}_{94}$	$0.05$ $\mathrm{\Delta }{E}_{ab}$	$0.05$ $\mathrm{\Delta }{E}_{00}$

2 different i1Pro 2 spectro

• 10 measurements of the same object were taken for each instrument
• $\mathrm{\Delta }E$ between first and other 9 measurements were computed for each instrument

	X-RIte i1Pro 2	X-RITE Ci62	Barbieri LFP qb
Inter-instrument agreement	Average $0.4$ $\mathrm{\Delta }{E}_{94}$ Max $1.0$ $\mathrm{\Delta }{E}_{94}$	Average $0.4$ $\mathrm{\Delta }{E}_{ab}$ Max $1.0$ $\mathrm{\Delta }{E}_{ab}$	Average $0.4$ $\mathrm{\Delta }{E}_{00}$ Max $1.0$ $\mathrm{\Delta }{E}_{00}$

Transmittance Measurement

• When we need transmittance ?
  • Light Filters
  • Printed Ads
  • Food Inspection

Inter-instrument agreement

• Compared measurements of 16 samples used for printing

Recap

• Many different (standardized) methods to measure Reflectance (and Transmittance)
• Unfortunately, measured Reflectance/Transmittance is not unique as it depends on the instrument you sued to measure it
• Type of instrument to used depends on what you want to measure, and how frequent you want to measure
• Only measurements tales under the same conditions can be truly compared. Therefore, it is necessary to note the following information in a color measurement report:
  • Color instrument (geometry, aperture, measurement condition)
  • Illuminant/observer standards, if you give
    $L\times a\times b$ values

Future trends: beyond color

Visual appearance of materials

• Reflection
• Transmission
• Absorbance

BRDF Measurement

Bi-directional Reflectance Distribution Function (BRDF) gives a more complete characterization of light interaction with the surface

We measure how light reflects in all directions

• BRDF allows characterizing the surface appearance at a microscopic level (used in Computer Graphics to render objects)
• Measurable with Goniophotometers

How to measure BRDF faster and cheaper ?

Sources

• Les boules
• Litteralement tout le cours
• Spectrophotometer
• Spectral measurment X-RITE

Metamerism

What's that ?

metamerism is a perceived matching of colors with different (nonmatching) spectral power distributions.

Most important types

• Illuminant Metamerism
  • Different spectral characteristic and
    • same color when viewed under one light
    • different color when view under another light
• Observer Metamersim
  • Different spectral characterisic and
    • same color when viewed by one observer
    • different color when view by another observer

Examples:

Car industry

Source

Other

Source

Metamerism vs Color Inconstancy

• Color inconstancy: A single object changing color with changes in the color of the illumination
• Metameric pair: Two objects having color inconstancy

Recap

• Metamerism is an effect we need to consider if a pair of objects will be viewed under more than one type of illuminant
• In the printing industry, neutral (grayscale) colors are more susceptible to illuminant metamerism as a mix of inks is used
• In the case of displays, illuminant metamerism is not a problem as they create their own light