Food colour consistency analyser (inline colour measurement)

01

Choose the instrument type

ASTM E1347 covers instrumental measurement of colour coordinates and colour differences with a tristimulus (filter) colorimeter, in reflected or transmitted light, using hemispherical (integrating sphere) or bidirectional (45:0 or 0:45) geometry. Because tristimulus colorimeters have limited absolute accuracy, it requires that, when colour coordinates are needed, the instrument be standardised against a product standard with similar spectral and geometric characteristics to the specimen. [1]

Tristimulus colorimeters give no reflectance or transmittance curves, so ASTM states they cannot reveal metamerism, and colour differences may only be measured between specimens with similar spectral and geometric characteristics. The method suits non-fluorescent, planar object-colour specimens. It does not cover spectrocolorimeters. ASTM adds that colorimeters with hemispherical geometry have a chromatic integrating-sphere error when standardised with a white standard. [1]

Where spectral data are measured, ASTM E308 provides the values and procedures to compute CIE tristimulus values X, Y, Z and chromaticity coordinates for the CIE 1931 and 1964 standard observers, with standard values for six CIE standard illuminants and three recommended fluorescent illuminants. [2]

Konica Minolta's colour booklet shows why spectral data matter for metamerism. In its example, two specimens read L* 50.93, a* 4.54, b* -5.12 under Standard Illuminant D65 (ΔE*ab = 0), but under Standard Illuminant A they read differently and ΔE*ab = 8.71. It states: "Tristimulus colorimeters can generally take measurements under only Standard Illuminant C and Standard Illuminant D65", which have similar spectral power distributions, so they cannot be used to measure metamerism; a spectrophotometer holds data for a wide range of illuminants and can. [5]

The same booklet describes two ways of handling gloss. SCE (specular component excluded) leaves out the specular reflection; SCI (specular component included) completes the sphere with a specular plug and measures specular plus diffuse reflection. HunterLab's SpectraTrend HT inline sensor, reviewed below, lists its specular component as excluded. [5] [7]

Give the supplier the products and their surface (flat, irregular, breaded, glossy), whether measurement is inline or on samples, the line speed and product spacing, the colour scale and illuminant/observer required, the reference standards you use, the tolerance to be held, and the data and alarm outputs needed.

02

Express colour and set the tolerance in L*a*b*

Konica Minolta describes the L*a*b* colour space (CIELAB) as one of the uniform colour spaces defined by CIE in 1976. L* is lightness; +a* is the red direction and -a* green; +b* is yellow and -b* blue. The colour difference ΔE*ab is defined as the square root of (ΔL*)² + (Δa*)² + (Δb*)², a single number that gives the size of the difference but not its direction. In the booklet's apple example, ΔL* = +4.03, Δa* = -3.05 and Δb* = +1.04 give ΔE*ab = 5.16. [5]

Konica Minolta's tolerance guide lists four ways to set limits around the master colour: a box tolerance with ± limits on L*, a* and b*; ΔE*ab, which forms a sphere; CIE94, published by CIE in 1995 and based on automotive coating experiments on smooth surfaces; and CIE dE2000, which adds five compensations (hue rotation, neutral colours, lightness, chroma and hue) and forms an ellipsoid. It recommends assessing candidate samples visually in a light booth, measuring them, and setting the limits from the values furthest from the master that were still accepted. [6]

The booklet explains why ΔE*ab and the eye can disagree: the eye's discrimination thresholds are ellipses that lengthen in the saturation direction for highly saturated colours, while ΔE*ab treats every direction as a circle. The CIE 2000 formula was developed to close this gap. [5]

For browned foods, this means asking which tolerance shape the analyser applies and which formula (ΔE*, ΔE CMC or ΔE 2000) it reports. HunterLab lists ΔE*, ΔE, ΔE CMC and ΔE 2000 on its Aeros instrument and ΔE*, ΔE, ΔC*, ΔC and ΔE CMC on its SpectraTrend HT sensor. [9] [7]

03

Compare inline and at-line sensor specifications

Two published datasheets show what non-contact sensors mounted over a line specify. Both measure 400–700 nm at 10 nm intervals and report CIE L*a*b*. The figures below are each maker's own published data, not results on food. [7] [8]

X-Rite's sheet recommends validation every 30 days or 50,000 measurements, whichever comes first, and more frequent checks if the sensor is not kept clean. [8]

Editorial calculation: at its 1 s cycle time, 50,000 measurements take about 13.9 hours of continuous running, so on a line measuring without pause the measurement count arrives long before 30 days.

Note that the X-Rite sheet gives the enclosure as NEMA 4 / IP67 in its feature list and NEMA 4 / IP56 in its general specifications; confirm the rating in writing. The sheet is dated 2007. [8]

For irregular products sampled beside the line, HunterLab's Aeros measures a sample on a platter rotating at 12 RPM: 35 measurements per rotation, 177.25 cm² measured per 5-second rotation, sample height 65–140 mm. It quotes colorimetric repeatability of ΔE* < 0.025 maximum on a white tile and lists indices including Baking Contrast Units and SCAA. [9]

04

Relate instrument readings to grading standards

USDA grade standards for frozen french fried potatoes score colour at 30 of 100 points. Colour is judged as visually similar to the USDA Color Standards for Frozen French Fried Potatoes, which the standard says may also be used to classify the fry colour of other cooked products. [3]

The USDA inspection instructions describe the colour standards as colours numbered 0 to 4 that depict changes caused solely by frying. Sample units are designated Extra Light (No. 0) through Dark (No. 4), and sample units designated No. 4 Dark are not allowed in Grade A. The instructions note that some large users base raw potato contracts on these fry colours. [4]

Where a buyer or grade is defined against visual colour standards like these, ask the analyser supplier how its readings are correlated with the reference standards on your product. ASTM E1347 asks for the instrument to be standardised with a product standard of similar colour. [1]

05

Light source, spectrometer, geometry and white-tile standards inside an inline colour analyser

06

Geometry, colour scale, ΔE formula, working distance and validation interval to fix in the quotation

Each row pairs a quotation field with what ASTM, Konica Minolta, HunterLab or X-Rite publish about it, and the question to put to the supplier.

07

Send the product, line and standard data together

Paste it into your enquiry below or into an email to a maker, then replace every bracket.

Products: [type; surface; size]
Mode: [inline over conveyor / sample]
Line: [speed; width; spacing]
Colour scale: [CIE values; illuminant; observer]
Reference: [product standards; grade standard]
Tolerance: [target and limits]
Outputs: [data; alarms; reject signal]

See manufacturer records Send an enquiry through CNFX