Hataka Early Luftwaffe Paint Set - Battle of Britain & September 1939 (HTK-AS02) - quality of color matches

[Casey]

Here is yet another episode of the review of the quality of color matches from paint vendors - this time, I've worked on another Polish paint set to verify their claims of the historical accuracy of color matches.

 

Link to the paint itself: https://www.hataka-hobby.com/HTK-AS02-Early-Luftwaffe-Paint-Set

 

As a comparison reference source, I've used:

• Luftwaffe Camouflage and Markings 1933-1945 vol I K.A. Merrick and Jurgen Kiroff - https://www.scalemates.com/books/luftwaffe-camouflage-and-markings-1933-1945-volume-1-classic-colours-luftwaffe-camouflage-and-markings-vol1-k-a-merrick-juergen-kiroff--102687
• Luftwaffe Colours 1935-1945 - Michael Ullmann - https://www.scalemates.com/books/luftwaffe-colours-michael-ullmann--102400

 

This time the comparison is made of three parts:

1. Visual comparison: flatbed scanner scans
2. Technical comparison: spectrophotometer and gloss meter readouts + some comments from my side

 

Drawdowns

All paints were mixed on a vortex mixer for a couple of minutes. Then I prepared drawdowns using BYK drawdown charts and a 10mil drawdown bar using unthinned paint, then let them dry for 72h. Due to the thickness of the paint layer, most of the samples form a concave shape on the drawdown charts; you can see it as a dark 'outline' around the color patch, which is an effect of scans of samples when they are pressed to the glass surface. All Hataka paints seem to be of SATIN finish where reference is full matt - which means visually they seem darker

 

Visual comparison

To facilitate visual comparison, I've scanned primary reference sources and drawdowns on a scanner with color correction turned off - it means the colors are unmodified by software and are as seen as my scanner device. Then I've cut&pasted the corresponding reference tiles to the drawdowns to let you see them close. I've left un-modified drawdowns on the 'spoilers' section.
 
Please note the colors seen by the flatbed scanner are not the same as seen by the spectrophotometer, but the difference between colors should be directly comparable. Also, be aware that the gloss finish of the Hataka paint is different than the reference samples.

 

Here is a scan of reference source:

 

And here are drawdowns with reference sources overlayed:

Spoiler

 

 Technical comparison

 

Below are spectrophotometer scans, compared with the references.

 

Explanation of the column names is in the spoiler section below:

Spoiler

Name: Paint name and code

Remarks - contain my Gloss Level measurements from gloss meter. Gloss Units (GU), of gloss meters is a scaling based on a highly polished reference black glass standard with a defined refractive index having a specular reflectance of 100GU at the specified angle. My Gloss Meter is 60 degree. More information: https://en.wikipedia.org/wiki/Glossmeter
L*, a*, b* - It is a color measurement defined by CIE standard; it expresses color as three values: L* for perceptual lightness and a* and b* for the four unique colors of human vision: red, green, blue, and yellow. More information - https://en.wikipedia.org/wiki/CIELAB_color_space

R, G, B - Conversion of the above value to the sRGB color space
Color Sim - A color swatch of RGB above
H, V, C (Munsell) - A color coordinate in the Munsell color system. CIELAB is a new digital version of this color system, but I have a Munsell color book, which lets me use the above coordinates as a color reference. More information - https://en.wikipedia.org/wiki/Munsell_color_system

dE*00 - Color difference. This number represents how the colors are visually different for an average human. The smaller the number, the closer the color match is. A value of 1.0 represents a just-noticeable difference. The value of 2.0 is different but acceptable in a lot of situations. More information - https://en.wikipedia.org/wiki/Color_difference and https://en.wikipedia.org/wiki/Just-noticeable_difference

Metamerism - Used for color comparison, shows how color differs in different light conditions. Some colors that match in daylight may look totally different in a different light.
Test mode - Describes specular (reflections) settings used by my spectrophotometer; SCI means Specular Component Included. Reads done in this setting incorporate all reflections to the measurement, which lets me ignore the color finish to perform readouts and evaluate the "true" color. SCI readouts of glossy and matt versions of the same paint will give the same readouts. More information - https://sensing.konicaminolta.us/us/blog/specular-component-included-sci-vs-specular-component-excluded-sce/
Light Source - Describes the simulated light conditions by my device and which standard observer is used for color calculations. D65/10 means D65 (Daylight) Illuminant; more info: https://en.wikipedia.org/wiki/Illuminant_D65. Number 10 describes an 'observer,' which is a model of a human vision used to analyze color. The 10-degree standard observer is the standard type used in spectrophotometry for evaluation and color formulation. More information - https://sensing.konicaminolta.us/us/blog/understanding-standard-observers-in-color-measurement/
UV Settings - UV400 means that in my spectrophotometer, I have a physical 400nm UV filter, which 'cuts out the ultraviolet component from readouts.

P/F dE*00 - Pass/Fail mark for comparison using dE*00 difference. Industry standard for paint matching assumes 2.0 as a threshold and this is same I've used for the P/F mark there, which is very strict but the vendor claims the most accurate colors there...

 

HTK-A015 - Black Green (RLM70)

Not a bad match for the Merrick publication, almost perfect for Ullmann

 

HTK-A017 - Dark Green (RLM71)

Not really bad but somewhat metameric matches for both publications, both are below 3.0DE

 

HTK-A023 - Grey (RLM02)

Very good match for Merrick and Kiroff publication.

 

HTK-A029 - Light Blue (RLM65/78)

Calling paint a match for two different colors is not not setting the best expectations, and this color closest (but far from being good) match is the later war RLM65 from Ullmann publication.

 

Edited February 1, 2023 by Casey