Measuring Paint Colors Using Adobe Photo Shop PART II

[David Rapasi]

Paint colors can be measured in Adobe Photo Shop to get the Color Hue°, and the percentages of Saturation and Brightness, ( H:S:B: ).

The paint chips below are all quite close in Color Hue°.

The Color (HUE) is between 198° and 208°.

The SHADE of these color chips is what makes them appear so different.

The SHADE of a color is determined by its SATURATION (SAT), the percentage of COLOR HUE to the percentage of GRAY.

The BRIGHTNESS (BRT) of GRAY is determined by the percentage of WHITE to the percentage of BLACK.

Below is the Adobe Photo Shop Color Picker, measuring a Model Master 2012 Cobalt Blue color chip.

The vertical bar in the center is the Color Hue°.

I can scan color paint chips and measure them for comparison.

I also use the Adobe Color Picker to mix paint.

I spray my paint mix on plastic card stock and scan it into Photo Shop.

The paint chips are compared to see what changes need to be made in Hue, Saturation, and Brightness.

Color photos can also be scanned and the color hues measured.

For example;

When I get a good sample of the color hue from a color photo, I can compare these measurements to the Color chip measurements.

This process will give a good indication of what color to paint your model aircraft.

Scanning the Corsair primer I can find the hue of the different shades of color on the aircraft in different lighting.

The scans of the Corsair in sunlight and in shadow have a hue of 9°.

I mixed MM paint colors in Photo Shop and created this paint formula that has a hue of 9°.

26% MM 1708 FS 33538

37% MM 1705 FS 31136

37% MM 1749 FS 37038

This color can then be mixed with white, black, or any shade of gray, to get the shade of this color you want.

FS 30108 is a close match with a hue of 13°.

Dave

PS

My monitor is calibrated using the Adobe Photo Shop Gamma Wizard.

PART I

Edited February 27, 2012 by David Rapasi

[Rex]

that sure beats the old "Make paint chips on 3x5 cards method" we had to do in the Floquil days

David, can those paint matches you come up with, like that Salmon, be output as CIE or Munsell, or any other paint designation systems?

no need to list them or anything, I am just wondering if that is possible

[Stona]

Are you trying to replicate the photograph? That seems all that this is relevant to.You're not going to replicate the original colour on the aircraft this way.

Steve

[Phil marchese]

Dave. Thanks. I ve dabbled in Adabe for a decade and of course had color theory in art school. This is a good quick explanation. I add that the basic color spectrum is arrayed by wave length and the is a phyics to how wave length effects both refltivity and sun fading. Red fades fastest and blue travels farthest.

The bar follows the nature color spectrum with many triiary color cobination in the hue range. Primary R b y; secondary o g v.

Giving that RAF PRU is not a US or AN family member and is proportly slightly aqua is what prompted my question on hue # movement on the scale bar. I posted on tint. Vs shade in tips.

Of course all of this is pigment color theory were all = black & white is absence. Light color theory is reverse...like stage lighting.

[Phil marchese]

It is the VALUE and INTESITY (pureness of hue0 in proper terms that effects the samoles in the 190 -210 range. Shade is the darkenig ; TINT is the brighting. When you shade and complementry colors the intesity of the hue will be purer that when done with gray or black. Thus the term complement ...that which makes something whole. These are: R&G; B&O; Y&V.

[Phil marchese]

Dave would you be kind enough to post some colors at and about H189 S22 B52 please.

[David Rapasi]

that sure beats the old "Make paint chips on 3x5 cards method" we had to do in the Floquil days

David, can those paint matches you come up with, like that Salmon, be output as CIE or Munsell, or any other paint designation systems?

no need to list them or anything, I am just wondering if that is possible

Rex

I can’t make any sense of the CIE color system.

The Munsell System only has 100 numbered color hues, and the chroma goes way above 100% when comparing it to the saturation used by Adobe Photo Shop.

I have made color chips according to the Munsell numbers, but they aren’t as accurate as the system used in Adobe.

The photo of the color picker shows the different paint numbers available HSB, RGB, Lab, and CMYK. The number in the lower left corner is the ink number.

When you click the “CUSTOM†button you get another nine color measuring systems.

The only numbers I am concerned with are the HSB, I can use these directly to mix paint by comparing my paint chips to the available paint chips online and adjust the colors accordingly.

Dave .

[David Rapasi]

Dave would you be kind enough to post some colors at and about H189 S22 B52 please.

Phil

Here is your color chip.

Dave

[David Rapasi]

Are you trying to replicate the photograph? That seems all that this is relevant to.You're not going to replicate the original colour on the aircraft this way.

Steve

Steve

In these two photos I measured the interior colors and exterior color of a Corsair.

The interior color has a hue of 150°.

When compared to known colors from that period it is a close match to ANA 612/ FS 34092.

The color measured on the exterior was hidden from fading and is a close match to known colors from this time, ANA 603 Sea Gray.

Dave

[Phil marchese]

Dave thanks for the chip. It is about what I expected...sllightly darker that I imagined. I was shooting for another go at the. RAF PRU color. But just as a guess not research.

S

Some of the other systems you ve listed are tradtional printers 3 and four color process mix to match stanards and one is for digitial priting. I suspect. One is for product design. Drapes matched to plastic pitcher. Many industriies of course are involved in color matching. Expertise comes in many forms.

[FM-Whip]

Paint colors can be measured in Adobe Photo Shop to get the Color Hue°, and the percentages of Saturation and Brightness, ( H:S:B: ).

Reading this post and your previous post on the subject I think your methodology is flawed.

First off, (as others have said) scanning photos to come up with accurate colors opens up a very large can of worms. The first thing is that you are not using a standardized light source for all of your "specimens". You don't have standardized input. Also, if you scan a photo of an airplane or piece of an airplane with some color you are interested in, what determines the specific area of color? Are you picking a specific pixel and interpolating from that? I can scan a color photo and I can then use Photoshop to pick hundreds of colors in any one photo - one pixel movement away and I will get a totally different color (hue/saturation value). Do you average out the values over a large area and go from there?

John Hairell (tpn18@yahoo.com)

[David Rapasi]

Reading this post and your previous post on the subject I think your methodology is flawed.

First off, (as others have said) scanning photos to come up with accurate colors opens up a very large can of worms. The first thing is that you are not using a standardized light source for all of your "specimens". You don't have standardized input. Also, if you scan a photo of an airplane or piece of an airplane with some color you are interested in, what determines the specific area of color? Are you picking a specific pixel and interpolating from that? I can scan a color photo and I can then use Photoshop to pick hundreds of colors in any one photo - one pixel movement away and I will get a totally different color (hue/saturation value). Do you average out the values over a large area and go from there?

John Hairell (tpn18@yahoo.com)

John

The photo of the Corsair tail section has a color in the hue of 150°.

The photo of the side of the tail the measurement was taken from the areas the size of the white circles and I get a hue measurement of 206-208°.

This aircraft was built before the change to the tri-color paint scheme.

Pick your colors from the list below of the colors used in this time period.

What color would you paint your model?

QMS #12 Blue Gray has a Hue of 203°

QMS # 41 Olive Drab has a Hue of 40°

QMS # 42 Medium Green has a Hue of 131°

ANA 603 Sea Gray has a Hue of 208°

ANA 605 Insignia Blue has a Hue of 231°

ANA 607 NS Sea Blue has a Hue of 220°

ANA 611 Interior Green has a Hue of 84°

ANA 612 Medium Green has a Hue of 153°

ANA 613 Olive Drab has a Hue of 40°

Dave

[Phil marchese]

Interior green at 84 is a SHADE of yellow. OD @ 40 would indicate it is an adultrated red oon its way to brown through shading deintensifing the hue...but not in the green family. That could be true for a brown drab and as the red is absorbed the shading color would appear...first purple then blue gray.

[FM-Whip]

[stuff snipped]

Dave,

I took a photograph I have of a closeup of a small area of an aircraft, which has one color of paint. I scanned the entire photo, and then I also scanned 4 sub-sections separately. For each of the four section scans, I did 8 to 12 color samples (in Photoshop) in a rough grid pattern.

The results are below, in H/S/B format:

file point h s b sample size (pixel)

scan1 1 20 24 95 1

scan1 2 28 26 95 1

scan1 3 25 25 96 1

scan1 4 29 28 94 1

scan1 5 25 32 90 1

scan1 6 21 32 87 1

scan1 7 22 24 88 1

scan1 8 25 35 86 1

scan1 9 26 34 84 1

scan1 10 25 32 85 1

scan1 11 28 35 84 1

scan1 12 26 35 80 1

total: 300 362 1064

mean: 25 30 89

scan2 1 30 45 73 1

scan2 2 21 40 69 1

scan2 3 21 39 68 1

scan2 4 18 36 70 1

scan2 5 26 43 67 1

scan2 6 27 40 66 1

scan2 7 24 38 65 1

scan2 8 26 43 67 1

scan2 9 25 49 64 1

scan2 10 21 44 64 1

total: 239 417 673

mean: 24 42 67

scan3 1 28 22 98 1

scan3 2 28 24 96 1

scan3 3 26 24 97 1

scan3 4 30 26 95 1

scan3 5 28 29 95 1

scan3 6 27 30 91 1

scan3 7 25 32 92 1

scan3 8 23 30 89 1

scan3 9 22 33 93 1

scan3 10 26 34 87 1

total: 263 284 933

mean: 26 28 93

scan4 1 25 45 69 1

scan4 2 26 49 64 1

scan4 3 24 46 72 1

scan4 4 23 46 66 1

scan4 5 25 52 64 1

scan4 6 21 42 75 1

scan4 7 21 43 65 1

scan4 8 21 47 60 1

total: 186 370 535

mean: 23 46 67

If you take the mean of each of the four sample blocks above and set those up in HSB values in Photoshop, you can construct

a 4-sector file with those specific colors to paint with. Two of the mean values are close to each other, but two aren't.

Also, I deliberately used a photo which has little color variance on the ACTUAL aircraft, to demonstrate that even with a

relatively uniform color at full size you will get variations at the pixel level in scans.

My point is that all of the values above came from one small area of one aircraft, and they (in the computer) all represent the same basic ACTUAL color. Even if you average the samples as I have done above, you still get quite a large variation in color over a small area. Based on your methodology, the final color above could be one of three (two lighter shades, and two darker shades which are very close together, i.e. samples 2 and 4 are so close as to be indistinguishable). Samples 1 and 2 are noticeably different from each other, and very, very different from 3 and 4.

If you do these colors in Photoshop you will get anything from a light tan to chocolate brown, and based on your methodology you would pick some sort of model color or combination of model colors to paint your model with which matches some of the samples above - but which ones?

Unfortunately the colors that Photoshop says are there are not the real color that's on the real aircraft. Can you identify the TRUE, real-life color from the scan values above?

I can e-mail you all my original scan test files if you want.

John Hairell (tpn18@yahoo.com)

[toniosky]

Would it be a good idea to normalize the different pictures with standardized colour samples (ex: white+blak+some other colour), that would allow further colour balance with Photoshop. Of course, it means taking pictures with this samples on the pics, so forget about interpreting ww2 colour pictures.

[Phil marchese]

Insignia red?

[WymanV]

I'm dying to know how you're going to paint scale models in Photoshop. This'll revolutionize the hobby.

Ken

[Steve N]

Well, I suppose you could create a "skin" as a .jpg file, then apply it to a CGI model.

As for the rest, I'm at a bit of a loss as to what David is trying to accomplish. He exhaustively analyzes the colors in photos and goes on and on about color specs, then turns around and says he doesn't give a rip what color the real aircraft was painted. If that's the case, just paint it whatever color looks good to you and be done with it.

SN

[Phil marchese]

I suppose this is way past PROP planes forum.

[WymanV]

No, it's way past scale modeling. I'll bet better than 99% of scale modelers are not equipped with all the toners required to mix and colormatch paints to the tones and hues and tints to EXACTLY match a color that was hit-or-miss to begin with. It's far easier and cost effective to do the same thing the guys who painted them did-pull paint off the shelf and start spraying.

Not that there's anything wrong with discussing color variations. And not that they don't exist-I know from personal experience that there can be 6 or more variations on a single Chrysler Corp. color for the same model year-but when it comes to painting a model there's not a lot of issues with colormatch. It is what it is.

Ken

[Falcon50EX]

Dave, thanks for the paint mixing formulae. i have a couple of F4U-1s in the pipeline and i will give your mixing formula a try. i might even try re-painting my -1 interior with 34092 , or similar.

david