PDA

View Full Version : Pantone colors in LW



Tim Parsons
10-31-2018, 10:52 AM
I was recently asked to render an item in Pantone 161c and I thought, okay no problem there are Pantone colors in the LW color picker. Problem is the color to me appeared way too light for the sample I was given. I went to Pantone's website to find their RGB equivalent and sure enough LW's is way off. Even using different color spaces it's off. Can anyone shed some light on this discrepancy and how to confidently use Pantone colors in LW in the future?

Thanks

143254

50one
10-31-2018, 11:35 AM
Well, the problem with any CMYK conversion tool is pretty much down to monitor colorspace...you might have correct RGB value there but there might be some colour bleeding in your scene too, like GI..plus your monitor have to be calibrated...most of the monitors are not. Therefore if anyone ever ask me to do some Pantone colour I just say it's technically impossible. You can get close tho.

Tim Parsons
10-31-2018, 12:21 PM
Well I figured Pantone would take the time effort to show a close approximation on their website of what 161c should look like and therefore thought LW's color picker would show something similar on a monitor. Even if my monitor is off by a long shot the on screen representations should be close because they are supposedly representing the same color.

raymondtrace
10-31-2018, 02:05 PM
Here's an example of various RGB values (0-255) I see for a common ink I work with:

PANTONE 327 C

Pantone web site: 0, 134, 117
Photoshop 6: 0, 148, 133 (this is actually recognized as 327 CVC)
Photoshop CS6: 0, 130, 117
InDesign CS5.5: 0, 146, 143
Lightwave 2018 (with a LCL I might have manually loaded): 0, 131, 135
FreeHand MX: 0, 135, 125
Affinity Publisher: 0, 134, 117

Even before discussing digital colorspace/ICC and the affecting lighting within the scene, we must recognize that PANTONE "owns" and manages the color definition. They hold the power to tweak it over the years. These RGB conversions in software are often done with old or unlicensed (inaccurate) data.

Your best bet is to check the pantone site. However, you may never know if your client is also using a faulty rendition of the pantone color when they review your work. You may somehow get the color correct but they may disagree with its appearance.

EDIT: one more consideration... the displayed RGB numbers in each of these programs may reflect the values before or after color profiles are applied within the software. It depends on the software. The values shown above for Photoshop CS6 are for Adobe RGB. In sRGB, they become 0, 132, 118. If you toggle color spaces in Affinity Publisher, the RGB values remain the same.

Matt
10-31-2018, 02:15 PM
Those Pantone libraries you can get for the Color Picker I converted from a very old version of Illustrator, when their libraries were not in binary format. Some conversion will lose accuracy, they were never going to 100% exact.

raymondtrace
10-31-2018, 02:19 PM
...I converted from a very old version of Illustrator...

For legal reasons, you may not want to admit that. :D

Matt
10-31-2018, 02:22 PM
Meh, they're old as hell now.

Better tell this guy then:

https://magnetiq.ca/pages/acb-spec/

https://magnetiq.ca/pages/freeware/#acb2xml

Edit: Looks like that is out of date too. That tool didn't work

Tim Parsons
10-31-2018, 02:29 PM
Those Pantone libraries you can get for the Color Picker I converted from a very old version of Illustrator, when their libraries were not in binary format. Some conversion will lose accuracy, they were never going to 100% exact.

Ok - that makes sense. I won't use them as such then. Thanks for chiming in.

MSherak
10-31-2018, 05:33 PM
I was recently asked to render an item in Pantone 161c and I thought, okay no problem there are Pantone colors in the LW color picker. Problem is the color to me appeared way too light for the sample I was given. I went to Pantone's website to find their RGB equivalent and sure enough LW's is way off. Even using different color spaces it's off. Can anyone shed some light on this discrepancy and how to confidently use Pantone colors in LW in the future?

Thanks

143254

Big problem of converting Pantone to RGB is that they use their own propetery color space. Hence why one has to pay for it. To get close to the pantone color from the CMYK values requires running through some hoops.

What to do:
1) CMYK -> RGB
2) Add +5 to each color channel
3) Convert the new RGB with a Gamma .4545
4) Multiply the new RGB values by 0.685
5) Now it should be close to the pantone color.

Example for Pantone 161c:
1) 16,67,100,71 -> 62,24,0
2) Add +5 so we get 67,29,5
3) Gamma 0.4545 results in 141,97,44
4) Multiply each channel by 0.685
5) end result 96,66,30 (Pantone value 96,61,32)

It's closer to the Pantone color than a strait CMYK conversion. Course at this point just check their website for what RGB value they think one should use to match their color.

-M

jwiede
11-01-2018, 12:49 PM
2) Add +5 to each color channel

So I get what you're doing with the other steps, but what is this step meant to address?

Is there really a "floor offset" or equiv. in the Pantone color space curves? Where did the 5 value come from?

Just wondered, thanks!

raymondtrace
11-01-2018, 02:19 PM
I got lost at step 1.


1) CMYK -> RGB

We really should not be dealing with CMYK. Pantone colors are sometimes specified because they cannot be replicated in CMYK inks. So why look at CMYK values?

Modern Pantone swatches are defined in the Lab color model. Old CMYK definitions in software were in whole numbers (even though I vaguely recall a Pantone mix book that displayed non-integers). Modern color tables use more precise values in Lab. If you're looking at pantone-RGB or pantone-CMYK reference tables, you're already a few steps behind a proper conversion.

MSherak
11-01-2018, 04:58 PM
So I get what you're doing with the other steps, but what is this step meant to address?

Is there really a "floor offset" or equiv. in the Pantone color space curves? Where did the 5 value come from?

Just wondered, thanks!

Pantone never lets a color channel hit 0 and they a wacky curve done a loooong time ago in analog days. So I go by rule of thumb to add +5 under 127 and +11 over. Kinda their averages of the analog curve.



I got lost at step 1.



We really should not be dealing with CMYK. Pantone colors are sometimes specified because they cannot be replicated in CMYK inks. So why look at CMYK values?

Modern Pantone swatches are defined in the Lab color model. Old CMYK definitions in software were in whole numbers (even though I vaguely recall a Pantone mix book that displayed non-integers). Modern color tables use more precise values in Lab. If you're looking at pantone-RGB or pantone-CMYK reference tables, you're already a few steps behind a proper conversion.

Yes since LAB is a better way to adhere to the analog curve. Yeah it really is a silly place.

Tim Parsons
11-01-2018, 08:02 PM
Thanks for all the great info everybody!

jwiede
11-02-2018, 03:18 PM
Pantone never lets a color channel hit 0 and they a wacky curve done a loooong time ago in analog days. So I go by rule of thumb to add +5 under 127 and +11 over. Kinda their averages of the analog curve.

Ah, okay, that makes more sense. Thanks!

jwiede
11-02-2018, 03:23 PM
Yes since LAB is a better way to adhere to the analog curve. Yeah it really is a silly place.

Hmm, is a CIELAB (Lab) LW CS even possible, or would it require a 3D LUT (versus 3x 1D LUTs)?

next_n00b
11-04-2018, 08:07 PM
I do not think that Pantone colours are mixed (made) inside the computer? It is too limited, especially the monitor. Pantone has a much wider gamut.

raymondtrace
11-05-2018, 10:17 AM
Pantone has a limited subtractive color gamut at only a few thousand colors per set ( https://www.pantone.com/color-intelligence/articles/technical/pantone-numbering-explained )

Some of those colors extend outside of the CMYK gamut. However, with only a couple thousand Pantone inks, you can produce more distinct # of colors within the CMYK inks.

You can, of course, produce millions of additive colors on the computer (and monitor).