All Your Pixels Are (Probably Not) Belong To Pantone

There’s a piece of news floating around the open IP and allied communities at the moment which appears to have caused some consternation. It comes from Adobe, who have announced that due to an end of their licensing deal with Pantone LLC, PSD images loaded into Photoshop will have pixels containing unlicensed Pantone colours replaced with black. What, Pantone owns colours now? Are we expected to pay a royalty every time we take a picture of a blue sky? It’s natural to react with suspicion when hearing a piece of news like this, but for once we think this might not be the unreasonable intellectual property land grab it may first appear. To illustrate this, it’s necessary to explain what Pantone does, and what they don’t do.

A Heinz baked beans tin
Heinz use Pantone to ensure their Viridian Green baked bean branding colour is consistent. Use it on a can of beans and Heinz will probably sue you, not Pantone.

For a company that bases its whole product line on colours, it might seem odd to say that Pantone do not own or sell colours. Instead their product is in effect a colour matching service, a library of defined and named colours which can be matched by designers, printers, ink manufacturers, paint companies, and anyone else who produces a coloured product. The bit they own is the name and index number for a colour in their library, not the colour itself. If a designer creates a logo for a customer and specifies a Pantone colour for it, the customer knows that they can order the paint for their trucks in that exact colour from a Pantone-licensed paint company, or have their packaging printed in the exact same colour by a printer using Pantone-licensed ink. Consistency in branding is important for companies, and it’s the consistency that Pantone sell, not the colours themselves. The customer is free to match colours themselves from any ink or paint, but as they will soon discover, exact colour matching is not an easy task. Pantone’s business lies in taking away that headache.

It would thus be extremely difficult for Pantone to argue that an image which happens to contain a load of pixels that match a colour in their library are infringing on their IP, so your pictures are safe from their grasp. The reason some Photoshop PSDs are now facing the problem is that Photoshop allows a designer to attach a Pantone index to a colour, and for files which have this applied what Adobe are saying is they no longer have the licence to act on that. There is a whole Pandora’s Box in asking why in 2022 a proprietary image processing package on a flawed monthly subscription model still has such a hold on designers, but as far as Hackaday readers are concerned there should be nothing to worry about. Nobody is coming for our precious #F3BF10!

Header image: Tuxyso (CC BY-SA 3.0).

80 thoughts on “All Your Pixels Are (Probably Not) Belong To Pantone

    1. There’s rules I use to decide if I should be paying a subscription.

      1: Am I getting something I need that I can’t get otherwise.
      2: Is it fair. Note that Tesla’s heated seat scam isn’t fair.
      3: Is the payee ethical. I’d rather die in a pool of my own blood than support a monster, and now that I mention it, I’m almost there right now, _literally._

      Have a great day all.

      1. Your points are valid and interesting, I’d like to say my #1 criteria concerns the ongoing ownership of my own work.
        What happens if the company folds, gets bought out, gets hacked, the subscription fails to be paid or suffers any of a dozen other presently unforseen reasons.
        All that effort and hours spent and I wonder if I can perpetually be guaranteed I’m able to pull my stuff back from the cloud and in a usable state for importation into another product.

        1. I use Adobe Lightroom – and grumble about the subscription model. But I’ve not found anything better and it’s a major reason I must have a windows PC (vs my workhorse Linux desktops). If there’s a truly comparable alternative someone please inform me. Anyway I archive the originals and then convert all my Canon, Olympus, and Panasonic proprietary format images into DNG which is open and has a great chance of being usable into the future. X-rays I store as TIFF. I don’t worry much about being able to access my files years from now. Images I send to clients are converted to high res SRGB color space JPG or PNG and archived in the cloud so even if all my computers and backup drives are burned or stolen I can at least get to the images my clients are using. Metadata contains all pertinent job details.

          1. You deemed darktable to be insufficient i reckon? As a hobby photographer, it seems to do most lightroom does. I am not too involved though, but maybe you simply didnt stumble across it yet.

        1. Isn’t Tesla the $15,000 “Full Self-Driving Capability” nonsense though? It’s already there in the hardware you bought so upgrade anytime? After paying another $15,000?

          1. In 2021, Tesla says it delivered 936,172 vehicles around the world.

            If everyone bought “Full Self-Driving Capability”, just last year, that would mean $14,042,580,000 paid to programmers for just this upgrade? Not saying don’t compensate programmers for their time and experience but how many programmers does Tesla employ?

            If you assume every single employee and every single subsidiary employee in 2021 was somehow a programmer and assume that everyone bought “Full Self-Driving Capability” at $15,000, that would mean that everyone got a $141,844.24 bonus, right?

            Obviously the numbers don’t reflect reality and not saying don’t compensate programmers (though programming true full self driving with the limited hardware is a different discussion) but the point is that Tesla seems to be charging (quite a bit) for a feature that could just work hardware wise and be turned on at any point in time.

          2. I think it’s one of the cases where a separate subscription makes sense : the hardware required for the self driving is just a small computer and a couple of cameras. You can put these on a bicycle and you won’t get a self driving bicycle.

            What’s actually needed is probably millions of hours of work coding, testing, simulating, running trials, etc. To perfect the self driving. And, as opposed to to he rest of the car that is more or less “done” once you but it (except for all the repairs that you pay yourself), the self-driving algorithms have to be continuously updated if you wan my them to stay relevant as the roads evolve.

            I don’t know if this is enough to justify the price that Tesla asks for. But a self-driving system is really not the same thing as a car.

            (This is very different from, say, an OS where you would need to pay more for it to support N > M users; that is really just configuration that can be activated at any time, not like self-driving)

      2. Quite often you have to pay several hundred of EURO’s to remove a software bandwidth limit from an oscilloscope. And for Rigol that was not enough. They make powersupplies with a color TFT that only shows green, unless you pay extra for an upgrade (which also enables an extra digit of resolution). I find this idea so appalling that I won’t buy Rigol stuff anymore.

        1. Its bad, but other, we can hack these and have amazing tools at bottom prices.

          My worries are when these lock downs happen so hard, that you can’t bypass it anymore as hobbyist/hacker …

          For now, I’ve had all their cheap scopes, and the PSU all happily unlocked.

          1. I was just having this argument with myself.
            Sure, I agree, it’s affordable for me to buy a bare bones MSO34, and if I need SPI decoding or higher bandwidth, to buy it later.
            At the same time, it’s profitable for Tektronix to sell me that same MSO34 with all that added capability, at the bare bones price I paid. They could sell me a fully loaded MSO34 for the same amount, and they’d make the same profit they are right now and I’d get a lot more capability. I think that’s the part that rankles.
            It is possible they’re selling the bare bones at a slight loss and relying on upgrades to recoup their costs. I don’t know.

      3. For me there is also a 4: Is the subscription paying for a defined service or also to ensure acces to the results of that service?

        Example: A subscription that would convert my images to various formats automatically is nice. However, I expect “the conversion” to be the service and that I’m able to export the results and use it however I see fit.
        I’m not paying for a service that would keep the results of such a conversion hostage and allows me access only as long as I keep paying.

    1. It’s not just any old index, but an index that applies to color; a specific color, and a name. That is the aspect that Pantone has protected. Just like apple in general can not be protected, but Apple as applied to name a company can be.

        1. That’s why I used the word, “protect”. Trademarks, copyright, and patents all fall under the category of IP protection. The point I was getting at was that use of an index to refer to a particular color is a unique and original expression and so protectable.

    2. Their index is novel. That’s why no one else sells “Pantone” compatible anything. You are free to index the color data anyway you like, but Pantone’s system is copyrighted, just like the arrangement of words in a book even though the individual words are just data.

      What’s sticky is that the people who depend on Pantone are professionals and they already license things like typefaces and clip art and stock images. It’s just that now either Adobe doesn’t want to pass the cost of a license to Pantone on to all the rest of their customers (and thereby take additional profit) or Pantone is feeling some pinch in the sales of their services and are trying to generate a new revenue path.

      Color matching is an ugly, complex business, which is why art departments used to spend sums of money on Pantone swatch books rather than hope their print service came close to what the customer approved. Think if a Coca-epsi ad ran with somewhat pink where the red should be and slightly cyan where the blue part was intended. Those advertisers don’t like that – not at all.

      1. Yes! Thank you, people who know. This is a service, Pantone provides physical chips to match color to what you think you see on your screen. If you only work on a screen, no worries. If you want to make a physical object, and expect it to look like what you imagined from your screen, Pantone has done a lot of work for you. You can do it yourself, sure. It’s work though, and you will pay one way or another. There’s a large gulf between your screen and IRL. Most people will never need this.

  1. It sounds like they own those names for colors. Or indices if you want to consider them as names for colors. So if I say that this is Pantone number 37882 then they can get fussy, but they don’t just own the integer 37882. Or something like that if I am getting a grip on things. I think I can live with all that.

    1. People don’t want to live with it. The key is that Pantone has a whole quality control and licensing operation with software, colorimeters, physical color swatches, pigment and ink suppliers, etc.

      The issue is if you make 100,000 bright orange laundry detergent bottles and because you used some imitation standard, the hue is off, the customer rejects it, and now you’re stuck with 10 tons of worthless plastic, plus penalties for missing the delivery schedule, plus loss of future business. That’s going to cost you far more than the subscription.

  2. So Adobe can’t interpret Pantone color indices because the indices are IP-protected and Adobe must have a license from Pantone to map an index to an RGB/CMYK value? And I’m guessing that when you specify a Pantone color in a PSD image, there’s no color value in an alternate color space like RGB/CMYK/whatever that’s also stored for that color. Wow. That’s unfortunate for Adobe users who now have to buy their own Pantone license in order to avoid losing their PSD images. I wonder if Pantone was really squeezing Adobe when negotiating the license renewal. Too bad Adobe didn’t offer their users a way to convert Pantone colors before the license expired.

    1. Adobe did the same thing several years back when they decided to pass the Dolby licensing on to Premier users. That’s when I stopped paying them their monthly fee since all I was getting from updates was less features than I used to have and a huge headache opening any of my old files.

    1. Yes, RAL is sort of the same. But each color is more losely defined. You could have two RAL-9000 (white) objects from different suppliers and one might look yellow compared to the other, or blue.

    2. Pantone is for ink, for printing. RAL is for paint, for painting. So that truck mentioned in the article will be painted a RAL colour not Pantone.

      And relatively little printing is done in Pantone as CMYK is far cheaper. So most places will convert your Pantone colours to CMYK anyway. Pantone is more consistent, but requires setup costs as the ink has to be specifically mixed – and the excess chucked.

      1. IP can mean one or more of three things:
        Patent (expires after about 20 years from date of application).
        Copyright (Longest is Mexico, expires after the last author of a creation dies plus 100 years).
        Trademark (do not expire after a set period of time).

        At a guess in this particular case IP, is referring to Pantone LLC trademarks on their brand and their names for colors/colours.

  3. To be fair. Color is complex and Pantone is doing quite some work to ensure consistency of colors.

    The color values are “meaningless” by themselves, and it would be fair if Adobe just “error” out that it can’t read Pantone without an appropriate license. (Hopefully they don’t remove the Pantone data from the loaded file itself (ie, destroy the file), since that would be more than just shooting one’s own foot.)

    However, one can argue that a color is just a graph of the amount of reflected light from a surface across the visible spectrum, often interpreted to an “RGB” format to be displayed on a screen, or interpreted to “CMYK” for printing.

    Technically speaking, a color can be verified using a known “good” reference surface and an spectrometer. Like the one featured in the very next article here on HaD:

    Have a surface with a known amount of reflectiveness and a short term stable light source. This would allow us to measure the relative reflectiveness of a surface across the visible spectrum. This will only give Chroma if we don’t also consider the brightness. (however, absolute brightness is mostly meaningless since it isn’t a property of the surface, but rather the lights. So calibrating Chroma is in my opinion more correct. Since surfaces generally aren’t emissive in themselves. However, it is good to standardize a monitor to the ambient room lighting, else one’s screen won’t show the same color as the physical object one works with. And that is annoying. And yes, even a “color calibrated” monitor isn’t helpful if your room lighting as well isn’t calibrated.)

    So ensuring perfect color just need a calibrated spectrometer, and a preferably white reference surface or perhaps just a calibrated half silvered mirror. Calibration is rather “trivial” using lasers (single wavelength easily achieved by optical filtering) and an optical power meter (preferably measuring post filtering). And here things got absolute suddenly. So we are back at measuring color, not just Chroma.

    But we could throw in a wrench. Since if we also consider how not all humans sees just red, green and blue, but some also see additional colors too. Then these people could for an example argue that two samples of orange are completely different, even if the majority of the population can’t see a difference. So even printed reference color swatches are inherently flawed since it relies on the fairly limited and high short term drift spectrometer called the human eye. (Want to test short term drift of the eye, just stare at a colored blob (preferably a primary color) on a white background for a while. The blob will soon be surrounded by its inverse color since the cones in the eye needs time to reset, so if the blob is blue, the blue cones won’t see the white, but the red and green ones do, so there is your yellow outline.)

    But in the end.
    One could make a standard that defines colors by the absorption spectrum of a given surface.
    This would practically be “open source” and license free. And interpolating the values to RGB is already a well defined thing. (Downside is that every color becomes a 2d plot with likely hundreds (if not thousands) of values, and not just a couple of color values. But “ideally” a project would only use these standardized colors as base colors, coverting them to RGB/CMYK and then working in more traditional fashions for further effects as far as color mixing and shading goes.)

    So do we need an organization standardizing color swatches and names, monitor calibration, etc?
    A calibrated spectrometer goes a long way by itself.

    1. “(Hopefully they don’t remove the Pantone data from the loaded file itself (ie, destroy the file), since that would be more than just shooting one’s own foot.)”

      That’s exactly what they did. Adobe up and deleted pantone color data from customers files, and did so in a very specific and intentional way to prevent customers from recovering it.
      People who have re-licensed the pantone data directly still have ruined images with black pixels.

      Had they stored two converted image files, one with pantone colors replaced with black, and another with pantone colors converted to RGB/CYMK, there would be no problems here.
      The RGB version would still have valid colors as long as it isn’t edited.
      The two together would let you convert back, the black pixel version having which pixels were pantone, and the rgb version having which pantone color at the time.

      So not only did they ruin customers files in a way that can’t convert back once re-licensed, they also did it in a way that destroys the image as-is for anyone that doesn’t re-license and didn’t need to edit the image.

      Finally, Adobe then started this PR campaign redirecting anger of their destruction of data over to “pantone owns colors!” knowing most people reading that won’t bother to learn it isn’t true.

      Adobe claiming so many specific “accidental mistakes” cumulating in this specific outcome reads more like some crime drama. “Well your honor, he slipped and fell down the stairs into the electrified pool we always keep there, on wednesday-knife-in-pool day. clearly a tragic accident”

      1. I did hear that Adobe supposedly did irreparably destroy customer data. (even on their own cloud service….)

        But regardless.
        No program should EVER under any circumstances edit a loaded file. Only exception is when the user explicitly states they desire the file to be edited. (by saving the data to the file.)

        A typical example is when one migrates a program to a newer version of a program, most programs then informs the user that the file is from an older program version and that saving to the file will make it incompatible with the older program due to the introduction of new features. (exactly how valid that is is debatable in itself.) But the file itself isn’t destroyed unless one saves it. The loaded file might be loaded without certain information, making it effectively corrupt. But the file itself should still be unaltered.

        A similar statement can be made about installing programs as well.
        If one installs a newer/different version of a program, then it obviously shouldn’t remove the older/other version already installed unless agreed on by the user.

        Whenever a program wants to edit/remove data, then the user should have knowledge of it and the ability to say no. (with an exception to files that the program itself has just created, like temp/config files, etc.) Loading a file isn’t an agreement that the file can now be edited/destroyed.

        1. Agree 100%
          I can’t imagine the level of mismanagement at Adobe where this sort of decision could have successfully made its way from the top down to the programmers that implemented it. There should have been objections up the wazoo

      2. Pantone colors are for spot printing; they don’t really match any CMYK screen colors, particularly fluorescent and metallic colors. The screen colors are to give the user some reminder, but the actual process to print will not use that information except for some really “kinda-sorta” matches, where close enough is close enough.

        It’s not clear to me exactly what Adobe is doing except no longer creating the on-screen reminder colors which it licensed from Pantone. It looks to me that the files retain the Pantone color index values and will display correctly when the licensed Pantone library is added.

        1. “Pantone colors are for spot printing; they don’t really match any CMYK screen colors”

          3rd bullet point from pantone’s own website for their ‘pantone connect’ service:
          ‘Convert RGB/CMYK/Hex/L*a*b* color equivalents to Pantone Colors’

          “It looks to me that the files retain the Pantone color index values and will display correctly when the licensed Pantone library is added.”

          You have tried this? You should report that on their support forms. So far people who had their .ACB files deleted haven’t been able to re-import them after attaching a newly purchased pantone license.

    2. Kinda, but you’re less than half way there. A friend worked in paint design and explained that matching the spectrograph was only one step. objects also have other properties which affect how they look (gloss, scatter, not to mention angle-dependant factors).

      Pantone encapsulates some of that in that they are based on specific papers (e.g. coated/h coated), but there’s more at least with paint.

      1. Yes, surface texture can affect things.

        Partly I would say that that surface texture isn’t color. For the same reason I don’t consider room lighting as part of color. Even if both room lighting and surface texture are needed for color to work in practice.

        1. Counterpoint #1: the color of a butterfly’s wing comes from its surface texture (iridescence/interference). Some paints/pigments work the same way.

          Counterpoint #2: the spectrum of the light source changes what color an object can appear as, because the color of an object is typically subtractive. In other words, it does not have a color without the environmental context.

          1. 1.
            Still a subtractive surface color. Since it filters out specific wavelengths. (discussing color for features at or bellow the wavelength of the EM wave in question is partly pointless.)

            Yes, color is practically always subtractive. With an exception to emissive surfaces.
            So yes, we can only see the portion of the spectrum that our light source provides.

            But since changing the light source doesn’t affect the surface, it only effects our perception of the surface. And this is why I generally think one should disregard the light source when talking about surface colors.

            In short.
            We can divide color into three groups:

            1. Surface color. (Its actual color if lit with a “flat” spectrum.)
            2. Emissive color. (The color of light sources.)
            3. Perceived color. (The combination of the above.)

            I for one consider it incorrect to call a red pigment as gray/black just because one lacks the red portion of the spectrum in one’s illumination.

            If light intensity should be regarded as important for colors or not is likewise debatable. (In my own opinion, absolute brightness shouldn’t matter, relative brightness however should matter. Since a 25% reflective color is obviously not the same as a 75% reflective one. But for emissive colors, even relative brightness starts becoming meaningless.)

    1. Ah, but you’d have to have a pantone license (or pirate it) to do that.

      The main people that will be affected by this are people in the graphic design or marketing department of big companies, and so, pirating is not an option (could get their employer sued).

    2. Pantone is for spot color – there is a layer where that is the only color and the graphics for that layer have no color associated with them. Spot color is solid print color, though I suppose one might try a half-tone screen to “shade” them. If you look at product labels, any solid area is likely a spot color – it won’t be a blend of C-Y-M-K – no half-tone dot pattern. They include metallic and fluorescent colors that cannot be CMYK printed or RGB displayed.

  4. Gimp and Paint.Net for the win! Stupid stuff like this is why I use free, open source software whenever and where ever I can.

    Micro$oft wants to charge a subscription fee just to use office? Bye! LibreOffice is free! My vendetta against rotten Apple began back around 2001, when 30GB iPods were first released. Creative Labs had one, did the same thing, HALF THE PRICE. I then realized that all of those poor sods that stand in line to replace their iPhone every year are just throwing their money away, paying for the name.

  5. What utter scum Adobe are being. Surely they could set a dialog box to appear when an image with pantone colours is open and have the user select for each listed pantone colour what, in the user’s opinion, is a decent match from a HSL colour space, doing this for each pantne colour used so as not to destroy the image’s information. As the user would then be matching pantone names to colours for themselves a pantone license has nothing to do with it. From their actions here Adobe clearly has some form of ulterior motive to exploit user suffering for profit.

  6. Pantone is not a color reference. It is a pigment reference. A pigment can produce a color when a light is shined on it and seen by a human eye. Which color is produced depends upon the light and many other factors. To produce a pixel representing a Pantone pigment on a light emitting display requires a number of assumptions.

    1. “To produce a pixel representing a Pantone pigment on a light emitting display requires a number of assumptions.” Seems to me that producing “a pixel” “representing a Pantone pigment” would be a very rare thing indeed.

  7. Was it in the original Adobe licence terms that access to the Pantone colour matching system could be revoked at any time?

    Yesterday, I’ve created/edited an image, contained in a PSD file, while my usage was fully licenced. That image complied with all licencing arrangements. Today, I open that image, and suddenly it’s not compliant?

    Adobe might be setting themselves up for violations of user IP. If only they’d set up a dialog that pointed out Pantone wasn’t licenced any more, here are the elements tagged with Pantone colours and their corresponding pantone numbers, and offered 1. a link to a pantone subscription, and 2. a close match in RGB and CMYK spaces. They do it automatically for missing fonts in InDesign.

  8. the reason people don’t jump ship from Adobe products is that the alternatives to a great degree are awful. GIMP’s interface is shockingly blocky (though I am not denying its abilities), Scribus certainly is not as user friendly as many of its rivals. Inkscape is certainly an improvement.

    Affinity Suite shows good promise as regards Photo and Publisher, but Designer is still far behind. It’s a great shame they do not release their applications for Linux or one of the BSDs.

    So the world is stuck with a model its users hate until something finally does come along to break the mold.

  9. The only thing that’s changed in Adobe’s agreement is they’re no longer distributing the color book files they used to that Photoshop uses to reference the color.

    If you already have one of those ACB files (you have a file that was created with one) just copy it into the new version it wasn’t included in.

    If you don’t already have one, how you get one is up to you.

  10. The main reason a proprietary image-processing package on a flawed monthly subscription model still has such a hold on designers is GIMP’s chronic inability to acknowledge that colorspaces other than RGB exist. This immediately disqualifies GIMP from several creative and professional workflows. In these workflows, Photoshop is essentially a monopoly, which is why Adobe can get away with forcing their users to pay a monthly subscription.

    Most computer people fail to understand how important it is for some people to have their colors come out right when they leave the digital world. This immediately makes RGB a non-starter, since you are at the mercy of whatever software or firmware will perform the conversion from RGB to CMYK or to the dyeing process used. Not to mention that not all sRGB colors map to CMYK or a particular dye process and backwards. This leads to several “interesting” problems such as certain teals coming out either as green or blue or certain bright colors coming out dull and washed out and generally just “off”, or in some extreme cases blues coming out purple-ish. With a system like Pantone, you specify in your PSD file that you want color #678 and then it’s the responsibility of whoever will transfer your PSD file to the real world (be it a print shop, a paint shop spray-painting a logo a truck, a factory dyeing a t-shirt) to match color #678 using the common reference Pantone provides. Ideally, you will also pick colors from a set of physical swatches bought from Pantone, and not by looking at what looks ok on your PC monitor. Some brand guidelines even specify a Pantone color to use.

    As regards to what happens now, someone has already made something called “freetone”, so it will be like that time Adobe tried to charge for PostScript and the industry came up with TrueType.

    1. Let us hope. Adobe has done a nasty thing to those in the trenches by calling Pantones’s bluff. This is all about translating the digital image into real physical objects. It is a space that has to be reckoned with, but often ignored by people who live in screens.

  11. Blissfully colourblind. Pantone got no hold on me. To me they guarantee that pink looks exactly like light gray. Anything over 24bit colour is wasted on me, so to me they’re just a whining company that numbers and names imaginary products and copyrights them. Ferrari red and Coca Cola red are identical to me.
    If you want a specific colour ink you can come mix it for me (so it remains your problem).
    To get around them though, mix in IR or UV. Measure the temperature of your monitor and add that as a constant 4th colour to the entire set. Don’t forget to claim copyrights and set exorbitant prices.
    Software can easily be adapted to completely ignore the 4th colour when rendering.
    You are ready for the future. At some point Pantone may even contact you to buy your system, when marketeers of screens convince consumers that IRUHDTV is really the way of the future (nobody can see it but that goes for 8K and 300FPS too).
    (Stay on the IR side. The competition will make a case of your UVUHDTVs causing skin cancer).

    Imagine a colourfan with “Best viewed at 298 K(elvin)”…

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.