From: INTERNET:clare@typeshop.co.nz, INTERNET:clare@typeshop.co.nz
Date: Tue, Nov 16, 1999, 11:52 PM
RE: Reverse Pantone Directory
I often receive simple logos in cmyk tif form. The customer wants a two colour spot printing job and I enjoy using Illustrator to generale gook sharp logos. Usually the logo has started its life as a vector file and then been saved in Photoshop. It is possible to flip through the Pantone solid to process book and get back to the original colors. Is there a file of some sort out there where you could key in the cmyk figures and have it spit out the Pantone colour? It would save a little time and I think it is the sort of thing someone might have done.
Thanks
From: Chris Murphy, INTERNET:lists@colorremedies.com
Date: Wed, Nov 17, 1999, 12:02 AM
RE: Re: Reverse Pantone Directory
>I often receive simple logos in cmyk tif form. The customer wants a two
>colour spot printing job and I enjoy using Illustrator to generale gook
>sharp logos. Usually the logo has started its life as a vector file and
>then been saved in Photoshop. It is possible to flip through the Pantone
>solid to process book and get back to the original colors. Is there a
>file of some sort out there where you could key in the cmyk figures and
>have it spit out the Pantone colour? It would save a little time and I
>think it is the sort of thing someone might have done.
Praxisoft Vector Pro will do both solid to process conversion and process to solid look-up. The later is more accurately referred to as finding the "closest color". You can input CMYK or RGB values (ICC profile for the device to be used is required) and it will locate the closest Pantone (coated, uncoated, prosim, pastel, and other), or FocolTone color. You can also select a Pantone or FocolTone color and it will tell you the appropriate CMYK or RGB (or five, six, seven or eight color combination if applicable) recipe for that specific output method. It will also compute Delta E (the difference between the twocolors, solid and process). It will also display a simulation on screen for the solid and the process colors. Naturally that feature depends on a calibrated and profiled monitor. There might be a demo at the Praxisoft web page. If not, talk to Stuart at Praxisoft and see if you can get a limited demo copy. Retail price is approximately $395.Chris Murphy
From: Luca Ragogna, INTERNET:luca@interhop.net
Date: Wed, Nov 17, 1999, 1:42 AM
RE: Re: Reverse Pantone Directory
PhotoShop will do that. While you're in the colour picker, type in your CMYK values then click "custom" and the closest PMS colour will be highlighted.
> I often receive simple logos in cmyk tif form. The customer wants a two
> colour spot printing job and I enjoy using Illustrator to generale gook
> sharp logos. Usually the logo has started its life as a vector file and
> then been saved in Photoshop. It is possible to flip through the Pantone
> solid to process book and get back to the original colors. Is there a
> file of some sort out there where you could key in the cmyk figures and
> have it spit out the Pantone colour? It would save a little time and I
> think it is the sort of thing someone might have done.
> Thanks
From: Jeff Yerkey, INTERNET:jeff@ydg.com
Date: Wed, Nov 17, 1999, 2:20 AM
RE: Re: Reverse Pantone Directory
X-Rite's Colorshop software will also do this.
Jeff Yerkey
---On 11/16/99 10:40 PM, luca@interhop.net so ably wrote:---
>PhotoShop will do that. While you're in the colour picker, type in your CMYK
>values then click "custom" and the closest PMS colour will be highlighted.
>> I often receive simple logos in cmyk tif form. The customer wants a two
>> colour spot printing job and I enjoy using Illustrator to generale gook
>> sharp logos. Usually the logo has started its life as a vector file and
>> then been saved in Photoshop. It is possible to flip through the Pantone
>> solid to process book and get back to the original colors. Is there a
>> file of some sort out there where you could key in the cmyk figures and
>> have it spit out the Pantone colour? It would save a little time and I
>> think it is the sort of thing someone might have done.
>> Thanks
From: Steve Upton, INTERNET:upton@wolfenet.com
Date: Wed, Nov 17, 1999, 3:36 AM
RE: Re: Reverse Pantone Directory
On Tuesday, November 16, 1999, Chris Murphy <lists@colorremedies.com> wrote:
>Praxisoft Vector Pro will do both solid to process conversion and process
>to solid look-up.
> Retail price is
>approximately $395.
Ooooh I wish! Vector Pro is a very capable program but its price is $1749 A less capable but useful alternative is X-Rite's Color Shop. It will also lookup Pantone matches and it has some other cool tools and a good color primer from Fred Bunting. Check it out here:
http://www.chromix.com/chromix/store/toolsearch.lasso?toolid=421
I think Vector Pro is one of the best formulators for matching Pantone and other products and coupled with ICC Autoflow is one of the best vector color solutions.
From: Chris Murphy, INTERNET:lists@colorremedies.com
Date: Wed, Nov 17, 1999, 10:39 AM
RE: Re: Reverse Pantone Directory
>PhotoShop will do that. While you're in the colour picker, type in your CMYK
>values then click "custom" and the closest PMS colour will be highlighted.The problem is that you can't easily go in reverse with Photoshop and get accurate results. Going from CMYK to Pantone in Photshop is dependent on
Photoshop's built-in CMYK equivalents for Pantone colors, NOT CMK Setup.
You can expect completely different Pantone solid suggestions if your basis for CMYK values isn't compensated for. The same CMYK value on
uncoated, vs coated, vs newsprint, vs magazine, vs supercalendered, let alone dot gain and different ink sets, will look different for different conditions; different enough you will end up specifying different Pantone colors for each of these conditions.
From: Chris Murphy, INTERNET:lists@colorremedies.com
Date: Wed, Nov 17, 1999, 10:51 AM
RE: Re: Reverse Pantone Directory
>Ooooh I wish! Vector Pro is a very capable program but its price is $1749
Ooops. Obviously I'm confused. I think I must be thinking of ICC Autoflow (the non-network version). I think it's the same price as Compass Pro XT.
Time to get a new price list from Praxisoft. I'll have to confirm this price with Praxisoft anyway, because it seems Vector Pro should be less money that the previous application it was based on, Compass Master. Compass Master had all kinds of features that many users found confusing. So they made a new application, Vector Pro, which has a simpler interface, as well as a different engine allowing the creation of RGB palette file.
Anyway, I've got it and it's a very cool program (both Compass Master, and now the new Vector Pro).
From: Chris Murphy, INTERNET:lists@colorremedies.com
Date: Thu, Nov 18, 1999, 3:00 PM
RE: Re: Reverse Pantone Directory
>Chris writes:
><<The problem is that you can't easily go in reverse with Photoshop and get
>accurate results. Going from CMYK to Pantone in Photshop is dependent on
>Photoshop's built-in CMYK equivalents for Pantone colors, NOT CMYK Setup.>>
>This is not correct. The Pantone-supplied CMYK equivalents are forced on
>Adobe for legal reasons, and are not used in computing color equivalence.
>To find the Pantone equivalent of a given CMYK value, Photoshop converts to
>LAB and looks at Pantone LAB equivalents. That conversion *does* depend on
>CMYK Setup.
So it does. However, it's not doing as good a job as Vector Pro at finding the correct Pantone color from a CMYK color. For a particular
blue, 70,50,0,0 Photoshop gives me a suggested Pantone color of 300CVC. Vector Pro suggests 660CVC. Quite different colors. When comparing the actual Pantone chip to the actual printed color (70,50,0,0), 660CVC is quite a bit better of a match than 300CVC. The difference in delta E is around 6. So if you're tolerance allows that to be acceptable, then OK
>In short, Luca is quite correct, Photoshop does what Clare is asking for.
>It never ceases to amaze how many suggestions can be found for spending
>money on products that duplicate what Photoshop already does for free (and
>faster).
Free yes. I disagree with faster. Vector Pro will compute solid to composite or composite to solid far far faster than Photoshop does.
Better conversions are also perfomed with Vector Pro as well.
So it's going to depend on your tolerance factor, and your budget (which can affect your tolerance factor).
From: Andrew Rodney, INTERNET:andrew@digitaldog.net
Date: Thu, Nov 18, 1999, 3:46 PM
RE: Re: Reverse Pantone Directory
on 11/18/99 12:59 PM, Chris Murphy at lists@colorremedies.com wrote:
>> In short, Luca is quite correct, Photoshop does what Clare is asking for.
>> It never ceases to amaze how many suggestions can be found for spending
>> money on products that duplicate what Photoshop already does for free (and
>> faster).
> Free yes. I disagree with faster. Vector Pro will compute solid to
> composite or composite to solid far far faster than Photoshop does.
> Better conversions are also perfomed with Vector Pro as well.
Plus Vector Pro with an accurate ICC profile will provide very accurate recipes for Pantone to devices that don't at all deal with inks that are anything close to process (like RGB for example). And with CompassXT, you can swap the correct values for vector art on the fly.
From: Dan Margulis, INTERNET:76270.1033@compuserve.com
Date: Thu, Nov 18, 1999, 9:23 PM
RE: Re: Reverse Pantone Directory
Chris writes:.
<<The problem is that you can't easily go in reverse with Photoshop and get accurate results. Going from CMYK to Pantone in Photshop is dependent on Photoshop's built-in CMYK equivalents for Pantone colors, NOT CMYK Setup.>>
This is not correct. The Pantone-supplied CMYK equivalents are forced on Adobe for legal reasons, and are not used in computing color equivalence. To find the Pantone equivalent of a given CMYK value, Photoshop converts to LAB and looks at Pantone LAB equivalents. That conversion *does* depend on
CMYK Setup.
<<You can expect completely different Pantone solid suggestions if your basis for CMYK values isn't compensated for. The same CMYK value on uncoated, vs coated, vs newsprint, vs magazine, vs supercalendered, let alone dot gain and different ink sets, will look different for different conditions; different enough you will end up specifying different Pantone colors for each of these conditions.>>
Photoshop does all this. If you enter the same CMYK value in the picker with two drastically different CMYK Setups, you'll get two different Pantone equivalents, just as you should. In short, Luca is quite correct, Photoshop does what Clare is asking for. It never ceases to amaze how many suggestions can be found for spending
money on products that duplicate what Photoshop already does for free (and faster).
From: Dan Margulis, INTERNET:76270.1033@compuserve.com
Date: Fri, Nov 19, 1999, 11:49 PM
RE: Re: Reverse Pantone Directory
Chris writes:
<<So it does. However, it's not doing as good a job as Vector Pro at finding the correct Pantone color from a CMYK color.>>
You're right, it won't, if you use the wrong CMYK Setup. if you use something sensible, it will apparentely do a better job.
<<For a particula blue, 70,50,0,0 Photoshop gives me a suggested Pantone color of 300CVC. Vector Pro suggests 660CVC. Quite different colors. When comparing the actual Pantone chip to the actual printed color (70,50,0,0), 660CVC is quite a bit better of a match than 300CVC.>>
I would agree that PMS 660 is a closer match than PMS 300, or, more accurately, that it sucks less. Neither is close to 70C50M in most CMYK conditions.
However, as noted previously, the quality of the conversion depends on what you have in CMYK Setup. At first, I thought you must have paid somebody to generate you a custom profile, because only a colorimeter or spectrophotometer is ordinarily capable of generating such a bizarre result. If you use Photoshop 5's default settings, the value returned is PMS 646. With any of the three settings I use in my normal work, I get PMS 652, which is also the value returned by the ColorMatch SWOP ICC profile. Both of these inks are closer to 70C50M than is PMS 660.
If, however, your quality standards are such that you are willing to accept PMS 660 as a match, then you may load into CMYK Setup either the negative or the positive proofing profiles built into Photoshop (using relative colorimetric). All of these return PMS 660. For the heck of it, I loaded that abomination, the ICC profile called "Generic CMYK" into CMYK, and, coincidentally, it returned PMS 300. I am quite sure that you would not be crazy enough to have *that* loaded as your
CMYK Setup, so I am left with my surmise that you must have had somebody generate a profile for you with a densitometer or whatever.Meanwhile, if Vector Pro is in fact returning PMS 660, Photoshop's defaults are better. I think I'll spend the money on a few cases of good wine instead.
From: Chris Murphy, INTERNET:lists@colorremedies.com
Date: Sat, Nov 20, 1999, 1:50 AM
RE: Re: Reverse Pantone Directory
You're right, it won't, if you use the wrong CMYK Setup. if you use
>something sensible, it will apparentely do a better job. I'm using the same ICC profile in CMYK Setup as I have set in Vector Pro. I'm also using the same CMM. So they're both getting the same info, just deriving different suggested solids.
>However, as noted previously, the quality of the conversion depends on what
>you have in CMYK Setup.
The problem is that Lab is not a perfect color space for doing these conversions in the first place. CIE94 is used to compute more accurate delta E values and arrive at closer matches in Vector Pro.> At first, I thought you must have paid somebody to
>generate you a custom profile, because only a colorimeter or
>spectrophotometer is ordinarily capable of generating such a bizarre
>result.
It's a bizarre inkjet printer. Not a press.
>For the heck of it, I loaded that abomination, the ICC profile called
>"Generic CMYK" into CMYK, and, coincidentally, it returned PMS 300. I am
>quite sure that you would not be crazy enough to have *that* loaded as your
>CMYK Setup, so I am left with my surmise that you must have had somebody
>generate a profile for you with a densitometer or whatever.
Generic CMYK is based on some weird inkjet printers I think because while I always get horrendous results with it, it has inkjet type behavior. Personally I wish it would go away. Profiles aren't made with densitometers, either colorimeters or spectrophotometers, and they are also used to make Photoshop separation tables, so I don't know what that has to do with this.>Meanwhile, if Vector Pro is in fact returning PMS 660, Photoshop's defaults
>are better. I think I'll spend the money on a few cases of good wine
>instead.
Wine better than Photoshop's defaults which might not do a bad job, but at the same time, the default is NOT SWOP even though it's called SWOP.
From: INTERNET:clare@typeshop.co.nz, INTERNET:clare@typeshop.co.nz
Date: Sat, Nov 20, 1999, 4:16 PM
RE: Reverse Pantone Directory
When trying to work back to Pantone from a cmyk colour, I don't want the *best* match, I want the *actual* Pantone colour that the artist used before changing the vector file to a bitmap. The colour, that if I searched through my process to colour swatch book, I would find. This isn't happening when I key in the cmyk in Photoshop and then click Custom. I might be getting a better match, but not the colour the artist used before throwing a spanner in the works and changing the vector file to a bitmap file.
From: Dan Margulis, INTERNET:76270.1033@compuserve.com
Date: Sun, Nov 21, 1999, 2:38 PM
RE: Reverse Pantone Directory
Clare writes:
<<When trying to work back to Pantone from a cmyk colour, I don't want the *best* match, I want the *actual* Pantone colour that the artist used before changing the vector file to a bitmap.>>
This is roughly like saying, if given a document that has been translated into another language, you want to be able not just to know what the English-language meaning of it, but what the *exact words* the original author used.
If the CMYK equivalent has only three inks, and black is one of them (CMK, CYK, or MYK), then you are looking at an "official" Pantone equivalent, and you can find the match if you look through the Pantone book until you find that specific match.
If, however, the color is made up either of CMY or CMYK, then you are looking at some kind of translation of the original value, and unless you can find out what software did the translation and exactly what kind of conversion settings were in place, you won't know what the original color was.
<<The colour, that if I searched through my process to colour swatch book, I would find.>>
You would have no means of knowing whether you had found it or not. In the example Chris gave me (client uses 70C50M) that could conceivably be a translation of at least half a dozen different PMS inks. One could not tell which one it was by examining swatches.
From: Bob Smith, INTERNET:rmsmith@calpha.com
Date: Sun, Nov 21, 1999, 3:56 PM
RE: Re: Reverse Pantone Directory
on 11/21/99 1:34 PM, Dan Margulis at 76270.1033@compuserve.com wrote:
> and unless you
> can find out what software did the translation and exactly what kind of
> conversion settings were in place, you won't know what the original color was.
Which is why profile embedding that started with PS5 is such a useful idea. The profile may not tell you everything you need to revert exactly but it sure beats flying completely blind.
From: Chris Murphy, INTERNET:lists@colorremedies.com
To: "Color theory", INTERNET:ColorTheory@listbot.com
Date: Sun, Nov 21, 1999, 5:13 PM
RE: Re: Reverse Pantone Directory
>I might be getting a better match, but not the colour the artist
>used before throwing a spanner in the works and changing the vector file
>to a bitmap file.Tell them, "No! Bad artist! Bad!" and then make it expensive to fix the file (or just give it an F and send it back and make them fix it). If you
really don't want those kinds of projects, make it expensive enough that it's worth your time to do. If it's too expensive, then the artist will
either find someone else to be a pushover, or they will get their act together.
From: "Ron Bean", INTERNET:rbean@execpc.com
Date: Sun, Nov 21, 1999, 6:38 PM
RE: Re: Reverse Pantone Directory
Clare <clare@typeshop.co.nz> writes:
>When trying to work back to Pantone from a cmyk colour, I don't want the
>*best* match, I want the *actual* Pantone colour that the artist used
>before changing the vector file to a bitmap.
Why not call them and ask? While you're at it, ask (politely) why they didn't send the vector art in the first place. If it's a corporate logo, try calling the company's advertising department-- they can probably tell you the exact color (and maybe send you the vector art as well). If this level of communication isn't possible, then you're just guessing anyway-- you don't even know if the CMYK file is the right color to begin with.
I've don't know why some CSR's seem so reluctant to call the customer with questions like this. I once worked with a guy (in a different industry) who wasn't afraid to double-check things with customers, and I always wondered why he was the exception and not the rule. He'd say something like "a question came up, and I just wanted to double check so we can get your order right". Saved us a lot of head-scratching.
From: Dan Margulis, INTERNET:76270.1033@compuserve.com
Date: Sun, Nov 21, 1999, 10:56 PM
RE: Re: Reverse Pantone Directory
Bob writes:
<<Which is why profile embedding that started with PS5 is such a useful idea. The profile may not tell you everything you need to revert exactly but it sure beats flying completely blind.>>
In the case at hand--given a CMYK value, find the Pantone specification that *produced* it, as opposed to matching it with a Pantone color--I would disagree that it beats flying blind.There are no technologies extant or even theoretically possible that will translate into CMYK and back again and get the same result. The problem is especially bad when the original color is not in the CMYK gamut, which is probably the case here.
As Chris pointed out, two different apps using the same profile are apt to make the CMYK>LAB conversion differently. There are a bunch of plausible PMS candidates that are going to be very close together indeed in CMYK under any reasonable conversion scenario. So, the mere presence of a profile, even if one were to assume the unlikely, that this client knows enough to embed the correct one, wouldn't be particularly helpful, unless we also knew the CMM, the rendering intent, and the program(s) he was using, and whether there was any previous conversion, such as in and out of a Photoshop RGB working space.Even if we did know *all* of these things, there would still be no guarantee we'd get the right PMS value. But if we did know all these things, then I agree, it would be marginally better than flying blind, provided we were willing to spend a couple of hours reverse engineering the process rather than a couple of minutes developing an educated guess by looking at a swatchbook.
From: Chris Murphy, INTERNET:lists@colorremedies.com
Date: Sun, Nov 21, 1999, 11:46 PM
RE: Re: Reverse Pantone Directory
>Bob writes:
><<Which is why profile embedding that started with PS5 is such a useful idea. The profile may not tell you everything you need to revert exactly but it
sure beats flying completely blind.>>
>In the case at hand--given a CMYK value, find the Pantone specification
>that *produced* it, as opposed to matching it with a Pantone color--I would
>disagree that it beats flying blind.
There may not be a Pantone specification that produced it. Pantone only publishes theirs which isn't ideal most of the time. They themselves admit the presses they use have lower dot gain than the average press and the most common analog proofs which simulate 22% dot gain. You have a better chance of going back to a Pantone color through an NCP (Named Color Profile). I don't know of any applications supporting this, but the technology is there. If you used an NCP (and then embedded the ICC profile to which the NCP is linked) then you could feasibly move back and forth between solid and composite versions of a named color.
The *catch* is when the original Pantone color was a good deal out of gamut for the device. In that case you get compromise conversions into a
composite space that can actually end up being far closer to some OTHER Pantone color. So conversion back into Pantone's color space isn't
exactly an easy task one way or the other. But Bob is correct that it would be useful to have an embedded profile in the image which would at least tell you what Lab colors those CMYK values equate to (roughly) and if yo assume the original Pantone color was in-gamut, then you could find a close Pantone color reference to match to.Now, other than operator error (common), why would you be using Pantone colors at all unless:
1.) You were going to use Pantone specified inks on the job.
2.) You had reasonable assurance that the Pantone colors you specified could be reproduced with process printing inks.
Otherwise, don't use Pantone colors.
>There are no technologies extant or even theoretically possible that will
>translate into CMYK and back again and get the same result. The problem is
>especially bad when the original color is not in the CMYK gamut, which is
>probably the case here.
I come close to agreeing that no technologies exist that could do this, but I'll add, "that I'm aware of," in case I'm wrong. Theoretically possible, totally it is. All I need is a single modification to the ICC's Named Color Profile specification to include two more tags:
1.) Delta E
2.) Delta E deriviation (what method)
If I know what method was used to compute delta E, and I have the delta E, I can figure out from any CMYK value what Pantone color was originally defined to arrive at those CMYK values. As long as that NCP is used for both conversion from Pantone, and the conversion from CMYK you could convert freely between the two regardless of whether the original Pantone specified colors were out of gamut.
For all I know, the ICC spec has these tags, or at least unused tags where these two items could be specified in future NCP's without sacrifice to current compatibility. The NCP is so rarely used and has a black hole of support in current apps, a major rewrite of the spec wouldn't even affect compatibility. But that's going off on a tangent...
>So, the mere presence of a profile, even if one were to assume the
>unlikely, that this client knows enough to embed the correct one, wouldn't
>be particularly helpful, unless we also knew the CMM, the rendering intent,
>and the program(s) he was using, and whether there was any previous
>conversion, such as in and out of a Photoshop RGB working space.
The CMM is an issue. The rendering intent isn't. It's common practice to use the absolute colorimetric rendering intent for use with solid to
process conversions. There is no other rendering intent even recommended, while some applications allow you to choose relative colorimetric for
reasons of avoiding white point mapping (which can result in undesirable effects when used on a monitor).So unless the file itself in addition contains an embedded profile and the CMM used to generate that conversion, embedding alone is not as much information as we would like. This doesn't mean it isn't useful however.
> But if we did know all these things, then I agree, it
>would be marginally better than flying blind, provided we were willing to
>spend a couple of hours reverse engineering the process rather than a
>couple of minutes developing an educated guess by looking at a swatchbook.
Why not use what we have and complain to the vendors to make it better than just ignore it though? We are human beings on planet earth in a Western industrialized country going into the 21st century. We shouldn't have to pick up anything to convert back and forth. The technology exists to do it, but apparently there isn't enough demand (or vocalized demand) to improve on what we have. NCP's have been around for a while but Adobe doesn't support them (shame! shame! shame Adobe!), along with anyone else I'm aware of. There's a Pantone plug-in for a few apps you can get that will let you use an NCP, but by default we don't have apps that use them. It's silly. It would make things so much easier to have an ICC and NCP profile for every device and then you could easily generate ideal hand tuned process equivalents from any named color system (pantone, focoltone, others).
From: Dan Margulis, INTERNET:76270.1033@compuserve.com
Date: Mon, Nov 22, 1999, 8:33 PM
RE: Re: Reverse Pantone Directory
Chris writes:
<<I come close to agreeing that no technologies exist that could do this, but I'll add, "that I'm aware of," in case I'm wrong. Theoretically possible, totally it is.>>
Only if you encrypt the black channel, and/or accept a wild swing toward yellow. Otherwise, there will be fewer digital variants of blue possible in CMYK than elsewhere, and it will be impossible to retain every distinction.
<<If I know what method was used to compute delta E, and I have the delta E, I can figure out from any CMYK value what Pantone color was originally
defined to arrive at those CMYK values.>>
Won't help. At best, you might eliminate a few candidates, but basically you wouldn't know from which direction the variant delta-E came.
<<The CMM is an issue. The rendering intent isn't. It's common practice to use the absolute colorimetric rendering intent for use with solid to
process conversions.>>
It may be common practice among those who really understand what's going on, but I repeatedly run across advice saying use perceptual no matter what. In reading over the ColorSync list, it appears to me that around half of the major posters don't understand what these rendering intents are all about. I would therefore certainly not assume that this particular client knows enough to use that rendering intent. One could call him up and ask him, but if one were going to do that, one might as well just ask him what Pantone color he was using. Ron suggests that, and I agree.
<<Why not use what we have and complain to the vendors to make it better than just ignore it though?>>
First, the problem of deciphering what Pantone color provoked a certain CMYK conversion is not one that most people lose sleep over. Second, in speaking of those who not only use an ICC profile for RGB>CMYK conversion,but also embed one, we are talking about a small fraction of a tiny minority. Complaints to vendors should, IMHO, be reserved for matters of more consequence.
From: Chris Murphy, INTERNET:lists@colorremedies.com
Date: Tue, Nov 23, 1999, 12:28 AM
RE: Re: Reverse Pantone Directory
Dan writes:
>Only if you encrypt the black channel, and/or accept a wild swing toward
>yellow. Otherwise, there will be fewer digital variants of blue possible in
>CMYK than elsewhere, and it will be impossible to retain every distinction.
You don't have to. We already have Pantone to process (with compromises). All you need is a look-up table from CMYK for a specific device to go
back to Pantone (or any other named color space). You really aren't converting, the CMS basically says, "these CMYK values could only be created if the original value they asked for was X."
>Won't help. At best, you might eliminate a few candidates, but basically
>you wouldn't know from which direction the variant delta-E came.
Doesn't matter. For a single CMYK value there are a limited number of colors in a given named color system that those CMYK values could have been derived from. If you specified a delta E value (x.xx or maybe x.xxx) there is more than adequate information for the CMS to figure out which Pantone color was originally specified. The delta E's are precomputed and placed in the NCP at the time of building. Delta E isn't being recomputed on the fly when going back to the solid color.It's not a question of *IF* it can be done, it's a question of in order to get it to work correctly, how much more of a pain is it vs. the help derived from the pain. In order for this to work, the converion profiles use from solid to process must also be used when going from process to solid. If you don't use the same exact set of profiles, thenyou'reout of luck.
But with the push for standard printing methods, perhaps we'll see not only ICC profiles for these standard printing methods (such as TR001), but NCP's as well. That would be quite helpful.>It may be common practice among those who really understand what's going
>on, but I repeatedly run across advice saying use perceptual no matter
>what. In reading over the ColorSync list, it appears to me that around half
>of the major posters don't understand what these rendering intents are all
>about.
At least with NCP's rendering intent selection is ignored. You get a conversion per the NCP contents; not that of the device profile. But yes clearly if only the device profile is being used its e different results is the least of the problem when trying to go from process back to solid - it just won't work when you have such limited infomation in device profiles. It's not a table that contains all the solid color possibilities of any given system and that's what would be required for this idea to work.
>First, the problem of deciphering what Pantone color provoked a certain
>CMYK conversion is not one that most people lose sleep over.
I don't know, I hear more complaints about equivalents specified in the Pantone Solid to Process Color Guide not matching as well as those in the guide (<sarcasm on> hmm I wonder why? <sarcasm off>) than seeing users on the ColorSync User's list insisting perceptual is the correct rendering intent to use for solid to process conversions.
> Second, in
>speaking of those who not only use an ICC profile for RGB>CMYK conversion,
>but also embed one, we are talking about a small fraction of a tiny
>minority. Complaints to vendors should, IMHO, be reserved for matters of
>more consequence.
Those who use ICC profiles may be in a minority, but those who complainabout color matching problems are in the great majority. Ask any tech support department from Adobe to MicroSoft. They get the constant complaints from users having monitor to inkjet, monitor to press, inkjet to press, monitor to web, press to web, and other mixes of color matching problems. Those kinds of incoming support calls are in the great majority and that is one big reason why there is a push for
*something* to if not solve the problem, to reduce the problem and associated costly support calls.
I also disagree that there is a minority of Photoshop users embedding profiles. Most users (I argue) are using default settings, and if they are using default settings they are embedding profiles whether they intend to or not.
Adobe Photoshop training classes are taught in the US by Sterling Ledet & Associates, Inc.