Update: Epson 2200 ICC profile problem

James Gifford wrote:

So my question boils down to: What is the difference between setting the PS print color space and setting the ICC profile in the driver? Why does PS not understand what ICC profile the printer is set to?

It’s all explained here:
< http://www.computer-darkroom.com/ps7_print/ps7_print_mac.htm>.

—
Regards
Madsen.

Thomas Madsen wrote:

So my question boils down to: What is the difference between setting the PS print color space and setting the ICC profile in the driver? Why does PS not understand what ICC profile the printer is set to?

It’s all explained here:
< http://www.computer-darkroom.com/ps7_print/ps7_print_mac.htm>.

Thanks. (I’m glad to see I wasn’t the only confused by the change in dialogs.)

—

| James Gifford * FIX SPAMTRAP TO REPLY |
| So… your philosophy fits in a sig, does it? |
| Heinlein stuff at: www.nitrosyncretic.com/rah |

Thomas Madsen wrote:

So my question boils down to: What is the difference between setting the PS print color space and setting the ICC profile in the driver? Why does PS not understand what ICC profile the printer is set to?

It’s all explained here:
< http://www.computer-darkroom.com/ps7_print/ps7_print_mac.htm>.

After reading through it, this appears to be saying that I should either specify the correct 2200 profile in Photoshop and set the printing mode to "No Color Adjustment" (i.e., do what PS tells you to do), or specify "Printer Color Management" in PS and set the ICC profile in the printer driver… but not to do both.

I understand the first two alternatives, but I don’t understand why both PS’s print management and the driver’s color management should be enabled. (I *seem* to have gotten excellent results with both specified, but I could do some experimentation to see what the variation is.) The site referenced above just says "don’t" with no real explanation – can anyone explain why I should not specify the ICC profile in both places?

—

| James Gifford * FIX SPAMTRAP TO REPLY |
| So… your philosophy fits in a sig, does it? |
| Heinlein stuff at: www.nitrosyncretic.com/rah |

James Gifford wrote:

I understand the first two alternatives, but I don’t understand why both PS’s print management and the driver’s color management should be enabled.

Sorry, "should not be enabled."

—

| James Gifford * FIX SPAMTRAP TO REPLY |
| So… your philosophy fits in a sig, does it? |
| Heinlein stuff at: www.nitrosyncretic.com/rah |

From: James Gifford

I understand the first two alternatives, but I don’t understand why both PS’s print management and the driver’s color management should be enabled.

Sorry, "should not be enabled."

Try this experiment … open any image and select a printer profile via a soft proof (View > Proof Setup > Custom …) and open the Info palette. Set the first eyedropper in the Info palette to ‘RGB color’ and click on the second (rightmost) eyedropper and change it to ‘proof color’.

Now select the eyedropper tool and hold it over a bright color in your image and you’ll see the RGB color (the number in your file) is different than the Proof color (the second set of numbers). This is what the file RGB values will be translated to in order to get the closest possible match (depending on the accuracy of the profile) when it’s printed.

If you do either alternative (either Photoshop OR the Epson driver controls the color management) you will get this translation. If you have *both* Photoshop and the Epson driver enabled for color management this translated number gets translated a second time and you are screwed.

To make it more concrete, open a blank file and select Epson’s PGPP profile to proof with ("SP2200 Premium Glossy_PK" in the menu). Fill a portion of the file with say 183/18/25 (red off a test pattern I have) and you should see a proof triplet of 239/25/70, which will be the correct values to feed to the printer if the profile is correct. Either Photoshop or Epson should give you this number. Fill another box with this translated value and you’ll see that a second (incorrect) translation of the proper RGB triplet gives you 255/53/126.

So basically you wanted 239/25/70 and got it with a single translation but if it gets translated again it’s the wrong number, in this case 255/53/126.

The tell-tale sign of this is a strong magenta cast over the entire image.

http://www.creativepro.com/story/feature/13605.html has a good explanation of why the translation is necessary in the first place.

Bill

Bill, thanks for the simple explanation. For the longest time, my problem was grasping the concept of resolution. Once the light bulb went on there, the problem was with color management…getting what I see on my monitor to come out of my printer. This helps a lot. Thanks again.
—

Linelle

"Bill Hilton" wrote in message

From: James Gifford

I understand the first two alternatives, but I don’t understand why

both

PS’s print management and the driver’s color management should be enabled.

Sorry, "should not be enabled."

Try this experiment … open any image and select a printer profile via a

soft

proof (View > Proof Setup > Custom …) and open the Info palette. Set

the

first eyedropper in the Info palette to ‘RGB color’ and click on the

second

(rightmost) eyedropper and change it to ‘proof color’.

Now select the eyedropper tool and hold it over a bright color in your

image

and you’ll see the RGB color (the number in your file) is different than

the

Proof color (the second set of numbers). This is what the file RGB values

will

be translated to in order to get the closest possible match (depending on

the

accuracy of the profile) when it’s printed.

If you do either alternative (either Photoshop OR the Epson driver

controls the

color management) you will get this translation. If you have *both*

Photoshop

and the Epson driver enabled for color management this translated number

gets

translated a second time and you are screwed.

To make it more concrete, open a blank file and select Epson’s PGPP

profile to

proof with ("SP2200 Premium Glossy_PK" in the menu). Fill a portion of

the

file with say 183/18/25 (red off a test pattern I have) and you should see

a

proof triplet of 239/25/70, which will be the correct values to feed to

the

printer if the profile is correct. Either Photoshop or Epson should give

you

this number. Fill another box with this translated value and you’ll see

that a

second (incorrect) translation of the proper RGB triplet gives you

255/53/126.

So basically you wanted 239/25/70 and got it with a single translation but

if

it gets translated again it’s the wrong number, in this case 255/53/126.
The tell-tale sign of this is a strong magenta cast over the entire image.
http://www.creativepro.com/story/feature/13605.html has a good

explanation of

why the translation is necessary in the first place.

Bill

Bill Hilton wrote:

The tell-tale sign of this is a strong magenta cast over the entire image.

So… the ICC correction is being applied twice if it’s specified as PS’s output and in the printer control? Okay, that makes sense.

As someone else already noted, thanks for a very lucid explanation. I really hate having instructions or tutorials just say "don’t do this" without at least attempting to explain the "why," and you’ve filled that in nicely.

—

| James Gifford * FIX SPAMTRAP TO REPLY |
| So… your philosophy fits in a sig, does it? |
| Heinlein stuff at: www.nitrosyncretic.com/rah |