What are the pure Epson pigment values ?

[larkis]

I want to be able to print the pure color from each cartridge on my Epson 4800 (sort of like the epson test sheet). Where can I find the actual rgb values + profile to make this happen ?

[hugowolf]

I want to be able to print the pure color from each cartridge on my Epson 4800 (sort of like the epson test sheet). Where can I find the actual rgb values + profile to make this happen ?

I don't think that will happen.

I doubt it is possible through the driver, you would have to use a RIP.

You could open empty cartridges and brush the ink onto a neutral paper like Epson SemiMatte Commercial Proofing paper. (There is about 10 ml of ink left in an empty cartridge.) You could measure that with a spectrophotometer, which would give you Lab values to probably to four significant figures. But with the same spectrophotometer, a second reading is most unlikely to be identical with the first. And even if you got an accurate reading, taking the Lab values into a program for printing, you would end up with integer Lab values. Photoshop, for example only does integer Lab values.

Integer Lab to integer RGB introduces fairly high rounding errors. And RGB values in what working space?

Qimage has a purge pattern print function. It isn't 100% pure, but it will print a high percentage of each color.

Brian A

[aduke]

Take a look at http://www.marruttusa.com/printer-maintenance/inkjet-printer-purge-files.php#col . I downloaded the file from Marruttusa.com some while ago to build my own printer purge file. The ink usage log from my 4880 indicates that most of the values are dead-on.

Alan

[huguito]

If you are looking for samples of color to exercise an individual nozzle you could just take a picture of the color tab in each cartridge next to a color target like the Greta Mackbet to correct for light cast and read it on ACR placing the cursor over the colored marker of the cartridge.

I imagine that epson would have printed those tabs on the ink cartridges very close to the actual color.

Not very scientific but the ones I did print as close as I can distinguish from the cartridges and the colors on the nozzle test.

Hope it helps

Hugo

[larkis]

Take a look at http://www.marruttusa.com/printer-maintenance/inkjet-printer-purge-files.php#col . I downloaded the file from Marruttusa.com some while ago to build my own printer purge file. The ink usage log from my 4880 indicates that most of the values are dead-on.

Alan

Thank you for this, it's what I have been looking for. What settings are you using in photoshop to print this file ? Is there a specific profile one should be using ?

[John Caldwell]

I created a PSD Pro Photo image based on the hex values shown in the linked file at MarruttUSA and printed the file from LR, with color management off. The file was printed on the Epson 4900. All colors looked as I expected, with the exception of green. Green rendered as very, very dark - with the obvious addition of black, yet other colors looked pretty much as expected (including orange). By that I mean all colors, with the exception of green, looked as though they were very close, if not accurate, to the pure Epson ink tones. When I print the same file from LR, using an appropriate ICC profile for printer and paper, green renders as expected.

So thus far, I'm quite happy with this "purge" file using the hex values communicated at the MarruttUSA site, but I can't figure of the green ink…

John Caldwell

[hugowolf]

Out of general interest...

I created a PSD Pro Photo image based on the hex values shown in the linked file at MarruttUSA ...

What color space did you assign to it?

The file was printed on the Epson 4900. All colors looked as I expected, with the exception of green...

But the orange looked ok?

Also, have you examined the prints with a loupe or microscope to see what the dot pattern actually contains?

Brian A

[John Caldwell]

Yes Brian, orange and other colors looked correct - with the exception of green which looked very, very dark. It was clearly printing green and black nozzles in the green column.

The document was created in PS as Pro Photo RGB, and was printed from LR5 using "Colors Managed By Printer", and Color Management was OFF in the Epson driver.

Does this sound like a good color management workflow?

When the same file is printed from LR using an ICC profile that is specific for the printer and paper, the green column certainly looks green, but none of the colors, except the blacks, are true. The shift in colors is accurately predicted by soft proofing the document within LR.

Any thoughts?

Many thanks,

John-

[hugowolf]

From what I understand, there is no way of fully turning off color management in Lr or Ps since CS4. In other words, 'printer manages color' and 'no color management' in the driver, isn't no color management. This is why Adobe had to release ACPU:

https://helpx.adobe.com/photoshop/kb/no-color-management-option-missing.html

It shouldn't matter that much, and I am doubtful that really using no color management would have that much affect. But it would be worth a try.

It is really highly improbable that the purge file patches are even 95% pure, and it wouldn't take much black to create a very dark green.

I think one of the problems of a purge patch file for O and G, looking at ink usage data, the driver appears to only use these two inks when no other possible combination of inks will create the requested color. In other words they are mostly used when the color would otherwise be out of gamut. RIPs can generally do better at this.

Brian

[howardm]

If you have a Mac, then the ColorSync Utility has the ability to print w/o any color management.

In fact, in the print pulldown, it's called 'Print as Target'

[martin0reg]

...
I doubt it is possible through the driver, you would have to use a RIP.
...

I think QTR (free) could do that in "calibration mode" ...although the patterns are small, not made for purging..

[cybis]

I use Quad Tone Rip (shareware) to print perfectly pure purging patterns:  http://lucbusquin.com/content/purge-all-epson-x900-ink-channels

[hugowolf]

I think QTR (free) could do that in "calibration mode" ...although the patterns are small, not made for purging..

You could, however, repeat it?

Brian A

[martin0reg]

You could, however, repeat it?

Brian A

While I agreed with your opinion I just wanted to point out that there is a RIP (not free but shareware indeed) which could help.
Now there is already a link in the previous post...you may search for "qtr calibration mode" to find more...

[John Caldwell]

I built a "purge file" in PS as a Pro Photo file from the hex values posted in the link above, and it prints very nicely with the exception of the green ink channel. I discussed this green failure a few replies above. The print is the same whether printed from LR5, with Color Managed by Printer and Color Off in the driver; or if the same file is printed from the Epson Color Print Utility. All the channels look "right" to my eye, meaning there is favorable comparison between the colors laid down in a nozzle check with the output of the purge file - with the exception of Green. The green, as you'll see in the attached scan, is very, very dark as if black is printing in that patch along with color. If the purge file is printed from LR5 utilizing a paper-sepecific ICC profile, the Green lane certainly looks green rather than black, but all the colors skew away from "accuracy", where accuracy is defined as matching the ink colors shown in nozzle check.

Attached is a scan of the two prints in side-by-side comparison: Color Management Off, and with paper-specific ICC profile.

My knowledge of color management is thin enough that I do not understand what's going on here wrt the green, or wrt to the conspicuous color shift of the other ink "colors".

John Caldwell

[John Caldwell]

For purposes of comparison, here is the nozzle check included in the print scan:

[hugowolf]

The green I don't know, but the color shift would be for a couple of reasons: the RGB or Hex values define no precise color without reference to a color space. sRGB(0,255,0) Hex 00FF00, is not going to look like AdobeRGB(0,255,0) or ProPhotoRGB(0,255,0).

For the two files I downloaded and examined in Photoshop as un-color managed images, the RGB and Hex values did not exactly correspond with the Marutt stated values. For example the 4900 cyan is stated to be #00FFFF (RGB (0,255,255), but is in fact RGB(0,254,252)

And, with the profile, you would expect a shift (a remapping), that is what the profile is for. Did you print using an absolute colorimetric rendering intent?

Seriously, as others have pointed out, a RIP is the way to go if you want pure ink colors.

Brian A

[John Caldwell]

...Seriously, as others have pointed out, a RIP is the way to go if you want pure ink colors.

Brian A

Not to doubt you in any way. I was just trying to learn why this green rendered in such an unexpected way. The print that did render green in close to the expected way was printed from LR5, engaging the profile for my proofing paper which is BC Vibrance Luster, using the Relative Intent (for no particular reason).

Thanks for your interest, Brian.

John-

[aduke]

My purpose for having the Epson pigment values was so that I could print a page of a single number and know that I would print a page of one of the eight ink in my machine. So, I always print with ACPU and no color management in the printer.

The printer reports that the numbers I posted above achieve that goal in seven of the eight colors, so I continue to use the numbers.

Since I don't print images daily, I build one page with all eight colors on it and print a couple of 8.5 x 11 copies of it on a daily basis an know that all of the channels are being exercised.

Alan

[John Caldwell]

Yes, your rationale made sense and was appealing and this is why I pursuer trying to get a version for the 900 series to work. Puzzle why all the colors look so good except the 900's green.