Inkjet RIP question

[Teh_Bm]

Addition:

MatchPrint EQUALS Offset Print
My Monitor EQUALS My Epson

My outputs are NOT EQUAL to others outputs
All devices here are custom profiled.

When my Eye One measures CMYK process printed color from offset and matchprint It gives a value in Lab which converted to CMYK gives values that are not correct. It even shows Yellow and blacK percentage which do not exist in that color.

When measured with SpectroEye it shows Lab value with converted to CMYK gives almost exact match to numbers in the file I gave for printing.

More and more I suspect my Eye One device has a problem.

I see no alternative but to "borrow" someones calibration device and compare its results to mine.

I have an option to "borrow" other guys EyeOne which he uses for for getting sucessful calibration of his monitor. I will also ask my dealer to inspect my EyeOne, maybe even ask him to temp pass me a test example of Spectrolino to run some tests with that device too.

Or maybe Im messing something while adjusting brightness in calibration process. Brightness setting is very tricky to get both arrows match while calibrating monitor. In two different calibration runs (one after another) i get totally different brightness settings on my monitor and yet in both runs, those black and white arrows show that Brightness setting is OK. Something definitely is wrong here.

tehBM

[AndrewRodney]

>Yes, color values in the file and the output from the offset is correct. My >monitor display is not correct.

Then why does the Epson and display match? You said you made a profile for the Epson using the EyeOne so I assume you have the Pro version. What software did you use with it to build the printer profile and how many patches? It doesn't make sense that the Epson to display works but the Matchprint to display doesn't IF the display is off. It's possible but odd.

> I figure that my monitor profile and my epson are off. Only two devices in
> whole process that are calibrated by my Eye One. Could it be that I got
> malufunctional Eye One device? It passes all self diagnostic tests though.

If you did, the output profile you made would produce a bad looking print.

>> How do you know the offset is correct?
> SpectroEye measurement shows it is. MatchPrint proof made at totally different
> place, (visually) tells me that offset print is ok.

I'm not sure how you feel this is true. But that's not important at this point.

> Epson is profiled, Im printing with PhotoShop using "Print With Preview".
> Image is RGB Adobe1998, Im using custom profile for my epson to print by
> setting it up as destination space in "Print with Preview" dialog. I think
> that Intent I use for print is Relative.

So this is just a nice RGB to RGB conversion to make a nice looking print. Fine. You're happy with the print and you're happy with the print to screen matching. Now what I want you to do is this. Take that Adobe RGB file and convert to CMYK with the profile you feel is working for you. THEN convert from CMYK to RGB using your custom Epson RGB profile (use an Absoulte Colorimetic intent). Does this Epson print look anything like the Matchprint?

> My display profile is bad. I can manually tune the monitor to have it kinda
> correct. I just tweak the brightness of the monitor a bit and only then I get
> the correct picture that looks like offset print.

You don't want to do that just yet until you are sure the display is the culprit. It very well might be but then why does the Epson print and the display match. You are saying the match when you setup a soft proof with the Epson profile (the Adobe RGB file will not and should not match your print).

> Yes, Epson gives me what I see on the monitor. Epson and Monitor profiles were
> made using same EyeOne Pro UVCut device.

That's why I'm not convinced you have a problem on your end. If the display profile were bad, you'd get a nice Epson print and a bad preview. Or a bad print AND a bad previews. That's not happening.

> CMYK output on offset print and matchprint is much darker.

Which could be due to the profile! If you have a wonderful profile but it doesn't accurately describe how the CMYK process really works, you'll see a great preview and get an off print. The profile expects the device to behave a certain way. It has no way of knowing the device isn't doing that. So it shows you what it expect to happen. Then you send those numbers to the output device and you get a darker print. It could be the profile. If you can make a custom RGB profile, why not try making a custom CMYK profile for this device?

> I have images saved with Adobe RGB 1998.
> - When printing to Epson I convert image from Adobe RBG 1998 to my custom
> epson printer profile (Relative colorimetric).
> - When printing to file for CMYK CtP I use custom printers ICC profile to
> separate images in PhotoShop. Adobe RGB 1998 -> CMYK profile I got from
> printer.

That's correct. But you have apples and oranges here. We still don't know if the issue is with the profiles you are building or the profile you were supplied.

> I still feel my Eye One is not measuring as it should.

That's possible. Still doesn't explain why the Epson is working as it should. You might want to either profile a different device and test it or try a different paper to the Epson to see if you can repeat the good print to good display matching. You should be viewing all prints under a D50 light box too of course.

> They say it is. Its ICC profile file. I beleve 'em its custom one because they
> usually send a new profile for every print job, regarding what paper someone
> is using, what dyes they have, etc.

That makes no sense. What's the dyes got to do with anything. You're printing to an offset press right? You're getting a contract proof made (what you're calling a Matchprint). The MatchPrint should be stable and behaving the same once the shop sets it up and runs it from then on. They may have to do daily calibration to keep it consistent but that's not your problem.

> Someone could think that they made incorrect profile for my exact printjob,
> but nope.. other process outputs like Matchprint make somewhere else shows
> they did ok.

But the initial RGB to CMYK conversion could be off and if both Mathprint's behave the same they will both match and look poor. That's why I still suspect it could be their profile but more testing is needed. The conversion can be off and since both MatchPrint devices match each other but do NOT behave as the profile assumes, you get two identical but wrong looking prints. Now if you are getting GOOD Matchprints but the preview of your RGB to CMYK file doesn't match, that's a different story.

> Altho Euroscale when looked as 3D and overlapped with profile I got from
> printer matches almost perfectly. Printer said I can easily use Euroscale
> without fear, but I asked em for custom profile which I got from em and used
> it at the end.

That doesn't insure or guarantee anything to do with those profiles producing the same color matching to the same device. It only tells us they have the same gamut size.

[Teh_Bm]

Heres the idea.
I have one ad for some kind of airliner.
Ill open it in photoshop.
Ill put offset Printed ad next to it. Ill put inkjet print next to it.
Ill put MatchPrint next to printed ad.
Ill take that same ad from some other magazine done by totally different printer than mine and sent to that totally different printer by totally different studio.

Then ill shoot it with digital camera.
You maybe wont see correct colors but youll see relations between colors.

Youll see that CMKY prints look identical, weather from me or someone else.

Youll see that my monitor image is much brighter. Youll also see that my inkjet print matches my monitor image but not printed or matchprint versions.

If everything is ok, how come my EyeOne reads Lab whose separation has yellow and black, which originally that color doesnt have (as sepatarions)?

tehBM

[ January 18, 2003: Message edited by: tehBM ]

[Teh_Bm]

>> Then why does the Epson and display match? You said you made a profile for the Epson using the EyeOne so I assume you have the Pro version. What software did you use with it to build the printer profile and how many patches? It doesn't make sense that the Epson to display works but the Matchprint to display doesn't IF the display is off. It's possible but odd.

I made printer profile using "Eye One Match 1.3" software.
It makes no sense, unless something is wrong with that exact EyeOneProUV device (IMHO).
When making profile, I used target that comes installed with EyeOne program. Its a TIFF of target used for calibrating Inkjet printers such as mine.


>> If you did, the output profile you made would produce a bad looking print.

Maybe, I wouldnt know. But fact is that my monitor colors are brighter then they should be.

Lets for a second drop Inkjet issue and focus on monitor profiling. Lets assume EyeOne device is OK. I shouldnt be able to go wrong while making monitor profile.
There are 3 settings to adjust and 2 settings to pick. White point and gamma to pick. For me its D65 and 2.2. Contrast to max, Brightness minimum and then raise till arrows allign. [img]images/icons/blush.gif[/img]nitor white point setting using OSD.

Only setting Ive noticed has some inconsistency is Brightness. Once I can set it up to like 40%, next time it says its OK at 28%.

Whats with that?

>> So this is just a nice RGB to RGB conversion to make a nice looking print. Fine. You're happy with the print and you're happy with the print to screen matching. Now what I want you to do is this. Take that Adobe RGB file and convert to CMYK with the profile you feel is working for you. THEN convert from CMYK to RGB using your custom Epson RGB profile (use an Absoulte Colorimetic intent). Does this Epson print look anything like the Matchprint?

Ill do that 1st thing on monday at work, then tell you the result.
Three Questions. Why Absolute colorimetric? What intent to use in 1st conversion from AdobeRGB to CMYK? Are we aiming to get image on screen that will look like Matchprint?

>> You don't want to do that just yet until you are sure the display is the culprit. It very well might be but then why does the Epson print and the display match. You are saying the match when you setup a soft proof with the Epson profile (the Adobe RGB file will not and should not match your print).

I wasnt too clear on that part. Adobe RGB 1998 with soft proof in photoshop set to Epson profile are very very similar.
Its not a perfect match tho, but they are very similar.


>> That's why I'm not convinced you have a problem on your end. If the display profile were bad, you'd get a nice Epson print and a bad preview. Or a bad print AND a bad previews. That's not happening.

Thats what Im suspecting. That my print and my monitor preview are incorrect.
Lets assume my monitor profile is incorrect. Since the image is corrected on my incorrect monitor and printed on my incorrect Epson where both profiles were done by same device, can we assume they both have same amount of 'incorecctness'?
So basically in the incorrect working conditions for all devices my images look as they should, but when ported to correct devices they fall apart, go dark. Can that be a case? If yes then only thing that is related to all incorrect devices is my EyeOne. That was my logic.


>> The profile expects the device to behave a certain way. It has no way of knowing the device isn't doing that. So it shows you what it expect to happen. Then you send those numbers to the output device and you get a darker print. It could be the profile. If you can make a custom RGB profile, why not try making a custom CMYK profile for this device?

Then why SpectroEye measurement shows printed on paper color is almost exact match?
Yes, if I would make the profile using my EyeOne, I feel that could work. But in my opinion i would just make another incorrect profile which implemented into my full system of incorectness MIGHT WORK. [img]images/icons/smile.gif[/img]
Yet Id prefer correct settings from start to an end.

>> That's correct. But you have apples and oranges here. We still don't know if the issue is with the profiles you are building or the profile you were supplied.

True, but we'll find out. [img]images/icons/wink.gif[/img]


>> That's possible. Still doesn't explain why the Epson is working as it should. You might want to either profile a different device and test it or try a different paper to the Epson to see if you can repeat the good print to good display matching. You should be viewing all prints under a D50 light box too of course.

AS I was saying, Epson calibration is done with same device as my monitor. Maybe Epson is dragged into that set of incorrect and differently set values, making it output correctly. Which works while my own calibration calibrated devices are involved, but when my files go to outside world it all falls apart.
It would be nice for me to have D50 box, but thats not an option at the moment.


>> But the initial RGB to CMYK conversion could be off and if both Mathprint's behave the same they will both match and look poor. That's why I still suspect it could be their profile but more testing is needed. The conversion can be off and since both MatchPrint devices match each other but do NOT behave as the profile assumes, you get two identical but wrong looking prints. Now if you are getting GOOD Matchprints but the preview of your RGB to CMYK file doesn't match, that's a different story.

01.41am - Syntax Error - too much data - Ill parse it tommorow, my brain is melting.

Anyway, as you say more testing is needed, but I feel that, as you said here at the end, Matchprints are good and my preview (soft proof) is wrong.
Anyway as I mentioned in that post just above, I got one ad that was printed in that same printjob, that was also prepared for print by some other studio and then printed by some totally different printer.
Its kinda hard beleving you can have 4 errors on the one side and me on the other side as only one thats has right and correct setting.

Matchprint (as you call it Contract Proof, but I made it after CMYK offset printjob) EQUALS cmyk offset Printjob
numbers on SpectroEye match the ones they should, and some other printer/studio printed same page thats very similar to this printed version

as opposed to my monitor and inkjet that work as closed system but have generally MUCH MUCH lighter color.

Andrew thank you alot for all this help. Im rather pushy in making it my own mistake at this issue, but somehow i sense its fault at my end due incorrect monitor calibration.

Print run for that job was rather big (print run/num of copies) and for the 1st time in my life I relied on Calibration device. Damage is here and all I want is not to repeat the same mistake again and at the same time I wanna use EyeOne, since I know that can be the right path.

thnx for all this, lets not stop this discussion.


tehBM
p.s. English is not my native language, so please bear with me.

[Teh_Bm]

we could ask Rob to move this thread to Color Management.

Nite nite, read you tommorow.


tehBM

[Teh_Bm]

Thank you for all the input.

I already can quite precisely Match what I see on the monitor to what I get on the Inkjet. I wanted to take it a step firther and simulate halftones.
I have Eye One Pro (UV) with Eye One Match 1.3 (thinking of getting ProfileMaker Pro software)

There is one other problem. I use Eye One to make monitor profile. Profile Im getting using EyeOne is good and if I make printer profile using Eye One, what I see on screen -almost- perfectly matches my Inkjet output.

**BUT**

... what I see on screen, doesn't match offset print. Printed images are much darker and less saturated. I can solve that by manually adjusting brightness to offset printed image (visually). Then my monitor matches offset print and not my inkjet printer. At first I was thinking that my offset printer, where I print my work, messes with something. But actually they do their job correctly and my Eye One produces incorrect monitor profile. I made films and matchprint at some other place than my printer and it perfectly matches to what printer is printing, therefore making Eye One monitor profile - incorrect.

Also, if I measure offset printed using GretagMacbeth SpectroEye, SpectroEye (which is top class very precise instrument) shows that my offset printer is printing color with dE of approx 1.0 - 2.2 (that is very accurate in my opinion).

I'm kinda tapping in the dark. I can make EyeOne Profile then manually lighten the screen using OSD to get acurately profilled monitor which matches printwork and doesnt match my inkjet. -OR- I can make EyeOne calibration to match my inkjet and monitor therefore not matching offset print.

I decided to go for matching offset since its my final output.
Bottom line is that this is not how Eye One is supposed to work and thats not what i paid for when i was purchasing it.

tehBM

p.s. Can Eye One Pro (UV) measure offset printed color on the glossy paper?
If yes, then my Eye One using i1Share software doesnt give me correct readout of the color on the paper.

Please advise me if anyone is familiar with the problem or may guess what is the problem.

thanks

[AndrewRodney]

-->... what I see on screen, doesn't match offset print.

Do you have a good (Better custom) ICC profile for the offset press? If not, how can it match?

In order to cross render to an Epson (or any printer), you need a profile for the printer and a profile for the device you are trying to proof on that Epson. So you need a good CMYK profile for the final piece. You need to do your RGB to CMYK conversion for this process. Then depending on what driver you'll be using to the Epson, you need to do a CMYK to Epson conversion (CMYK to RGB if you are using the Epson driver since Quickdraw and GDI printers can't deal with CMYK data).

-->Also, if I measure offset printed using GretagMacbeth SpectroEye, SpectroEye (which is top class very precise instrument) shows that my offset printer is printing color with dE of approx 1.0 - 2.2 (that is very accurate in my opinion).

I'm not following you here. deltaE is only a measure of difference between to spectral readings. You've got a measurement of the press but what are you comparing that to in order to get this deltaE reading?

You've made a custom ICC profile for the CMYK device in question? The output is fine but the file doesn't match the output on screen?

--> Can Eye One Pro (UV) measure offset printed color on the glossy paper?
If yes, then my Eye One using i1Share software doesnt give me correct readout of the color on the paper.

Sure. What makes you say the paper reading in Share is off?