how to measure dmax on paper

[digitaldog]

In the meantime, here are scans of two prints, one printed at 1440dpi, the other at 2880dpi.

The 1440dpi print is darker than the 2880dpi print.

You're utterly missing the point! This discussion is about printing targets for profiles. See « Reply #18 on: July 31, 2019, 07:40:59 am »
NOT if one set of DPI produces a visual difference. I'd expect there would be otherwise what's the point of differing settings. You're completely missing the difference in printing solid square patches for profile creation and the instructions provided by myself and others in printing said targets and how to test all this.
I've just output 918 patches since it fits one page, on a 3880 at both 1440 and 2880. To the naked eye, the appear identical as some of us have attempted to report to you. After they dry down, I'll measure them and provide a colorimetric report as I outlined one, with the tools and knowledge should do so for analysis.
If you're sending targets for two profiles with two different resolutions, you're wasting your time, ink and paper along with whoever is measuring them. That you SEE a difference using two settings on images is utterly off topic and to be expected.
Do you understand? 

[elliot_n]

I am not missing the point. The visual difference I would expect when changing from 1440 to 2880 would be a slight increase in image acuity — and certainly not a change in image brightness.

Thank you for running a test.

[digitaldog]

I am not missing the point. The visual difference I would expect when changing from 1440 to 2880 would be a slight increase in image acuity — and certainly not a change in image brightness.

As one who doesn't create profiles, or understands the results of the colorimetry of the patches used to build such profiles, or one that recognizes that the targets appear identical, yeah, you're missing the point as asked and answered in post #18. But that's OK, the person who asked about differing targets for profile creation got the correct answers from several sources.

[elliot_n]

Have they dried yet?

[digitaldog]

Have they dried yet?

Pigmented inks? No, they require (news flash) 24 hours for minimum deltaE colorimetric testing. At least those of us who've printed and measured targets right after output, an hour later and 24 hours later colorimetrically understand. But they still appear IDENTICAL (so much for visual testing). 

UltraPremGlossy first meaurment compared to 1 hour later:

dE Report
Number of Samples: 1728
Delta-E Formula dE2000
Overall - (1728 colors)
--------------------------------------------------
Average dE:   1.07
Max dE:   8.59
Min dE:   0.04
StdDev dE:   0.90


Best 90% - (1554 colors)
--------------------------------------------------
Average dE:   0.84
Max dE:   2.06
Min dE:   0.04
StdDev dE:   0.45


Worst 10% - (174 colors)
--------------------------------------------------
Average dE:   3.19
Max dE:   8.59
Min dE:   2.06
StdDev dE:   1.14
--------------------------------------------------


UltraPremGlossy 2nd meaurment (1 hour dry down) compared to 24 hours later:


dE Report
Number of Samples: 1728
Delta-E Formula dE2000
Overall - (1728 colors)
--------------------------------------------------
Average dE:   0.21
Max dE:   0.95
Min dE:   0.01
StdDev dE:   0.12


Best 90% - (1554 colors)
--------------------------------------------------
Average dE:   0.18
Max dE:   0.35
Min dE:   0.01
StdDev dE:   0.08


Worst 10% - (174 colors)
--------------------------------------------------
Average dE:   0.45
Max dE:   0.95
Min dE:   0.35
StdDev dE:   0.11
--------------------------------------------------

[elliot_n]

Perhaps my two prints will look the same when I view them tomorrow.

[Doug Gray]

In the meantime, here are scans of two prints, one printed at 1440dpi, the other at 2880dpi.

The 1440dpi print is darker than the 2880dpi print.

You shouldn't expect prints made with different settings to exactly match and, how close they are depends on individual variations in printers and how well the OEMs match the colors for different weave densities.

I have an Epson 9800 which offers settings of 1440x2880 as well as 720x1440. I always print using the same printer settings used to make the profile unless I'm just rushing something out like a large 60"x40" draft and need faster printing.

I also see a difference on the 9800 between the two modes on glossy with no other settings changed but 720x1440 -> 1440x2880. But it goes in the opposite direction you describe. The highest setting yields a 3.7 lower L* on an arbitrary gray patch. Average dE 2000 of the two targets (480 patches each) was 2.2

That's going to make a slight difference between prints viewed side by side.

Generally, follow the rule to only print using the same printer/paper/settings you made, or had made, a profile for. You might luck out using the same profile with different settings.

[elliot_n]

Thanks Doug. Would making two custom profiles, one for each print resolution setting, fix this problem?

[digitaldog]

Did you read this?

02/06/2007 1400 DPI vs. 2800 DPI

According to Epson's claim, the new screening technology incorporated in the 3800 "compensates for various ink densities between resolutions - allowing for a single ICC profile to be used per media type".

In practical terms this means that you no longer need two separate custom profiles, one for printing at 1440dpi and one for printing at 2880dpi.
I never really liked the 1440dpi option of the Epson printers, other than for its speed. I always complained about it because to my eye it printed with some noticeable microbanding.
I therefore read Epson claim with much interest, and I was eager to verify it with my own measurements.

To do this I created a small color Testchart -283 color patches- and I printed it using both the 1440dpi and the 2880dpi resolution modes.


I then measured the Testcharts and compared the measurements files.
An acceptable correlation between the two measurements would be contained within delta E<1.
This limit is considered the (average) minimum distance that two colors must have for the human eye to be able to separate them.
The results are shown below:

By using different delta E formulas the results varies but the best formulas to determine small colors differences are the delta E CMC (2:1) and the newest delta E 2000. With both these formulas the Total Average is well within the 1 limit (0.82 using delta E CMC (2:1). Also the distribution of the yellow boxes (which indicate larger color differences) seems reasonably random, with the exception of the vertical column on the right, which contains low chroma colors and their difference should not be as apparent to the eye as it is to the spectrophotometer used to measure the samples.

The conclusion of this simple test is that Epson claim is right on target.
It is perfectly acceptable to use one ICC profile for both 1440 and 2880 dpi printing.

Even wet, the differences on this end thus far, between the two are an average of 0.9 which is invisible. 

[Doug Gray]

Did you read this?

02/06/2007 1400 DPI vs. 2800 DPI

According to Epson's claim, the new screening technology incorporated in the 3800 "compensates for various ink densities between resolutions - allowing for a single ICC profile to be used per media type".

In practical terms this means that you no longer need two separate custom profiles, one for printing at 1440dpi and one for printing at 2880dpi.
I never really liked the 1440dpi option of the Epson printers, other than for its speed. I always complained about it because to my eye it printed with some noticeable microbanding.
I therefore read Epson claim with much interest, and I was eager to verify it with my own measurements.

To do this I created a small color Testchart -283 color patches- and I printed it using both the 1440dpi and the 2880dpi resolution modes.


I then measured the Testcharts and compared the measurements files.
An acceptable correlation between the two measurements would be contained within delta E<1.
This limit is considered the (average) minimum distance that two colors must have for the human eye to be able to separate them.
The results are shown below:

By using different delta E formulas the results varies but the best formulas to determine small colors differences are the delta E CMC (2:1) and the newest delta E 2000. With both these formulas the Total Average is well within the 1 limit (0.82 using delta E CMC (2:1). Also the distribution of the yellow boxes (which indicate larger color differences) seems reasonably random, with the exception of the vertical column on the right, which contains low chroma colors and their difference should not be as apparent to the eye as it is to the spectrophotometer used to measure the samples.

The conclusion of this simple test is that Epson claim is right on target.
It is perfectly acceptable to use one ICC profile for both 1440 and 2880 dpi printing.

Even wet, the differences on this end thus far, between the two are an average of 0.9 which is invisible. 

Perhaps Epson improved things with the 3880 or perhaps you have a really good one but Epson also claimed that their profiles for the 9800 could be used with any resolution. See this:

https://epson.com/faq/SPT_C595001UCM~faq-64641?faq_cat=faq-topFaqs

But, at the end and apparently in contradiction with their own statement, they added some profiles specifically for 2880 resolution. I'm not sure when the additional profiles were added but they are only a subset of papers. I still have all the Epson OEM profiles from when I bought it. There are no 2800 resolution profiles. It's one size fits all.

OTOH, even though there are clear, if slight unless viewed side by side, differences, they pale in comparison to the differences made printing the same, in gamut print, with the Canon 9500II.

But if I print the same, in gamut image with custom profiles on the 9500 and 9800 I see no difference at all unless I look for differences in bronzing and gloss differential by tilting the paper against the light.

[elliot_n]

Did you read this?

Yes, please see my earlier post:

https://forum.luminous-landscape.com/index.php?topic=131486.msg1122314#msg1122314

My 3880 is clearly not as well-behaved as the 3800 under discussion at Digital Outback Photo.

[digitaldog]

Yes, please see my earlier post:

https://forum.luminous-landscape.com/index.php?topic=131486.msg1122314#msg1122314

My 3880 is clearly not as well-behaved as the 3800 under discussion at Digital Outback Photo.

But mine is.....

[elliot_n]

3800: Match
3880 (yours): Match
3880 (mine): Don't Match
9800 (Doug's): Don't Match

I call it a draw.

[digitaldog]

3800: Match
3880 (yours): Match
3880 (mine): Don't Match
9800 (Doug's): Don't Match

I call it a draw.

Sorry no. Looking just at the folders of spectral data files for 3880's I've profiled for customers, there are 105 there. Got 39 for the 3800. And then there's all those folks who used the profiles I built for Epson for Exhibition Fiber.
http://pixelgenius.com/epson/
Not a draw. 
Nothing to do with matching. Having to do with how to print targets for creation of profiles. I'd love to suggest to customers they need multiple profiles for differing resolutions, I'd make a lot more money. But the colorimetric facts and a bit of honesty will not allow me to do so. So not matching; printing profile targets. Not what the image looks like afterwards but we've been over this ad nauseam.

[elliot_n]

Are you saying that whilst my two prints of your test file look different (1440 is darker than 2880), if I were to print a profile target through ACPU, at 1440 and 2880, the two prints would look the same? That the driver behaves differently when hooked up to ACPU rather than Photoshop?

[digitaldog]

Are you saying that whilst my two prints of your test file look different (1440 is darker than 2880), if I were to print a profile target through ACPU, at 1440 and 2880, the two prints would look the same? That the driver behaves differently when hooked up to ACPU rather than Photoshop?

The data (wet) already suggests this fact with respect to solid patches found on printer targets. As does the web pages referenced. I don't know why you find this colorimetric data so difficult to either accept or understand. Again, WET, the two targets measured here today have an average dE 2000 of under 1. That will be even lower but still an invisible difference tomorrow!

[elliot_n]

I don't find the colorimetric data difficult to either accept or understand. I accept that your 3880, and the 3800 tested on the linked website, print the same whether 1440 or 2880 is selected. But my 3880 does not. I am simply trying to figure out why this is the case, and whether separate custom profiles for 1440 and 2880 would ameliorate the problem.

[digitaldog]

I don't find the colorimetric data difficult to either accept or understand. I accept that your 3880, and the 3800 tested on the linked website, print the same whether 1440 or 2880 is selected.

FOR PRINTING SOLID COLOR PATCHES LARGE ENOUGH TO MEASURE ON A SPECTRO TO MAKE PRINTER PROFILES!!!!! That's it. I don't know, shouting or otherwise how to make this clearer and stated below (or above  ) what you see on an image, which is composed of thousands or hundred of thousands of single color patches scattered all over an print of an image can visually differ.
The topic AGAIN IS PRINTING TARGETS FOR PROFILES. YOU DO NOT NEED TO PRINT TWO (1440/2880) NOR PRODUCE TWO PROFILES.

[elliot_n]

FOR PRINTING SOLID COLOR PATCHES LARGE ENOUGH TO MEASURE ON A SPECTRO TO MAKE PRINTER PROFILES!!!!! That's it. I don't know, shouting or otherwise how to make this clearer and stated below (or above  ) what you see on an image, which is composed of thousands or hundred of thousands of single color patches scattered all over an print of an image can visually differ.
The topic AGAIN IS PRINTING TARGETS FOR PROFILES. YOU DO NOT NEED TO PRINT TWO (1440/2880) NOR PRODUCE TWO PROFILES.

No need to shout. There are solid colours on the test chart I printed - e.g. the grayscale step wedge along the top. They don't print the same when I switch from 1440 to 2880.

[digitaldog]

No need to shout. There are solid colours on the test chart I printed - e.g. the grayscale step wedge along the top. They don't print the same when I switch from 1440 to 2880.

Then something is off on your end. I'm not going to dig any deeper into your possible problems (visual or mechanical) but did answer the question asked, consistently and upon agreement with other's who also colorimetrically tested several printers that printer targets at differing resolutions are below the threshold of a visual difference (under dE of 1).
I don’t know if you are purposely trying not to understand this, or if you are really struggling with it.