Epson R3000 poor detail test: is my printer FAULTY?

[mac_paolo]

I've more than satisfied from my printer, especially color-wise.
Today I decided to make a detail test, as I've never been satisfied by the high frequency detail in very sharp images.

I took the test image from the bottom of this page: HERE


Opened in Photoshop, set as 1440 ppi and printed on Ilford Galerie Smooth Gloss (may not the best paper, but very good achievable detail).
Printed at 5760 ppi, unidirectional print. Finest detail checked by default.



Here is the scanned result (Epson V500 @ hardware 4800 dpi):


Having printed at 1440 dpi, the smallest 1px black lines should be 720 black lines/inch, each of them 1/1440" wide.
To me, horizontal 1px and 2px lines are not readable. 3px is starting to look better and the first reasonable pattern is 4px (= 180 black lines/inch, each 1/360" wide).
Vertical lines look even worse, especially black ones.

How would you evaluate these results? I'm asking you to help me figure this out before contacting Epson support.

Thanks,
Paolo

[mac_paolo]

I didn't meant to "steal" the test image. I left the author's name and the URL.
I'll leave a link only.

Update: removed the direct picture link. Added the page URL.

[mcbroomf]

You printed it as a 5x7 as well as 1440 ppi.  It looks like a 600 pixel image, so you uprezed it a godawful amount.  I think that's your problem.

[mac_paolo]

You printed it as a 5x7 as well as 1440 ppi.  It looks like a 600 pixel image, so you uprezed it a godawful amount.  I think that's your problem.

How could the paper size affect the image?
The whole pattern at 1440 ppi is about 1cm long.
I don't understand what my problem should be… 

[elolaugesen]

How did you save the image from his site...   cannot find any files to download

cheers elo

[mac_paolo]

How did you save the image from his site...   cannot find any files to download

Uh? Right click -> Save as? 

[raydee]

Did you change the image mode from Indexed (gif) to 8bit RGB before printing?
There also seems to be some ink bleed, like the ink limit of the paper might be exceeded...

[mac_paolo]

Did you change the image mode from Indexed (gif) to 8bit RGB before printing?
There also seems to be some ink bleed, like the ink limit of the paper might be exceeded...

I didn't set the mode as RGB. May this cause any issue?
I see quite a bit of ink bleed as well. I just followed the Ilford recommended settings for the specific paper.

Here it is the driver setting for Paper configuration:

[raydee]

Set the image mode to 8bit and print again. Then some iteration with less density, just give -10% a shot. If you move the test graphic around inside a page sized canvas in PS, you won't waste too much paper on the fine tuning process, as you can use the same page multiple times...

[mac_paolo]

Set the image mode to 8bit and print again. Then some iteration with less density, just give -10% a shot. If you move the test graphic around inside a page sized canvas in PS, you won't waste too much paper on the fine tuning process, as you can use the same page multiple times...

I'll do some testing on density, and maybe even platen gap.
I'm already printing on different position of the same sheet.

[elolaugesen]

I copied the image in question.   when changed it to 300 dpi it came out 1.86 inches by 1 inch.   

I would never try to print that kind of detail in that small a space nor would I resize

get a good quality image at a good size then print a proper test...    then judge the quality of your printer....

cheers elo

[mac_paolo]

I copied the image in question.   when changed it to 300 dpi it came out 1.86 inches by 1 inch.  

I would never try to print that kind of detail in that small a space nor would I resize

get a good quality image at a good size then print a proper test...    then judge the quality of your printer....

cheers elo

Elo, at 300 ppi the highest frequency become 150 black lines per inch, which is quite low. Any low-end consumer printer is able to achieve that.
I mean high frequency detail: the one you achieve with high end printers, as the R3000.
This may not be the best test overall, never said that, but I'm find interesting result... posting in a few minutes.

[mac_paolo]

New test

I converted the image to RGB mode and made some prints playing around Paper Configuration.
They all have been printed at 5760 ppi, unidirectional print.
Scanned at 4800 dpi as before.

Wide Platen


Standard Platen


Standard Platen, -10% density


To me they are ordered from worst to best.
The biggest improvement in in vertical lines. At the wide platen gap they separate on 6 px, while on the other two already at 3-4px.
The pentacolor test at the top of each sample shows a lot of bleed in the Wide gap, while is a lot better in standard.
I can see an obvious reduction in saturation for the last sample. I'll have to print some real world images.

I use to leave Wide platen gap because I mostly print of Ilford Gold Fibre Silk. I may have to reduce it for semigloss papers.
I still haven't tested the baryta paper as I don't feel it may guarantee the same level of detail.

[raydee]

So the conversion to RGB made indeed quite a difference, given the first specimen is the same as the initial test only converted to RGB.
The standard platen improves things quite a bit, though there is still ink bleeding evident. Maybe it isn't needing the -10% but I'd give some decreasing iterations a try.

This settings should result in an improvement in real life images too...

[mac_paolo]

I did a comparison between two paper settings for the same Ilford Galerie Semi Gloss paper.

Premium Glossy vs Premium Semigloss


The semigloss settings seems to introduce less ink bleeding.

[raydee]

The printer driver implies a different total ink limit depending on the chosen paper type.

[mac_paolo]

The printer driver implies a different total ink limit depending on the chosen paper type.

Exactly. To tell the truth I'm not willing to experiment every single paper configuration.
I'm testing glossy and semigloss alone. This time both with 100% density and 95% density.

[mac_paolo]

Paper Configuration setting

Printed three variations of the same real-world high-frequency image and scanned the same portion at 2400 dpi.
Between premium glossy and premium semigloss paper settings is very hard to tell any difference. A draw.
Between any of the two and the same, but with wide platen and bidirectional print, I definitely see difference in detail.

Ink density

Printed three variations: 0% (standard), -5% and -10%.
-5% is a tiny bit better than 0% and there's not visible reduction in gamut.
-10% has almost the same detail as -5% but the saturation loss is visible.

[AFairley]

It seems to me that your main problem is incompatibity betweent the paper and the Epson inks. I would say your starting point should be to test on an Epson paper, since you know you won't have ink/paper compatibility problems and that the default driver settings for that paper will be correct.  If I recall correctly, I have some Ilford paper that just will not print on the Epson 3800, it just seems to be incompatible, amost like it repels the ink.  I can't remember what kind it is off the top of my head, but I think it's one of the Galleries. 

Paper Configuration setting

Printed three variations of the same real-world high-frequency image and scanned the same portion at 2400 dpi.
Between premium glossy and premium semigloss paper settings is very hard to tell any difference. A draw.
Between any of the two and the same, but with wide platen and bidirectional print, I definitely see difference in detail.

Ink density

Printed three variations: 0% (standard), -5% and -10%.
-5% is a tiny bit better than 0% and there's not visible reduction in gamut.
-10% has almost the same detail as -5% but the saturation loss is visible.

[mac_paolo]

It seems to me that your main problem is incompatibity betweent the paper and the Epson inks. I would say your starting point should be to test on an Epson paper, since you know you won't have ink/paper compatibility problems and that the default driver settings for that paper will be correct.  If I recall correctly, I have some Ilford paper that just will not print on the Epson 3800, it just seems to be incompatible, amost like it repels the ink.  I can't remember what kind it is off the top of my head, but I think it's one of the Galleries. 

I respect you're opinion but I totally disagree. Ilford provides ICC profile for my printer for each and every paper they sell.
No 3rd party paper manufacturer will ever release a paper which is not compatible with Epson/Canon printers.
Beside that I'm still struggling to understand where could any incompatibility arise beside using photo black ink an a matte paper or a matte black ink on a glossy paper.