ET-8550 incoming

[glasseye]

After using this printer now for a little more than a week, I don't know if I could go back to a cartridge-based printer. Ink management is pretty much a thing of the past, and the print quality is beyond what I expected.

Image quality aside, the best part about the inks is loading them. I love how it auto-detects the bottles, sucks in the ink and shuts off when the tanks are full, leaving a little bit in the bottle. And the can't-miss bottle keying prevents mis-fuelling, something I did once on my 9800.

My L-850 was very finnicky to load and I always made a mess. Not so with this beauty. I gave my still functional L-850 to a fellow photographer. We're both delighted.

 

[nertog]

I printed a patch sheet with 283 colors with the ET-8550 - ultraglossy quality best and with an L1800 - 6 color dye inks incl. LM and LC - ultraglossy quality high.

Are both patches printed at the printer's highest quality setting? The results seem to match my observations here, minus the obvious banding in the L1800 sample which seems to be random and might be related to an ink flow or piezo actuator issue. I've seen similar banding patterns on old R3000 printheads.

It would be interesting to repeat this for desaturated (light) R, G and B colors.

 

[Ink stained Fingers]

A question was raised recently if there is any visible difference in print 'quality' between the ET-8550 and the XP-15000 - both w/o light links. Since there is no clear definition what 'print quality' is in this context beyond a visible assessment I was trying to look to some color spots - on a target sheet - and measure the chroma and luminance spread of the color spots - this more as a proof of concept evaluation than an actual printer comparison. The ET-8550 offers threee quality ettings - standard - high - best and the L1800 just runs with 'standard' and 'high', the driver does not offer other settings like 'RPM' longer time ago with the R800 and similar printers, so I just looked to the highest settings available to compare with. There could be another limiting element in the test - I scanned with 2400 dpi but I just don't know the real and effective resolution of my age old 3490 photo scanner, I vaguely remember from earlier tests that the scanner does not deliver more than 1500 dpi real resolution measured with a high res scanner target on film with a Siemens star. But anyway - the color/lum spread of the printed colors appear to be as a usable characteristic to compare printouts/printers.
There is some banding in the L1800 printout - I never use the 'high' quality setting on this printer for regular prints , and the effect disappears with the standard quality setting, and it prints faster. I assume it's the ink spread - the driver probably prints with 1.5pl droplets in the high quality mode, and with 3pl or similar in the standard setting - the larger ink drops cause a wider ink spread on the paper and just cover up any banding. I have seen that this banding depends on the type of paper - the Aldi paper shows it but another paper (out of a paper sample box) does not, the coating probably influences the width of the ink spread.
It would be necessary to look to other colors as well to get a more complete view of this approach, it is pretty time consuming and I don't even have a good test flow at this time yet so I may pick up this subject any time later, and with more printers

 

[Ink stained Fingers]

When you google for 'print quality' you get millions of links, which could already be an indication of the complexity, and there is a publication which describes in some detail the difficulties and the complexity to define a measure for 'print quality' - the idea what it encompasses varies widely between users, and this article does not make it simpler and easier.

https://www.researchgate.net/public...y_metrics_for_the_evaluation_of_print_quality

I couldn't stop and had another look to the differences between an ET-8550 and an L1800 - this printer with light inks. I'm using the approach from above - scan 2 light cyan color zones of patch sheets printed with these printers and display the data as a pixel cloud in a color space, I'm using this time the ColorInspector 3D which does not let me overlap several images but can display the data in various different color spaces, MonacoGamutWorks only in the Lab space.

https://home2.htw-berlin.de/~barthel/ImageJ/ColorInspector/help.htm
https://imagej.nih.gov/ij/plugins/color-inspector.html

'To run it, download ColorInspector3D.jar and double click on it. On Windows, Java 5.0 or later must be installed.'

It's an old program but free, a similar display would be possible with the Colorthink software by Chromix - but at a charge.

These are crops of the color spots - this from the L1800



and this from the ET-8550




The noise pixel clouds look like these - in the Lab color space, this for the L1800




and this for the ET-8550



It is directly visible that the ET-8550 pixels have a wider variance in vertical direction - indicating a wider spread
of luminance values , and the eye is more sensitive to these than to chroma variations, this makes such variations more visible to the viewer. The color spots for the display of the stray colors and luminance are 7,5 x 8 mm, the spots shown above are cropped to about 2,5mm, the 2400 dpi at scanning works like a strong magnifier, the images are about 240 pixels wide - 1 tenth of an inch = 2,5mm.