SureColor P9570

[Jim Kasson]

Here are the links to the updated/tweaked TIFF versions, I'll remove the PNGs to avoid confusion:
Test target for 600 PPI printers (e.g. Canon / HP)
Test target for 720 PPI printers (e.g. Epson)

Sorry, but I don't see the links.

[Jim Kasson]

Interesting point you mentioned about aliasing, sharpening.

Well i use the images Bart van der Wolf published on LuLa forum in 2015:

Here are the links to the updated/tweaked TIFF versions, I'll remove the PNGs to avoid confusion:
Test target for 600 PPI printers (e.g. Canon / HP)
Test target for 720 PPI printers (e.g. Epson)

And those images do show aliasing when sharpened in Lightroom of Photoshop.

I found the targets here:

https://forum.luminous-landscape.com/index.php?topic=99905.0

And you are right; there is aliasing in both targets, so we can't use the presence or absence of aliasing as a test for whether the driver is resampling or not with those targets.

Jim

[JRSmit]

In previous post I mentioned it looks like some sharpening is applied somewhere in the Epson pipeline.
Attached are 2 photos of the print of a res chart printed at 600PPI
First one shows the chart completely (13 by13cm)
Second one a close up of the center of the Siemens star of sinus waves (not just black lines).
In the first photo you se. in the part marked with “4” a small variation in what should be equally spaced lines.
In the close up you see aliasing artefacts that are really there. You also see the “lines” not really reaching the 600PPI mark. Partly because of the limited resolution of the iPad camera, partly because it is true.

To simulate those artefacts I played with sharpening in Lightroom and at setting between 11 and 16 I get a very similar result.

Also : Note that the center shows a sort of oval variation in blackness, this is a sign of a not perfect head alignment. I have not been able to get it completely round yet.

What I will do today is print the res chart with different PPI settings in Lightroom (I use W10 as platform) by scaling the size differently, and copy the resulting PPI in the print setting. But the Epson setting will be quality level 5 and fine details set to ON (600PPI/2400DPI) This to see how the Epson pipeline will deal with this.

In photoshop if you rescale this res chart  with NN it will show strong vertical and horizontal bars.
Therefore if the pipeline rescaled with NN it is logic that these bars will appear.

First a coffee ;-)

[Jim Kasson]

What I will do today is print the res chart with different PPI settings in Lightroom (I use W10 as platform) by scaling the size differently, and copy the resulting PPI in the print setting. But the Epson setting will be quality level 5 and fine details set to ON (600PPI/2400DPI) This to see how the Epson pipeline will deal with this.

In photoshop if you rescale this res chart  with NN it will show strong vertical and horizontal bars.
Therefore if the pipeline rescaled with NN it is logic that these bars will appear.

You might want to try a simpler target:

https://blog.kasson.com/technical/injet-printing-on-epson-part-3/

Jim

[JRSmit]

You might want to try a simpler target:

https://blog.kasson.com/technical/injet-printing-on-epson-part-3/

Jim

Hi Jim, those sets of lines are included in the print res testchart I use.
I also use a checkerboard like image, but with varying sizes black blocks, smallest being 1x2 pixels. This is my quick acid test. The print res chart i use for more in depth analysis.

Being trying to get consistent output today, no such luck. The driver version 6.72 is not sticky, so I printed 3 times , made sure all the settings were set, yet 3 different outputs. One is dramatically different. The outlier I cannot reproduce. It is a weird printer.
I found a way to save the settings, and this looks like I get consistent prints now. But also dealing with customers, so in between I am doing testprints.

The rescaling appears not to be NN but more like the preserve details 2.0  in Photoshop.

[Jim Kasson]

The rescaling appears not to be NN but more like the preserve details 2.0  in Photoshop.

That would be an improvement.

[JRSmit]

That would be an improvement.

Will get back on this, i did test, and the outcome is interesting.

[JRSmit]

Getting very weird now. I cannot get a proper head-alinnment with Velvet Fine Art Print as selected media type. Tried 3 times, no such luck. No change whatsoever in the head-alignment.
Then i selected Fine Art Matte as media type and lo-and-behold head-alignment is way better, still not perfect, but close enough.

The cause, i do not know for sure, but i do know that before i deleted EMI from my computer and did a Factory Reset on my printer, i did "register" via EMI Velvet Fine Art Print. Could it be that the Factory Reset did not clear everything?

Whatever the cause, it is the preferred media type for Hahnemuhle Fine Art Matte papers, which i use a lot, so i am blocked.

I just by telephone reported this to the head of the Dutch Epson office, and made clear that i will make the formal notice active again. I am totally fed-up with this printer, and the way Epson Netherlands is communicating.

[drmrking]

Was this an auto head alignment? Did you try manual?
I had that problem with Canson Platine yesterday, failed an auto-alignment.
Did a manual one instead (tedious), but because 99% of the cases is was spot on, was pretty quick to check and enter data.
Afterwards ran the resolution chart to double check alignment and looked good.

Mike

[JRSmit]

SureColor P95xx/75xx, indeed 300/600 PPI device, but Head-Alignment not perfect!

Triggered by Jim Kasson's remark in earlier post, i did the following test to determine the PPI resolution of the SCP95xx printer.
I use W10 and Lightroom Classic, and the 600PPI res-test file of Baert van der Wolf.

I found interesting aliasing artefacts, basically proofing the printer is a 300/600PPI device (translated to 1440/2880 DPI on the print).
But also that the Head-alingment needs some improvement.


In Lightroom print module i placed on an A4 the res-test, with image dimensions adjusted to get different PPI:
567PPI
598PPI
600PPI
602PPI
615PPI
The PPI setting in the print job is set accordingly, and no output sharpening applied. In this way Lightroom creates a print file wih that PPI resolution. To double-check i also did print to file, with the same PPI settings, to be sure that the print output file does not introduce aliasing artefacts.

In the SCP9500 printer driver (v6.72) the print quality setting is set to level=5, Fine Details=ON, the paper edge quality set to Standard, the res-test image centered on the page via the driver, to stay away fromm any paper edge print quality issues, typical for the double-head concept.
As media type i used Epson Premium Glossy 250gsm. In the media settings is the button "custom Settings" that allows you to save your media settings. As i am not using a custom media, this functions ok.
For each PPI value i saved the whole setting in a printjob in Lightroom, and double checked that the settings in the driver stayed as set.
Note that the SCP95/75 driver up to 6.72 is notoriously poor in that respect.

I did a feed-alignment and then a head-alignment, then I issued the print jobs one by one. Prints are on a glossy PE-paper.
Then i found a way to take a picture of the print without introducing aliasing artefacts, by using my 2019 iPAD and with a distance of 7cm it still focused well, and did not introduce visible artefacts. So the artefacts you see in the images are as they are in the print.

The pictures show the center of the "siemens" star, including the 300PPI and 600PPI rings. One picture shows the bottom-left (linksonder) part of the res-test image, there test "4" shows black lines in 4 different directions, and if perfect each quadrant should give an equal greyish impression when viewed from a distance.


Conclusion:
Send files with 300PPI or 600PPI to the printer, else you are in for some negative surprises.
Head-alignment is techniclly speaking close to perfect, BUT, NOT SO FOR THE BLACK INK. Vertical lines get always printed a that fatter than horizontal lines, also some false colors are visible.


The number of attachment is limited to 4, so:

SCP9500-GlossyPaper-598PPI-600PPI-1.jpg :
some horizontal and vertical "vague banding" across the 300PPI ring as well as stronger aliasing artefacts.

SCP9500-GlossyPaper-600PPI-600PPI-1.jpg :
very modest aliasing artefacts. When shown on my EIZO CG248-4K zoomed in 100% is see the same artefacts.
You also see, on the real print a bit better, that the printer resolution , its modulation transfer, sort of is 0 at the 600 PPI ring.

SCP9500-GlossyPaper-615PPI-600PPI-1.jpg :
again several horizontal and vertical "vague banding" across the 300PPI ring as well as stronger aliasing artefacts.

SCP9500-GlossyPaper-600PPI-600PPI-Linksonder-1.jpg:
diagonal lines are almost equal, the vertical lines are fatter than the horizontal lines, and some false color is visible in the horizontal lines.
Note that the aliasing artefacts are quite symmetrical in the "Siemens" star (is composed out of a sinus variation and not just black lines),
but in the center rings one gets a sort of oval impression (more black ink), this is the effect of the not perfect alignment of the Black Ink.

Personally i use primarely Lightroom Classic for printing, and allways have the print output PPI setting set to the printer resolution, 300/600 or 360/720 depening on the printer used. I also sometimes use other pinting applications, in these too i set the output PPI accordingly.
So for me  i accept the fact that the SCP95/75 is a 300/600PPI printer.
The only issue left is the Head Alignment, which is not good for the BLACK INK.

[Jim Kasson]

Conclusion:
Send files with 300PPI or 600PPI to the printer, else you are in for some negative surprises.
<snip>
So for me  i accept the fact that the SCP95/75 is a 300/600PPI printer.

Thank you for doing this. I think what you have demonstrated is that the Epson driver resamples to 300/600 ppi, not anything about the printer itself.

Jim

[Jim Kasson]

I found interesting aliasing artefacts, basically proofing the printer is a 300/600PPI device (translated to 1440/2880 DPI on the print).

Are you sure about the marking engine resolution? That would be a departure from the way the other Epson drivers work. I would expect the printer to be a 1200/2400 dpi printer.

Jim

[JRSmit]

Hi Jim,

it is 1200/2400DPI, i attached a photo (iPad) of a print where FineDetails are OFF. Yet there are some aliasing artefacts visible in te range from 300PPI/1200DPI to 600PPI/2400DPI. So the printer has resolution in that range, although there is no "siemens" star data because of the downrez to 300PPI of the image prior being converted to dpi. So by mine understanding the the marking engine is 1200/2400 DPI, as it is in essence quad-tone, the PPI resolution is 1/4 so 300/600 PPI.

Or are we talking about different things?

Are you sure about the marking engine resolution? That would be a departure from the way the other Epson drivers work. I would expect the printer to be a 1200/2400 dpi printer.

Jim

[Jim Kasson]

Hi Jim,

it is 1200/2400DPI, i attached a photo (iPad) of a print where FineDetails are OFF. Yet there are some aliasing artefacts visible in te range from 300PPI/1200DPI to 600PPI/2400DPI. So the printer has resolution in that range, although there is no "siemens" star data because of the downrez to 300PPI of the image prior being converted to dpi. So by mine understanding the the marking engine is 1200/2400 DPI, as it is in essence quad-tone, the PPI resolution is 1/4 so 300/600 PPI.

Or are we talking about different things?

We are now on the same page with the marking engine resolution. However, the halftoning algorithms used by Epson printers is not anything that could be called "quad-tone", but is error diffusion with blue noise dither. It is an error to discuss the resolution of the marking engine in ppi. The marking engine has no knowledge of the contone resolution of the image, and therefore its resolution should be expressed in dpi.

[JRSmit]

We are now on the same page with the marking engine resolution. However, the halftoning algorithms used by Epson printers is not anything that could be called "quad-tone", but is error diffusion with blue noise dither. It is an error to discuss the resolution of the marking engine in ppi. The marking engine has no knowledge of the contone resolution of the image, and therefore its resolution should be expressed in dpi.

You are right, wrt to the dpi. W.r.t. to the halftoning algoritms , that is beyond my understanding, but in essence it is a cmyk print system, that is what i ment with quad-tone.

What i am referring to with PPI is the PPI of the image/photo that is put into the pipe-line. This is what in Lightroom print module the output PPI setting is used for. And what is apparently expected by the printer driver at the input side. The test i did show that if you do not provide the 300PPI or 600PPI in the image/photo to be printed, in the pipe-line somewhere funny things happen.
The dpi setting is done in the driver with the Print Quality Level setting, and there it "talks" about dpi.

[Jim Kasson]

You are right, wrt to the dpi. W.r.t. to the halftoning algoritms , that is beyond my understanding, but in essence it is a cmyk print system, that is what i ment with quad-tone.

What i am referring to with PPI is the PPI of the image/photo that is put into the pipe-line. This is what in Lightroom print module the output PPI setting is used for. And what is apparently expected by the printer driver at the input side. The test i did show that if you do not provide the 300PPI or 600PPI in the image/photo to be printed, in the pipe-line somewhere funny things happen.
The dpi setting is done in the driver with the Print Quality Level setting, and there it "talks" about dpi.

The 7570 is not a CMYK printer. The inkset is more complex than that. I agree about not letting the driver do the resampling, but doing it yourself or letting Lr do it. I wish the Epson driver didn't resample, but it does.

Jim

[dchew]

Guys,
Thank you very much for this discussion. It has significantly helped my understanding of how the printing engine works.

Dave

[Rand47]

Guys,
Thank you very much for this discussion. It has significantly helped my understanding of how the printing engine works.

Dave

+1

Rand

[unesco]

Hi all,

Any new good news about printing on non-Epson glossy baryta media e.g. HM or Ilford? I have quite a big stock of Harman Gloss Baryta (incl. warmtone version) which are my best papers for B&W and would like to continue using them.

There must be something in this since I have checked Epson Digigraphie web page and these new printers has been added to the list of certified printers for Digigraphie standard but with annotation: "SureColor SC-P7500 (currently only EPSON certified media)"

I just wonder about going to higher formats from my current 17 inch and now I am stuck between P7000 and P7500 but not want to be a beta tester (also for Z9).

Any thoughts will be very appreciated. The new Sure Color series looks as a dream machine on paper... if it works properly...

[JRSmit]

No news.