Pages: [1] 2   Go Down

Author Topic: Different results from Custom Media settings on P9570 PhotoShop, Lightroom etc  (Read 2661 times)

Mick Sang

  • Full Member
  • ***
  • Offline Offline
  • Posts: 170

My colleagues and I have been testing our new 9570 with particular focus on custom media settings as created by two different methods for this printer. One can be entered via the control panel on the printer and the other is set up via a software utility called Epson Media Installer (EMI). We print from MAC with Mojave (10.14.6). The printer firmware is up to date as is the driver.

We printed tests to evaluate whether or not the 2 different media settings had an effect on printed results and if so, to quantify it. Prints were made from Photoshop, LightRoom , ACPU and Print Tool. The results from ACPU and Print Tool were as one would expect, whereas those from Photoshop and Lightroom were definitely not. The average Delta-E of the latter tests was between 2.93 and 6.93.

Both custom media settings were made to accommodate the media Canson Aquarelle. The same reference media, Velvet Fine Art was used for both settings. One custom media setting was made using the control panel method (P.56 in the guide). It was named Aquarelle. The other media setting was made using Epson Media Installer (EMI) (P.57). It was named Aquarelle-2. The EMI method requires an ICC profile to be linked to the custom media setting. We applied our custom profile for Canson Aquarelle. The actual tests were printed on MOAB La Salle Matte in order to save cost and gather the required colorimetric data.

The test chart consisted of 2,033 colour patches on one sheet.  The charts were printed  from PhotoShop, Lightroom, ACPU and Print Tool. For one set of prints, no profile was linked; for another set our custom profile for La Salle Matte was applied; for the last set, our profile for Aquarelle was applied. The use of the two profiles was intended to determine whether or not the profile applied in the printing application would over-ride the profile linked in the media setting made via EMI and to see how the EMI media setting would deal with a duplicate profile coming from that application. The prints were read by an xRite i1Isis-2 and the data was compared and evaluated in Patch Tool. 

A Few Results:
The comparison of 4 prints from ACPU for which neither profile was embedded compared as one might expect or would hope from prints from an identical file without an embedded profile. Prints using each media set were the same within Avg. dE .25 which shows a normal variance between 2 prints from identical files without embedded profiles.

Four sets of comparisons of 4 prints from Print Tool where each had one of either the Aquarelle or the La Salle Matte profile embedded and each was printed using one of the 2 media settings, had an Avg dE of 3.19 to 3.21. So, this clearly and consistently reflected the difference between those 2 profiles and nothing more.

Then, 4 prints were made from Print Tool with each media setting and each of the 2 profiles embedded (either La Salle Matte or Aquarelle profile). Each pair of prints with the same profile but with one of the two media settings was compared. The Avg dE was .24 and .23. So again the result was as it should be from Print Tool.

However, when we got to PhotoShop differences became stark and the reasons are unclear. For instance, we sent 2 prints from PS with our custom Aquarelle profile linked to each and with one of the two 2 media sets selected for each print. The avg dE was 7.07.

In disbelief, we ran this test again – prints and readings. The same result was found. Instead, they should have been virtually the same as the results from Print Tool (Avg dE: .23).

A Comparison of 2 prints from PhotoShop with both media settings and with La Salle Matte profile linked to each print has an avg. dE 5.63. Prints from LR agree.  As a reminder the same test from Print Tool gave the result Avg dE: .24.

So far, we have a total of 42 tests and results range from dE 2.93 to 7.49. At no time did we see an Avg dE below 1.0, as we had from Print Tool, for any test from Photoshop or LightRoom. More testing is required with the latter. But, for now, all we have seen is that something is playing a role here in PS and LR which is not the case with Print Tool.

Visually, prints using the EMI generated media setting look closer to the screen display than those printed with the one made via the Control Panel. But, they should match. From Print Tool they do but not from PS or LR.

Early on, we put this to Epson and were essentially advised that this is Adobe’s problem. Of course, we’re not done with Epson. But, in the meanwhile,  If anyone has thoughts on what might be causing this, I’d appreciate your feedback.

Thanks in advance,

Mick
« Last Edit: May 12, 2020, 11:14:53 am by Mick Sang »
Logged

Rand47

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1882

Quote
A comparison of 2 prints one of which was sent from LightRoom and the other from Photoshop both of which were printed using the Aquarelle-2 media set (EMI) and both had our Aquarelle profile linked resulted in Avg dE: 17.99. Again, this test pair was run again in disbelief. The results were confirmed Avg dE: 17.97.

Mick,

Thanks!  As you know, I’m following this in the EWF group.  What absolutely boggles my mind in the tests is the above.  How could two “identically handled” prints vary so wildly between PS and LR?  It would seem impossible.  I print primarily out of LR but am going to test your schema on my SC P7570.  I don’t have measuring equipment for prints themselves, but with a dE of 17+ a test print of Bill Atkinson’s printer evaluation file should show differences that hit me like a tone of bricks.  I’ll be curious to see which is a closer screen to print match.

I do know that an update of EMI is soon to be released, according to my contact at Wide Format Advanced Tech Support.  But I don’t think they’ve contemplated this issue.  And, it would not seem as though EMI “alone” can account for the apparent disconnect between LR and PS print results - all other factors being equal.

The lesson for me, thus far (and I thank you for your testing!) is that I’ll create custom media via EMI to be registered to the printer and driver for the time being, since you apparently had a better screen to print match in that mode.

I can say, for those reading/following/wondering, that using Epson papers and OEM ICC profiles has produced outstanding results for me on the 7570 so far.  The OEM ICC profiles for Legacy Baryta, Legacy Platine, and good old Enhanced Matte seem excellent visually when printing standard test files that I’m accustomed to evaluating over the years in my GTI light booth.  So this isn’t a matter primarily of whether the printer is capable of outstanding, class leading results, IMO - but more an issue of what the new “black box” of the Epson Media Installer’s interrelationship is between the printer’s internals, the driver, and now - between those and applications’ print modules such as PS and LR in an “application based color management” work flow - when creating media types for 3rd party papers. 

Rand
Logged
Rand Scott Adams

Mick Sang

  • Full Member
  • ***
  • Offline Offline
  • Posts: 170

Quote
So this isn’t a matter primarily of whether the printer is capable of outstanding, class leading results, IMO - but more an issue of what the new “black box” of the Epson Media Installer’s interrelationship is between the printer’s internals, the driver, and now - between those and applications’ print modules such as PS and LR in an “application based color management” work flow - when creating media types for 3rd party papers.

Rand:

No argument from me. We are very happy with this printer. So far, it has been our best experience with Epson printers of which we currently have several and have had many others over the years. Prior to purchasing this 9570, we looked closely at the Canon Pro6000. Colour was not up to snuff. So we waited and are very glad we did. Shortly after, the 9570 and 7570 became available. Many of its features have been long awaited and it is superior to any other of our Epson printers.

Nevertheless, the issue which we have encounter here is concerning and needs to be vetted. If something we are doing or have done is responsible for the differences we have found and I hope this is the case, then we will gain a better understanding from the process. One way or the other, it is severe enough that it can not be simply ignored.

Mick
Logged

JRSmit

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 922
    • Jan R. Smit Fine Art Printing Specialist

Mick, if I understand it correctly, the print from PS with an EMI custom media is quite different from the print with an Printer custom media?

I use W10 as compute platform, and can look into the data (files and SQLite dbs) , technically the EMI media has ICC profiles attached, 1 RGB and up to 3 CMYK . I tested what would happen if I replace the profiles with totally different profiles, for RGB a B-W profile. The print was no different. Then I changed the ink density in the EMX specification file(XML) from 100 to 50 and that made a clearly visible difference. So somehow it uses the EMI settings.

Note that I now have removed EMI from my system, and did a factory reset on my SCP9500, because in W10 EMI did not work on multiple computers. You guys are using Mac platform, and from the ewp forum posts I understand that the driver shows a different behaviour compared to W10.
This was now the thirty time I removed EMI, and now I at least have a consistent ”behaviour” of the printer, using printer made custom media.
The “behaviour” of the printer is nevertheless still not even close to that of an SCP9000.

Logged
Fine art photography: janrsmit.com
Fine Art Printing Specialist: www.fineartprintingspecialist.nl


Jan R. Smit

Mick Sang

  • Full Member
  • ***
  • Offline Offline
  • Posts: 170

These posts are made in LuLa and EWF Group with the intention of informing owners of the Epson 7570 or 9570 of the issue which we have discovered and to ask for assistance from anyone who may know why the effects we have seen are happening.

Yesterday, our testing as described in my first post was completely redone in order to verify their accuracy and to correct anything that may have been done incorrectly. In our previous tests, the colour image was an untagged target consisting of 2,033 colour patches which we printed from Photoshop, LightRoom, ACPU and Print Tool using two ICC profiles. Thanks to a suggestion from Roy Harrington, for the revised tests, the image was converted to ProPhotoRGB.

Aside from one test, all the results from these revised tests were confirmed to be accurate and the results are very similar to those from the first set of tests. The test result which had shown quite a large difference was one in which 2 prints were made of the patch target – one from PS and the other from LR. Each of those was printed with one of the two custom media settings - EMI version versus the one made via the Control Panel. The actual results of that revised test show that both prints match (dE .20). These tests have shown that both LR and PS function equally as printed with either of the 2 custom media settings.

Otherwise, at the end of the day the results are perplexing. For example: 2 prints both from PS with identical settings except for the Custom Media settings as selected via the printer upon loading the paper i.e. EMI version vs Control Panel made version are very different Avg dE: 5.51. They should be the same.

The same test of 2 prints printed from Print Tool with each of the 2 custom media settings and the same ICC profile etc. were in fact the same: dE.20. So, why Photoshop behaves in the way that it does is beyond us.

There are several other very troubling results which come to light upon the introduction of one different ICC profile as compared against the profile which is linked to or embedded in the EMI created version of the custom media setting. For example, 2 prints are made from PS with the same ICC profile. That profile is different from that which is attached to the EMI Custom Media setting.  Each print is made with one of the 2 custom media settings. They differ by dE: 5.51. The results from these prints should not differ.

Next we have 2 prints of the same target image printed from PS with the same ICC profile. Each includes one of the 2 custom media settings. The profile linked to each print is also the same profile that is combined into the EMI custom media setting. They differ by Avg dE: 6.93. What?
This makes no sense whatsoever. That identical test printed from Print Tool resulted in Avg dE: .23. This does make sense.

So, while we can proceed, as others have, to simply use the EMI created media settings and get very good results, as we have seen from our client's prints to date, this still does nothing to answer these questions: Why do these differences exist with prints from Adobe products? Could our prints from those products be better if those differences did not exist? Why do prints from Print Tool not suffer from the same anomalies? Can the status quo method of media selection for printing as used with older printers be trusted with these new printers? Have we done something incorrectly in our testing? If so, what?

Perhaps the answers are simple, after all of this. Further testing from Print Tool and Photoshop and/or Lightroom is required to try to determine an answer to the second question.

The reason for our testing was not fueled by a desire to find fault with our 9570 or to complain about it if we did. It was simply to explore and compare the two paths by which media settings are created for printing with this printer and perhaps to learn something new via the process. In the past, the majority of print settings, aside from platen, were set in the driver and one can still print using that method. But, now, we are encouraged to use EMI which registers the selected media and respective settings to the printer along with its specific ICC (ICM for PC) profile. So, is using the old method still accurate? In order to gain confidence in the new methodology it needs to be tested. At least, that's how I and my team prefer to work.

If anyone has thoughts on what may be going on with these custom media settings in combination with Photoshop or Lightroom, I’d appreciate your help.

Mick
« Last Edit: May 11, 2020, 07:59:28 pm by Mick Sang »
Logged

kers

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4388
    • Pieter Kers

As the Printtool manual describes: Photoshop, the OS and Epson deal with the colour management of the print; - that could (should) be done with one agent... Printtool par example
Logged
Pieter Kers
www.beeld.nu/la

JRSmit

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 922
    • Jan R. Smit Fine Art Printing Specialist

As the Printtool manual describes: Photoshop, the OS and Epson deal with the colour management of the print; - that could (should) be done with one agent... Printtool par example
Pieter, how does printool then work? Does it bypass the Epson driver?
I have no Mac, so I have no idea.
Logged
Fine art photography: janrsmit.com
Fine Art Printing Specialist: www.fineartprintingspecialist.nl


Jan R. Smit

kers

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4388
    • Pieter Kers

I only quoted some part of the manual of printtool 2.0 and that dates back to 2016 so to be sure you have to contact the maker(s)
https://quadtonerip.com/html/QTRprinttool.html

But one thing is sure: the interaction of all three actors ( OS, Epson driver and printing software) make the outcome and they have to play well together to get consistent results.
Not knowing what is happening under the hood you would want to stay with one pipeline; the one you know works well. Especially if it is your work.
In this case it is a new printer and maybe the epson software is not yet fully stable - perfect.
Once it is you can set up one computer with the OS on it that works and a printing software that works fine and leave it like that. ( than make sure you have a good backup in case of failure)
I would not connect it to the internet to be sure it stays the same.
So making the computer part of the printer. A computer capable of printing is not expensive.
« Last Edit: May 12, 2020, 07:22:23 am by kers »
Logged
Pieter Kers
www.beeld.nu/la

Mick Sang

  • Full Member
  • ***
  • Offline Offline
  • Posts: 170

Quote
But one thing is sure: the interaction of all three actors ( OS, Epson driver and printing software) make the outcome and they have to play well together to get consistent results.
Not knowing what is happening under the hood you would want to stay with one pipeline; the one you know works well.

We have selected one of the pipelines for our work. My concern is that since there is a serious inconsistency between the 2 pipelines this leaves us wondering about the degree to which we can trust either pipeline. The fact is that they should be the same i.e. provide the same results. But, they do not.

Mick
Logged

JRSmit

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 922
    • Jan R. Smit Fine Art Printing Specialist

I have had a stable pipeline for years, using Epson printers, the latest stable ones being the SC-P6000-7000-8000-9000 and 20000. And is still stable as we speak. But the problem is that for the new line of printers the Epson software (firmware, driver, EMI) are far from well- functioning. And this is already the case since it’s introduction on the market 5 months ago.
Also Epson has changed the ink formulation specific for these new printers , which renders these printers incompatible with fine art papers from leading companies. Specifically the glossy papers. On top of that the paper transport, again a crucial element , is faulty.







I only quoted some part of the manual of printtool 2.0 and that dates back to 2016 so to be sure you have to contact the maker(s)
https://quadtonerip.com/html/QTRprinttool.html

But one thing is sure: the interaction of all three actors ( OS, Epson driver and printing software) make the outcome and they have to play well together to get consistent results.
Not knowing what is happening under the hood you would want to stay with one pipeline; the one you know works well. Especially if it is your work.
In this case it is a new printer and maybe the epson software is not yet fully stable - perfect.
Once it is you can set up one computer with the OS on it that works and a printing software that works fine and leave it like that. ( than make sure you have a good backup in case of failure)
I would not connect it to the internet to be sure it stays the same.
So making the computer part of the printer. A computer capable of printing is not expensive.
Logged
Fine art photography: janrsmit.com
Fine Art Printing Specialist: www.fineartprintingspecialist.nl


Jan R. Smit

kers

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4388
    • Pieter Kers

New is not always better. Sad but true. I hope things will clear up on Epsons side.
I read on this forum also the 'new' Z9 of HP needs refinement.
Good software is as crucial as hardware.
13 years ago I jumped on the wagon of the HP Z3100 when it just came out. Now I understand I was lucky that they sold me a refined product.
It is hard to understand Epson selling a new printer to professional users with such faults in the software.
But then under working pressure they even send even airplanes in the sky of which they know it is not ready to fly, so...good its just a printer.
Logged
Pieter Kers
www.beeld.nu/la

Mick Sang

  • Full Member
  • ***
  • Offline Offline
  • Posts: 170

Quote
It is hard to understand Epson selling a new printer to professional users with such faults in the software.

We actually don't know as yet what the cause of this issue is or which company's software or hardware is responsible for it or even if it doesn't come down to some misunderstanding or mistake on our part. All we know at this stage is that both sets of our tests as described have produced the results which we have posted so far. We are working to get to the bottom of it. Roy Harrington is on board and is assisting.

Mick
Logged

JRSmit

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 922
    • Jan R. Smit Fine Art Printing Specialist

Mick, the issues you and Paul are faced with, is but one of the issues with this new line of printers.
As I told Paul I will in my W10 environment replicate the test procedure you follow, and share findings.
I hope it is just a misunderstanding, else it is another serious problem with this new line of printers.
Knock wood etc and above all stay positive 😎😊👍

We actually don't know as yet what the cause of this issue is or which company's software or hardware is responsible for it or even if it doesn't come down to some misunderstanding or mistake on our part. All we know at this stage is that both sets of our tests as described have produced the results which we have posted so far. We are working to get to the bottom of it. Roy Harrington is on board and is assisting.

Mick
Logged
Fine art photography: janrsmit.com
Fine Art Printing Specialist: www.fineartprintingspecialist.nl


Jan R. Smit

Mick Sang

  • Full Member
  • ***
  • Offline Offline
  • Posts: 170

This issue has now been fully tested, checked multiple times and confirmed with the help of Mr. Roy Harrington for which we are very grateful. There are 2 methods by which to create custom media settings on the 9570. Both should print to equal results. But, colour managed prints from PhotoShop or Lightroom with either custom media setting differ. Custom Media settings are made either directly on the control panel (CP) or through the use of an Epson app called Epson Media Installer (EMI). The app communicates with the printer and registers the newly created CM with the printer. For both the CP and EMI method, a reference media is selected, copied and renamed. For the EMI method, an ICC profile must be assigned to the CM. For the CP method, this is not required. Once a Custom Media has been established by either method it is available on the control panel and in the driver in different places and they are identified differently. While prints from the EMI method appear to more closely match the on-screen image, prints from the CP version are darker. Both CM settings should match or at least there certainly is no information anywhere in the manual which states otherwise. Hence, if you have a new Epson SC P7570 or 9570 beware of this. Custom media should be created via the Epson Media Installer - not the control panel.

We spoke with Epson regarding this and were directed to take it up with Adobe.

Mick
Logged

Rand47

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1882

This issue has now been fully tested, checked multiple times and confirmed with the help of Mr. Roy Harrington for which we are very grateful. There are 2 methods by which to create custom media settings on the 9570. Both should print to equal results. But, colour managed prints from PhotoShop or Lightroom with either custom media setting differ. Custom Media settings are made either directly on the control panel (CP) or through the use of an Epson app called Epson Media Installer (EMI). The app communicates with the printer and registers the newly created CM with the printer. For both the CP and EMI method, a reference media is selected, copied and renamed. For the EMI method, an ICC profile must be assigned to the CM. For the CP method, this is not required. Once a Custom Media has been established by either method it is available on the control panel and in the driver in different places and they are identified differently. While prints from the EMI method appear to more closely match the on-screen image, prints from the CP version are darker. Both CM settings should match or at least there certainly is no information anywhere in the manual which states otherwise. Hence, if you have a new Epson SC P7570 or 9570 beware of this. Custom media should be created via the Epson Media Installer - not the control panel.

We spoke with Epson regarding this and were directed to take it up with Adobe.

Mick

Mick,

As one who owns a 7570 and who has been following along here, and in the EWF group, I’d like to thank all involved for the serious work involved in chasing this issue to sound final conclusions. 

This new generation printer is amazingly good, but is not without its “mysteries,” many of which Epson was not even aware.  At this point, at least, we do have a “best practices” in terms of making and registering custom media in the printer.

Since Epson has blithely told you to take it up with Adobe, I’m wondering if a test of the two CM’s printing out of Epson’s own “Print Layout” software would produce the same differences in output?  If so, another call to Epson might generate an interesting conversation.  When I was testing “other problems” - I found the output of “problems” was identical in PS, LR, and Epson’s Print Layout!

Thanks!
Rand
Logged
Rand Scott Adams

Mick Sang

  • Full Member
  • ***
  • Offline Offline
  • Posts: 170

Quote
At this point, at least, we do have a “best practices” in terms of making and registering custom media in the printer.
I’m wondering if a test of the two CM’s printing out of Epson’s own “Print Layout” software would produce the same differences in output? 

Epson also suggested we test with their print Layout software. Unfortunately, it does not support the Use Printer Settings (UPS) driver option. The rep we spoke with was unaware of that. The tests would not therefore be equivalent. However, we did print from Print Tool and Image Nest which did not show the same results as were seen from prints from either Adobe app. Prints from those apps were unaffected by the two CM settings.

It is not clear that the issue is a "fault" with the printer or with the two Adobe apps. What is clear is that there is a difference between prints made with the two CM settings from Photoshop or Lightroom where none should exist. In a production environment, it may be tempting to set up a CM on the control panel in the interest of expediency or other reason. EMI is rather slow. A warning should therefore be present to indicate the fact that prints using CM settings made this way may not provide the desired results depending upon the software from which the prints are made. Furthermore, we can't help but wonder if other glitches may be present with the EMI method of CM setting due to the discrepancy between the two methods.

Mick
Logged

webline

  • Newbie
  • *
  • Offline Offline
  • Posts: 3

I can not talk about the P9750 - but I do have the same problem with my new SC-P700.
The prints came out perfect when I use the Epson print layout app on my Mac. If I use LR or PS they are to dark, show a too much saturation and less contrast - not much, but noticeable. Maybe both new Epson printer share some code for the new software driver?!

Detlef
Logged

Mick Sang

  • Full Member
  • ***
  • Offline Offline
  • Posts: 170

Quote
The prints came out perfect when I use the Epson print layout app on my Mac. If I use LR or PS they are to dark, show a too much saturation and less contrast - not much, but noticeable. Maybe both new Epson printer share some code for the new software driver?!

That might be the same issue or it could easily be due to many other things. Our issue becomes apparent when prints from PS or LR are made with a Custom Media setting. That CM setting can be made either on the printer's control panel or by using an App called Epson Media Installer. I'm not familiar with the P700. So, are your prints made with a custom media setting? If so, was that CM setting made on the control panel of the printer or through the use of Epson Media Installer (EMI)? When you print, do you select a custom media upon loading the paper? And then in the driver do you Use Printer Settings (UPS)? Those are the operations which bring the issue that we have seen to the fore.

Mick
Logged

webline

  • Newbie
  • *
  • Offline Offline
  • Posts: 3

Hi Mick,

I didn't use the Epson Media Installer but I did use a custom ICC profile - I would assume that the ICC profile is the important part of a CM setting too?!

Best,
Detlef
Logged

Mick Sang

  • Full Member
  • ***
  • Offline Offline
  • Posts: 170

Quote
I didn't use the Epson Media Installer but I did use a custom ICC profile - I would assume that the ICC profile is the important part of a CM setting too?!

Yes it is Detlef and with each of our tests the same profile was used. Our situation is particular to the 7570/9570. The primary difference in the printed results occurred with the selection of a custom media setting for the 9570 - not the profile. That setting was available to us in two ways.  Each was created using a different method as per the Epson manual. One was made directly on the control panel and the other through EMI. Otherwise, they are identical in terms of general settings. But, the EMI method requires that a custom profile be incorporated with the settings. This can be overridden if you wish. But, we used the same profile with both settings. The prints should have been identical. They were not.

So, in terms of the issue that you are facing, I can't agree that it is the same issue because the same process may not have been used. However, for your prints did you use the same profile, media setting, resolution and select Application Manages Colour for all 3 apps?

Mick
Logged
Pages: [1] 2   Go Up