Ad
Ad
Ad
Pages: « 1 [2] 3 »   Bottom of Page
Print
Author Topic: Epson x800 print softness  (Read 10269 times)
Brian Gilkes
Sr. Member
****
Offline Offline

Posts: 431


WWW
« Reply #20 on: December 04, 2007, 05:06:54 PM »
ReplyReply

Quote
If I leave the values for margins alone, I end up with a soft print. Instead, I've discovered today that if I set all four margins to 0, the resulting sharpness is good. But this should not be the case, because margins should not have anything to do with printing sharpness (unless, maybe, you use auto expand or something like this and the driver resizes your picture). Besides, using the Windows driver I can set whatever margins I like and the output is always as expected.

I never would have thought of that one.
To solve a problem I had with my Epson 4000, I was advised by the Epson Techo to set margins at zero. That has been my pracice since out of habit and stll using the 4000 for some jobs.
That could explain why no one has noticed a sharpness problem. I'll test that too asap.
Nothing further to add at the moment.
Brian
Pharos Editions
Logged
ares
Newbie
*
Offline Offline

Posts: 23


« Reply #21 on: December 04, 2007, 05:25:42 PM »
ReplyReply

Quote
That could explain why no one has noticed a sharpness problem. I'll test that too asap.

If you have always printed with zero margins then you should be ok. Lucky you. Anyway, I've made other prints, and while they seem identical to the naked eye, examining them under a 32x microscope I think I'm able to spot a very slight difference in sharpness, always in favor of prints made on a default paper size. But I'm not sure, the difference is so small that I probably would need a proper double blind test to draw a definitive conclusion.
Logged
Wayne Fox
Sr. Member
****
Offline Offline

Posts: 2846



WWW
« Reply #22 on: December 04, 2007, 06:27:03 PM »
ReplyReply

Quote
At this point, I obtain what is shown in picture 2.

[a href=\"index.php?act=findpost&pid=158208\"][{POST_SNAPBACK}][/a]

The pop-up menu in that picture allows you to select a printer, and it will set the default margins to the maximum the printer can handle for that size of paper, based on sheet feed.  Helpful especially on the bottom where some printers require some area for gripping the paper.

Since all of my custom setups are for roll paper, I usually leave them at 0.
Logged

Wayne Fox
Sr. Member
****
Offline Offline

Posts: 2846



WWW
« Reply #23 on: December 05, 2007, 01:57:07 AM »
ReplyReply

For what it's worth, tonight I resized to 360 ppi without interpolation the test page that has been used, and printed 3 versions on my 11880 printer on 24" roll kodak Glossy paper at 2880 dpi.

Version one was using the standard 24x30 roll option.
Version two was with a custom paper size with all margins set to 0 (which is also the default for the 11880 printer)
Version 3 was a custom paper size with manual margins of .5 inches on sides and 1 inch top/bottom.

I marked the back, and cut the prints to the same size, cutting off the identifying text from the front.  Mixed them up, and spent about 10 minutes with a 10x loupe - not knowing which one I was looking at ... just trying to see if one was slightly better or worse than the others.

While there are some subtle differences in dot patterns, none equate to a difference in sharpness.
« Last Edit: December 05, 2007, 01:57:46 AM by Wayne Fox » Logged

Sven W
Sr. Member
****
Offline Offline

Posts: 514


« Reply #24 on: December 05, 2007, 03:59:27 AM »
ReplyReply

Quote
If you have always printed with zero margins then you should be ok. Lucky you. Anyway, I've made other prints, and while they seem identical to the naked eye, examining them under a 32x microscope I think I'm able to spot a very slight difference in sharpness, always in favor of prints made on a default paper size. But I'm not sure, the difference is so small that I probably would need a proper double blind test to draw a definitive conclusion.
[a href=\"index.php?act=findpost&pid=158254\"][{POST_SNAPBACK}][/a]

But who is looking at a print under a microscope  
/Sven
Logged

Stockholm, Sweden
madmanchan
Sr. Member
****
Offline Offline

Posts: 2108


« Reply #25 on: December 06, 2007, 11:32:51 AM »
ReplyReply

Quote
But who is looking at a print under a microscope  
/Sven
[a href=\"index.php?act=findpost&pid=158354\"][{POST_SNAPBACK}][/a]

The photographer, of course!  
Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6938


WWW
« Reply #26 on: December 06, 2007, 11:42:14 AM »
ReplyReply

If truly microscopic differences are all this thread is about, one wonders what's the fuss, but I thought the OP was about differences of sharpness you see with the naked without having to look too hard.

Wayne - really great you are taking the trouble to do those tests and with such a state of the art printer. The fact that you can't replicate the problem would seem to suggest that the issue is not generic but case-specific, and that would suggest "ares" really needs some high quality concentrated tech support from Epson to sleuth what's going in his set-up and workflow to cause this.

Wayne - also - not clear that the resampling to 360 is done exactly the same way (mathematically) in the Epson driver versus in Photoshop. There's been much inconclusive discussion about this in the past, for earlier versions of Photoshop and Epson drivers. It's an interesting question that perhaps should be revisited - again making a test or two for a SUBSTANTIAL resampling to 360 both ways, and seeing whether there is a visible (naked-eye) difference of image quality (sharpness, tonal gradations) between the two approaches. Time permitting I'm thinking of doing it, but please beat me to it.  
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Wayne Fox
Sr. Member
****
Offline Offline

Posts: 2846



WWW
« Reply #27 on: December 06, 2007, 12:54:22 PM »
ReplyReply

Quote
Wayne - also - not clear that the resampling to 360 is done exactly the same way (mathematically) in the Epson driver versus in Photoshop. There's been much inconclusive discussion about this in the past, for earlier versions of Photoshop and Epson drivers. It's an interesting question that perhaps should be revisited - again making a test or two for a SUBSTANTIAL resampling to 360 both ways, and seeing whether there is a visible (naked-eye) difference of image quality (sharpness, tonal gradations) between the two approaches. Time permitting I'm thinking of doing it, but please beat me to it. 
[a href=\"index.php?act=findpost&pid=158727\"][{POST_SNAPBACK}][/a]

In this case I don't think any resampling took place ... I resized the print without resampling to 360ppi, then printed it at the resulting size.  

I think your talking about when printing to a specific size is it better to resample with photoshop to 360 dpi, or is it better (or as good) to let the driver resample, which is the recommended workflow presented by Jeff Schewe in the LL Camera to Print video.

There is some logic in my mind that the driver may be able to screen and resize the image in one step better than resizing in photoshop and screening in the driver.  But then I'm clueless into the actual math of it all, so my logic may be out in left field

I have recently been using Jeff's approach, sending the master file at native resolution to the printer, and letting the printer driver handle resizing, which outstanding results.  The only exception is if I go below about 220ppi, I then uprez the unsharpened version 200%,  sharpen that, and again send that to the printer and let the driver size to the final print size.

I have not tested to see if there is any perceptible difference in image quality.  The resulting prints are very very good, and it saves so much time in the workflow with great results I haven't really thought about "testing" it.  I suppose upon micro examination there may be a difference, but Jeff is pretty clear on the video that normal viewing distances and the ability of the human eye are an important reason why he recommends this workflow.

Since I'm having a lot of fun playing with my new printer, maybe I'll try some tests.
Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6938


WWW
« Reply #28 on: December 06, 2007, 01:28:45 PM »
ReplyReply

Wayne, good if you could run a test or two. I'm in the thickets of testing my 1DsIII just now and have an issue or two I really need to get comfortable about in a hurry.

I agree - Jeff did/does espouse that workflow. However, I know of at least one other technically expert practitioner (un-named to protect the guilty) who resamples to 360 first - but mainly to get an exact match to PK Sharpener Output Sharpening, rather than a specific concern over resampling differences between Photoshop and the Epson driver. I've tried it both ways on different images at different times and I find both totally acceptable. But I haven't tested it on the same image at the same time recently.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
ares
Newbie
*
Offline Offline

Posts: 23


« Reply #29 on: December 06, 2007, 04:07:51 PM »
ReplyReply

Hello,
just to clarify: when I was talking about the microscope I was referring to the difference between a print made on default paper size and one made on custom paper size with all four margins set to zero. In that case there is no visible difference, but I wanted to be sure that really there is no difference, as it should be.
The difference between a print made on default paper size and one made on custom paper size with margins other than zero is readily apparent to the naked eye.

So, to summarize my findings:

Under Windows:

print on default paper size -> sharp
print on custom paper size with margins set to zero -> sharp
print on custom paper size with margins other than zero -> sharp

Under OSX:

print on default paper size -> sharp
print on custom paper size with margins set to zero -> sharp
print on custom paper size with margins other than zero -> soft (visible to the naked eye)

I've now tested this on two 7800s and one 9800 with consistent results.

Maybe Epson has fixed the issue with the drivers for the new x880 printers. Maybe it has something to do with the new dithering algorithm, I don't know.

Anyway, thank you all for your help, it's been really appreciated.
Logged
madmanchan
Sr. Member
****
Offline Offline

Posts: 2108


« Reply #30 on: December 07, 2007, 08:19:25 AM »
ReplyReply

Glad you found a solution/workaround. That's some extensive testing you did!
Logged

Wayne Fox
Sr. Member
****
Offline Offline

Posts: 2846



WWW
« Reply #31 on: December 07, 2007, 05:55:29 PM »
ReplyReply

Now I am somewhat puzzled.  I still have my 9800, so curiosity got the better of me.

I followed the exact same procedure as I did with my 11880, printing 3 versions.  What puzzles me is the one printed with a custom paper size with 0 as margins is the soft one.

Yes, it's really hard to see with the naked eye, very easy to see with a loupe or scanned image.

I have no idea why the printer driver would be handling the data differently based on the border settings/custom paper settings.
Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6938


WWW
« Reply #32 on: December 07, 2007, 06:00:09 PM »
ReplyReply

Wait a minute - Wayne - you are getting the OPPOSITE problem from Ares on the 9800? He said the soft image is the one WITHOUT the 0 borders, and you're saying the soft one is the one WITH the zero borders? Both of you are using OSX? Are both of you using the same version of the Epson driver? Same photo? Same settings? Something is REAL WIERD!
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Wayne Fox
Sr. Member
****
Offline Offline

Posts: 2846



WWW
« Reply #33 on: December 07, 2007, 09:14:51 PM »
ReplyReply

Quote
Wait a minute - Wayne - you are getting the OPPOSITE problem from Ares on the 9800? He said the soft image is the one WITHOUT the 0 borders, and you're saying the soft one is the one WITH the zero borders? Both of you are using OSX? Are both of you using the same version of the Epson driver? Same photo? Same settings? Something is REAL WIERD!
[a href=\"index.php?act=findpost&pid=159123\"][{POST_SNAPBACK}][/a]

yes ... that's what has me puzzled.  I am using OS X 10.5 and the new leopard beta drivers for all the tests I have printed.

Going to retest tonight just to make sure I didn't mess up the labeling on the 3 prints.

BTW, checked this with a Canon ipf6100, and couldn't see a difference.  So at this point, my 3800, my 11880, and my ipf6100 all print the same regardless of paper size/border size choices, but my 9800 seems to be exhibiting the problem discussed here in the thread.
« Last Edit: December 07, 2007, 09:16:09 PM by Wayne Fox » Logged

Wayne Fox
Sr. Member
****
Offline Offline

Posts: 2846



WWW
« Reply #34 on: December 11, 2007, 12:47:56 PM »
ReplyReply

Just an update to this, I retested this on the 9800.

I was out of 24" paper so I used 16" paper.  The custom paper size with 1" margins required the image to be clipped, so this was a difference between the two tests.

This time there was no visual difference between any of the 3 prints.

I have no conclusions at this point other than there is a problem, it isn't exactly repeatable, and no idea how to insure a custom paper size will deliver a sharp print other than testing each one that is created.

I have a distrust for custom paper sizes at this point however ... I may do some more testing on my 3800 and 11880 just to make sure this problem hasn't migrated to those printers.
Logged

Schewe
Sr. Member
****
Offline Offline

Posts: 5475


WWW
« Reply #35 on: December 11, 2007, 01:01:07 PM »
ReplyReply

Quote
I have no conclusions at this point other than there is a problem, it isn't exactly repeatable, and no idea how to insure a custom paper size will deliver a sharp print other than testing each one that is created.
[a href=\"index.php?act=findpost&pid=159898\"][{POST_SNAPBACK}][/a]


The one thing I have _NOT_ seen mentioned is what starting page size was selected before going into the custom page size...if one started on a standard page size that represented a borderless printing and made a custom page size from that, I could see where the driver would be doing on the fly resize because that's what borderless page sizes do.
Logged
Doombrain
Full Member
***
Offline Offline

Posts: 206


WWW
« Reply #36 on: December 11, 2007, 01:22:28 PM »
ReplyReply

has anyone tried to print some digital technical test prints, like a resolution test? surely the best way to test for this, not scanning in images.

I've got my hands for for a few day now but if someone can quickly summarize the settings i'll try some tests on my Epson's.
« Last Edit: December 11, 2007, 01:23:01 PM by Doombrain » Logged
ares
Newbie
*
Offline Offline

Posts: 23


« Reply #37 on: December 11, 2007, 02:14:43 PM »
ReplyReply

Quote
The one thing I have _NOT_ seen mentioned is what starting page size was selected before going into the custom page size...if one started on a standard page size that

I've tested this too, by changing between various sizes of sheet, roll, and borderless roll.

The result is always the same, at least for me: borders other than 0 -> soft print, with the degree of softness varying depending on margins size.

By the way, now I've tested a 4800 under Tiger (not sure what the driver version is, sorry) and the bug affects it too.
Logged
ares
Newbie
*
Offline Offline

Posts: 23


« Reply #38 on: December 11, 2007, 02:19:00 PM »
ReplyReply

Quote
has anyone tried to print some digital technical test prints, like a resolution test? surely the best way to test for this, not scanning in images.

I've printed some SMPTE test images, just to be sure that the issue isn't image dependent, but I'm not interested in doing measurements, and the softness is easily recognizable on the test image from Bill Atkinson I've always used.
Logged
kovacj
Newbie
*
Offline Offline

Posts: 2


« Reply #39 on: December 11, 2007, 02:52:50 PM »
ReplyReply

Quote
I've attached a JPG version of the original image I'm using for my tests. It's already sharpened (0,5-160). Could you be so kind to try with this one?

[attachment=4123:attachment]
[a href=\"index.php?act=findpost&pid=158106\"][{POST_SNAPBACK}][/a]

It's not clear from your post if you're actually using the jpeg attached to your post for your tests?  Given that the attached jpeg is a roughly 4"x6" print at 360ppi, I would think that the results would be highly dependent on how it was up-sized if making a large print.
Thanks,
-jk
Logged
Pages: « 1 [2] 3 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad