Ad
Ad
Ad
Pages: [1] 2 »   Bottom of Page
Print
Author Topic: Article: Debugging Profiled Inkjet Printing  (Read 5220 times)
Diane
Newbie
*
Offline Offline

Posts: 3


« on: July 19, 2013, 03:57:08 PM »
ReplyReply

I have an Epson R3000 and am printing with Lightroom 4.  I just read the above article on the Lula site and want to know what I am supposed to select in the "Mode:  (J)" drop down list of the printer drive when trying to get through Procedure 1.  Here are my choices:

Epson Standard
Adobe RGB
PhotoEnhance
ICM
OFF (no color adjustment)

Or does it make no difference at all what I use since, in Lightroom, I am selecting under Print Job, Profile:  "Managed by Printer?"  Maybe I missed it, but I don't see where it tells me under Procedure 1 what I am supposed to select on the driver under "Mode: (J)."

I apologize if this has already been asked.  I did a search and did not find anyone else asking my question.
Logged
Diane
Newbie
*
Offline Offline

Posts: 3


« Reply #1 on: July 19, 2013, 04:12:52 PM »
ReplyReply

I have the same question for part 2 of the procedure.
Logged
stefohl
Jr. Member
**
Offline Offline

Posts: 59


WWW
« Reply #2 on: July 20, 2013, 03:30:24 AM »
ReplyReply

Here are my choices:

Epson Standard
Adobe RGB
PhotoEnhance
ICM
OFF (no color adjustment)

Or does it make no difference at all what I use since, in Lightroom, I am selecting under Print Job, Profile:  "Managed by Printer?"  Maybe I missed it, but I don't see where it tells me under Procedure 1 what I am supposed to select on the driver under "Mode: (J)."


Ihaven't seen the LULA tutorial, but the standard method is to choose the correct printer profile for your paper instead of “Managed by printer” and then choose “OFF (no color adjustment) in the printer driver.

/stefan
Logged

Stefan Ohlsson
Projektor
www.projektorutbildning.se
Rhossydd
Sr. Member
****
Offline Offline

Posts: 1889


WWW
« Reply #3 on: July 20, 2013, 04:02:13 AM »
ReplyReply

Hi Diane, welcome to LuLa :-)

I'm afraid the article you're puzzled by isn't the clearest bit of writing here and could do with some clarification. It also assumes that you're using a modern printer supplied with great profiles, but sadly not all printers are, even now. The article also seems to assume you've installed the full set of manufacturer's drivers that includes all their supplied profiles*, not everyone will have done that either.

From what I can understand, the idea behind the article is to go back to basics to fault find printing workflows.
1. The starting point is to use the manufacturer's own paper and profiles and default settings, applying the profile via the printer driver to get a base result.
2. Then make a print with the manufacturer's supplied profile applying it in your own software and leaving the printer driver not applying any profile at all.
3. Finally to build your profile or use some other custom profile, print with it and compare with the previous results.

The problem here is that some printer drivers allow more control than the article expects/mentions, as you've found out. So it's not clear what exact settings to use.

From what I can see the 'debugging process' goes in three stages;
 
In stage 1. You set your software to manage colour by printer and leave colour management to the driver defaults. In your case use Epson standard. (In the specific case of your driver you could also use ICM and select the correct profile* in the appropriate dialogue or you could also use Adobe RGB assuming your image is Adobe RGB).

In stage 2 you'd set the driver to OFF no colour management and apply the manufacturer's profile* in your printing software(Lightroom).

In theory 1 & 2 should give identical results.

In stage 3 you'd use the settings from stage 2, but apply a profile you were having issues with or build a new profile from scratch.
Frankly, anyone thinking of building their own profiles shouldn't need this kind of debugging help, but it may be useful for people having problems using non-standard profiles.

*Manufacturers can use some pretty cryptic naming on their profiles, so you may need to Google to find out which profile to use for your chosen paper.


Hope this helps
Logged
Rhossydd
Sr. Member
****
Offline Offline

Posts: 1889


WWW
« Reply #4 on: July 20, 2013, 04:03:30 AM »
ReplyReply

Ihaven't seen the LULA tutorial
Maybe you should read the article before trying offer help, rather than further confusing the issue.
Logged
Diane
Newbie
*
Offline Offline

Posts: 3


« Reply #5 on: July 20, 2013, 11:09:36 AM »
ReplyReply

Rhossydd, thank you for  responding! 

I just began to print from LR 4, but I am not totally unfamiliar with icc profiles since I used to print from Photoshop using an R1800.  The problem I'm having is that my prints are all coming out greenish, and I am spending a lot of time and money on ink/paper making small test prints before I print big.  Even using an Epson paper and having the printer manage the color, the prints are coming out a little greenish.  So I decided to do a head cleaning test.  After doing a head cleaning test, there were some black voids (but I think it's the matte black cartridge and not photo black which is what I've been using on my prints).  After doing a cleaning, the second head cleaning test was even worse than the first one (even more black voids).  This also used to happen with my R1800.  If I left it alone for a day usually the voids went away after a head cleaning.  I'm hoping it will be the same with the R3000.

I have not touched anything in the "color controls" section of the driver.  Anyhow, I am sure I am doing something wrong, but I cannot figure out what it might be.  This is quite frustrating.
Logged
Rhossydd
Sr. Member
****
Offline Offline

Posts: 1889


WWW
« Reply #6 on: July 20, 2013, 11:44:12 AM »
ReplyReply

First step is to ensure your printer is working correctly.
If the nozzle check isn't correct you really need to get that sorted before proceeding further. There's some reports that running lots of cleaning cycles sequentially can have detrimental effects on some printer heads, so maybe try printing some pages with images that use all the carts, you can get away with making the prints on any old cheap paper whilst just running ink through the heads. Then do another nozzle check to see if everything is working correctly. Leaving the printer overnight after the test prints before doing the nozzle check may also be beneficial.
Unfortunately the R3000 hasn't  a stellar reputation for reliability compared to some other Epsons, so it's possible that it may have developed a fault.

Once you have got it working correctly, then test with an industry standard test image. There are several across the web that are a reliable benchmark to test with.

The routine described in the 'debugging profiled inkjet printing' is probably best to follow.

If you still have problems having got the printer working properly, ask again here. It's important to provide every detail you can to help people understand what's going on and offer relevant help.
OS version
Printer driver version
Software version
Profile
paper
test image used
Then we may be able to help further.
Logged
eronald
Sr. Member
****
Online Online

Posts: 3870



WWW
« Reply #7 on: July 25, 2013, 08:17:24 AM »
ReplyReply

Hello Diane.

For what it's worth I am the author of the cited article.

For Step 1:
- Please select "Epson Standard" in the drop down list.
- For this step 1 which serves to create a reference print, please be certain that you have a genuine paper from Epson available, and that the right paper type is selected in the printer settings.
- You are correct that the application -here Lightroom- should say that the color is "managed by Printer". This is called "Printer Manages Color" or "Vendor Color" in my article.

All the best.

Edmund

I have an Epson R3000 and am printing with Lightroom 4.  I just read the above article on the Lula site and want to know what I am supposed to select in the "Mode:  (J)" drop down list of the printer drive when trying to get through Procedure 1.  Here are my choices:

Epson Standard
Adobe RGB
PhotoEnhance
ICM
OFF (no color adjustment)

Or does it make no difference at all what I use since, in Lightroom, I am selecting under Print Job, Profile:  "Managed by Printer?"  Maybe I missed it, but I don't see where it tells me under Procedure 1 what I am supposed to select on the driver under "Mode: (J)."

I apologize if this has already been asked.  I did a search and did not find anyone else asking my question.
Logged

Edmund Ronald, Ph.D. 
eronald
Sr. Member
****
Online Online

Posts: 3870



WWW
« Reply #8 on: July 27, 2013, 04:33:50 PM »
ReplyReply

I guess the lack of reaction to this article means profiled printing is working perfectly for everybody. What a relief!

Edmund
Logged

Edmund Ronald, Ph.D. 
Eric Myrvaagnes
Sr. Member
****
Offline Offline

Posts: 7789



WWW
« Reply #9 on: July 27, 2013, 06:55:44 PM »
ReplyReply

I guess the lack of reaction to this article means profiled printing is working perfectly for everybody. What a relief!

Edmund
Edmund,

Those of us who have finally gotten this stuff figured out may feel that (at last) we don't need your article (or: "Why wasn't it there for me X-years ago?").
But there are undoubtedly many thousands of photographers who really need your article right now, but have no idea that they need it.

I like the way you reduced the process to the essentials. The queries in this thread are largely to do with the details of terminology between different brands of printers, and if you had tried to cover all the little variants, the central ideas would likely have been lost in the details.

A good and useful piece, for those that need it.

Eric M.
Logged

-Eric Myrvaagnes

http://myrvaagnes.com  Visit my website. New images each season.
aduke
Sr. Member
****
Offline Offline

Posts: 360


« Reply #10 on: July 27, 2013, 07:12:36 PM »
ReplyReply

I actually tried the procedure, with my 4880, up to the point creating my own profile, which I do not do. I did a few other variations just to see what the results would be. My standard printing paper has been Epson Hot Press Bright and I am considering using PremierArt Matte Duo 44 for a double-sided project. Also, I have some Epson Ultra Premium Presentation Paper Matte so I through it into the mix along with some plain, 24lb, laser paper. I have Epson profiles for their papers as well as a custom profile for the HPB and the company's profile for the Matte Duo.

The results were interesting but not particularly surprising. I've not completed analysis of the results.

For me, the importance of the article is laying out a solid procedure for problem-determination in a very complex system of profiles, papers and inks. I would hope that many people keep the procedure in the back of their heads for the time when printing begins to go haywire.

Thanks, Edmund, for the article.

Alan
Logged
eronald
Sr. Member
****
Online Online

Posts: 3870



WWW
« Reply #11 on: July 27, 2013, 08:33:54 PM »
ReplyReply

I actually tried the procedure, with my 4880, up to the point creating my own profile, which I do not do. I did a few other variations just to see what the results would be. My standard printing paper has been Epson Hot Press Bright and I am considering using PremierArt Matte Duo 44 for a double-sided project. Also, I have some Epson Ultra Premium Presentation Paper Matte so I through it into the mix along with some plain, 24lb, laser paper. I have Epson profiles for their papers as well as a custom profile for the HPB and the company's profile for the Matte Duo.

The results were interesting but not particularly surprising. I've not completed analysis of the results.

For me, the importance of the article is laying out a solid procedure for problem-determination in a very complex system of profiles, papers and inks. I would hope that many people keep the procedure in the back of their heads for the time when printing begins to go haywire.

Thanks, Edmund, for the article.

Alan

Bravo for trying Smiley 

 I just got an email query; this photographer needed to be informed that his original picture needed conversion from his preferred ProphotRGB to sRGB for the procedure to get going, as Epson drivers work best with sRGB files (and if that is not the case, let's pretend it is). I guess when I'll have been informed of a few more necessary explanations, I'll amend the article.

Edmund
Logged

Edmund Ronald, Ph.D. 
hugowolf
Sr. Member
****
Offline Offline

Posts: 589


« Reply #12 on: July 28, 2013, 12:20:13 PM »
ReplyReply

I guess the lack of reaction to this article means profiled printing is working perfectly for everybody. What a relief!
The thing about the profile workflow is that it is consistant. You the same workflow whether using printer manufacture's branded paper or third party paper. So in a way it is simpler.

Brian A
Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 8608



WWW
« Reply #13 on: July 28, 2013, 12:21:16 PM »
ReplyReply

I guess the lack of reaction to this article means profiled printing is working perfectly for everybody. What a relief!

IF done correctly, yes! Anything done incorrectly, hit or miss, usually miss. True for CMS or just about any process.

Printer Manages Colors as recommended in the article (which appears to be aimed at least in part to those who CAN build their own profiles) seems counter productive. Forget soft proofing for one. Forget control over rendering intent. Forget your investment in profiling hardware and software. It IS very good advise for those that haven't much of a clue about color management or don't have a device/driver that can even support ICC output profiles.

Anyone that has the equipment to build their own profiles can easily figure out if the results are poor due to anything but the profile. Just measure the same target used to build the profile setup with the questionable result process, compare the measured data to the data used to build the profile. Having a tool like ColorThink can show us exactly where in color space there's an issue, and to what degree. Low Avg/max dE, it isn't the CMS, it's something else (document data for example). No need for this audience to use a reference image. Compare expected to resulting measured data.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
eronald
Sr. Member
****
Online Online

Posts: 3870



WWW
« Reply #14 on: July 28, 2013, 01:42:24 PM »
ReplyReply


Anyone that has the equipment to build their own profiles can easily figure out if the results are poor due to anything but the profile. Just measure the same target used to build the profile setup with the questionable result process, compare the measured data to the data used to build the profile. Having a tool like ColorThink can show us exactly where in color space there's an issue, and to what degree. Low Avg/max dE, it isn't the CMS, it's something else (document data for example). No need for this audience to use a reference image. Compare expected to resulting measured data.

Yeah, sure. And printing the target twice is going to give you two copies of the same target equally polluted by the same foobarred Mac color management "features".

My debugging procedure is carefully thought out to get around a lot of the gotchas which have been introduced in the past few years.

Oh, and by the way, Andrew did you know that the first print made after every launch of out of the box CS5 on Mac did not take the printer media settings into account properly? Switch on your machine on different days, launch CS5, print a target,  and you get exactly the same foobarred target.


Edmund
« Last Edit: July 28, 2013, 01:45:55 PM by eronald » Logged

Edmund Ronald, Ph.D. 
digitaldog
Sr. Member
****
Offline Offline

Posts: 8608



WWW
« Reply #15 on: July 28, 2013, 01:45:37 PM »
ReplyReply

Yeah, sure. And printing the target twice is going to give you two copies of the same target equally polluted by the same foobarred Mac color management "features".

If you print the target incorrectly twice, yes it will. If you do anything incorrectly more than once, you'll get incorrect results. IF you are creating your own profiles, you have the tools to evaluate if the process faults are outside that profile creation process and there are all kinds of areas where this can occur (driver update, incorrect driver settings, clogged heads, incorrect media settings etc).

Quote
My debugging procedure is carefully thought out to get around a lot of the gotchas which have been introduced in the past few years.

Some yes, many no. Using Printer Manages Color is useful for those not having any ability to build or use ICC profiles. Outside of that, not so useful.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
eronald
Sr. Member
****
Online Online

Posts: 3870



WWW
« Reply #16 on: July 28, 2013, 01:48:32 PM »
ReplyReply

IF you are creating your own profiles, you have the tools to evaluate if the process faults are outside that profile creation process and there are all kinds of areas where this can occur (driver update, incorrect driver settings, clogged heads, incorrect media settings etc).

In the interest of education, please go ahead and describe use-cases of these tools.

Edmund
Logged

Edmund Ronald, Ph.D. 
digitaldog
Sr. Member
****
Offline Offline

Posts: 8608



WWW
« Reply #17 on: July 28, 2013, 01:56:39 PM »
ReplyReply

In the interest of education, please go ahead and describe use-cases of these tools.

I don't understand the question.

That said, IF you have tools to measure patches, and you have anything that can provide a metric of difference (ColorThink), you can always evaluate if the expected values and the resulting values are close or a mile off. You don't even have to use the target that built the profile. You can create your own, big or small, from any source color space you wish. The profile has a reference value it expects it should create using any or all rendering intents. If you print that target and compare what you get with what it expects, if the values are off, it's something OTHER than that profile! Profiles don't age like cheese. The converted patches (using a rendering intent you select) is your color aim if we use the CHROMIX Maxwell language. That aim is what is expected from the conversion. And of course the aim assumes a good profile. Not one built with clogged ink heads, wrong printer settings etc. So at this point, any issues that show up can be seen in any area of color space, and we can see if the issue is or isn't the profile but rather something in the print path.

So if you have a Spectrophotometer (an assumption you outline in step 3), you have the tools to inspect process control outside the profile itself.

You either built the profile incorrectly or you built it correctly and the resulting 'issues' are a myriad of possibilities as outlined.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
eronald
Sr. Member
****
Online Online

Posts: 3870



WWW
« Reply #18 on: July 28, 2013, 04:37:05 PM »
ReplyReply

Andrew,

I don't quite get it - do you want to measure two printings of the  profile target patches, or do you want to measure a different quality assessment target, or do you want to do a reverse conversion of the profiling target measured values through the computed profile into a source image and then print the source via the profile to verify that it compares with the original profiling target?  I'm a bit lost here - but then every time I use ColorThink I feel this irresistible urge to kiss the feet of Steve Upton, and then wake up and notice three hours have passed and I need to visit the restroom Smiley

Edmund

I don't understand the question.

That said, IF you have tools to measure patches, and you have anything that can provide a metric of difference (ColorThink), you can always evaluate if the expected values and the resulting values are close or a mile off. You don't even have to use the target that built the profile. You can create your own, big or small, from any source color space you wish. The profile has a reference value it expects it should create using any or all rendering intents. If you print that target and compare what you get with what it expects, if the values are off, it's something OTHER than that profile! Profiles don't age like cheese. The converted patches (using a rendering intent you select) is your color aim if we use the CHROMIX Maxwell language. That aim is what is expected from the conversion. And of course the aim assumes a good profile. Not one built with clogged ink heads, wrong printer settings etc. So at this point, any issues that show up can be seen in any area of color space, and we can see if the issue is or isn't the profile but rather something in the print path.

So if you have a Spectrophotometer (an assumption you outline in step 3), you have the tools to inspect process control outside the profile itself.

You either built the profile incorrectly or you built it correctly and the resulting 'issues' are a myriad of possibilities as outlined.
« Last Edit: July 28, 2013, 04:41:32 PM by eronald » Logged

Edmund Ronald, Ph.D. 
digitaldog
Sr. Member
****
Offline Offline

Posts: 8608



WWW
« Reply #19 on: July 28, 2013, 05:11:50 PM »
ReplyReply

I don't quite get it

We start with a good profile (how one determines we have a really good profile is another discussion). If the profile is an issue, we're screwed from the get go.

The profile defines the color aim: what the values should be. We take a target of patches that we can build or we could take the target used to build the profiles (that ARE advantages to building our own, that's another discussion too). It's in RGB and we tag it with whatever working space we are working with. Let's say Adobe RGB (1998). We convert using our profile in the rendering intent we hope to test. We send that through the entire print process: driver, condition of the printer etc. We compare the color aim (what profile predicts) to what we actually get from the print process. We can now see an Avg and max dE, where that shows up etc. If you're really smart about it, you can repeat some of the colors in different areas of the target and you can even see where on the page there are errors which is kind of useful if you're working with presses.

This tests the entire print path: application settings, driver settings, condition of the printer at this point etc.

There are ways to just test the accuracy of the profile (tables). That's a different process which gives us different info. The process I describe helps debug the print process, it allows us to also trend the process (how far do we go outside a dE limit we decide upon, how often does this happen)? You update a new driver. You buy new paper. You put in new inks. This process debugs them as you compare the color aim to the measured results.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Pages: [1] 2 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad