Pages: [1]   Go Down

Author Topic: Starting Profiling  (Read 3020 times)

HowardG

  • Full Member
  • ***
  • Offline Offline
  • Posts: 105
    • http://www.hgrillphotographic.com
Starting Profiling
« on: September 09, 2016, 11:31:52 am »

I am just starting to try my hand at printer profiling using an i1 Pro2.  For my first attempts, and to make sure I am doing things correctly, I would like to try to reproduce (and hopefully improve upon) the 'canned' profiles from a few third party paper makers. Is it safe to assume that if the profile from, say, Hahnemuhle calls for a media choice of Heavyweight Fine Art in the Canon driver, that Heavyweight Fine Art would be the best media to profile, as opposed to starting from the question of which media setting to use? Also, any good references to be had about how to choose the best media setting if, say, I wanted to try profiling some Epson papers for a Canon printer where no media choices or profiles are available. I would love to see some examples beyond choose the one that looks best or closest to the target if possible.

Thanks.
Howard
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Starting Profiling
« Reply #1 on: September 09, 2016, 11:34:54 am »

Is it safe to assume that if the profile from, say, Hahnemuhle calls for a media choice of Heavyweight Fine Art in the Canon driver, that Heavyweight Fine Art would be the best media to profile, as opposed to starting from the question of which media setting to use?
Probably a good starting point but always test if unsure. Here's a simple file you can output using various media settings, examine the steps and pick one that separates the best:
http://digitaldog.net/files/InkDensity.zip
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Starting Profiling
« Reply #2 on: September 09, 2016, 02:26:30 pm »

I am just starting to try my hand at printer profiling using an i1 Pro2.  For my first attempts, and to make sure I am doing things correctly, I would like to try to reproduce (and hopefully improve upon) the 'canned' profiles from a few third party paper makers. Is it safe to assume that if the profile from, say, Hahnemuhle calls for a media choice of Heavyweight Fine Art in the Canon driver, that Heavyweight Fine Art would be the best media to profile, as opposed to starting from the question of which media setting to use? Also, any good references to be had about how to choose the best media setting if, say, I wanted to try profiling some Epson papers for a Canon printer where no media choices or profiles are available. I would love to see some examples beyond choose the one that looks best or closest to the target if possible.

Thanks.
Howard

A number of things here:

(1) The best guidance on Media Type selection for printing your profiling targets is the Media Type that the paper manufacturer uses for its own profiles for the printer/paper combination at hand - and they tell you what it is, either abbreviated in the profile name or in the information sheet accompanying the profile or both. The most important point here is that once you select this Media Type in the printer driver for printing your profiling targets, always use that same Media Type setting when you make prints using that profile.

(2) I presume you are using i1Profiler software for making the profiles. As X-Rite has never seen fit to produce a manual accompanying this application (a huge act of commercial and professional irresponsibility) and as there are no third party manuals on the market, and given the very poor design quality of the i1Profiler user interface, using this application to optimize your profiles can be challenging. So a few key things to watch out for.

(a) The operating mode: select the "Advanced" radio button in the lower right of the Home page to open up all the options that make using this application worthwhile.
(b) Patch set: While one can get away with a small number of patches and make an adequate profile, you'll be more assured of more accurate results using a target of at least 1600 patches. Some practitioners recommend as many as 2200 or more, but I have found 1600 works very well. Remember, the larger the target the more measuring you will be doing.
(c) Create and save workflows: For every different set of conditions you profile, create and save the workflow before you proceed to make the profile. This is partly for the convenience of future reference, but more importantly, i1Profiler is a buggy application that can quit on you at the most inconvenient moments losing everything. If you save the workflow as you go along, you can revert to the read data without having to rescan it. These are little things one learns the hard way - take my word for it; done a lot of this.
(d) Beware of the "M" Factor: The M factor deals with differences of profiling based on whether or not the paper contains OBAs (Optical Brightening Agents) or FWAs (same thing) - chemicals added to the paper that make it fluoresce producing brighter looking whites. If you know the paper contains OBAs (it likely will unless the manufacturer states unequivocally that it does not) you will want to profile under M0 and/or M1 conditions. If you are certain the paper does not contain OBAs, you can safely use the M2 condition. As long as you suspect there may be OBAs, you should do your target scanning in dual scan mode and save the workflow. In this scan mode, it will save the data for all three M conditions at once. Once you have completed measuring and you are at the Profile making tab, you can successively create three profiles from the same set of data by going back to the Measurement tab, selecting another M factor, then revert to the Profile making tab and regenerate a new profile based on the selected M factor. Make sure to give each profile a separate name that states the M factor in the name, so you will know which is which. Then to test, make a test print from each of the three profiles to see which you think best reproduces your photos.

With careful attention to detail in the workflow, you can make profiles that will surpass OEM profiles for accuracy, simply because you are characterizing the specific behaviour of your printer in your working environment. Notwithstanding the many rough edges with i1Profiler, the combination of this software and the i1Pro2 instrument can produce very high quality results.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

howardm

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1984
Re: Starting Profiling
« Reply #3 on: September 09, 2016, 02:38:11 pm »

A few personal observations:

Manually scanning sheet after sheet gets really boring.  You may want to dial down your first attempt (or 3) down to 1 sheet (which on US paper and minimal size) gets you to 728 or so patches.  When you get your feet wet, you can bump up the patch count.

I'm not sure as I'm at work right now but I think saving the workflow does *not* save the read data so I always save that separately.  Also, i1Profiler has a super annoying habit of forgetting stuff cm/in or other settings so do doublecheck.

I put a couple of tiny marks on the white surface of the scanning 'table' so I know where exactly to place the chart (left/right).  Let the charts dry 24-48 hours *covered* before scanning.  Make sure you use a utility that prints WITHOUT color management

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: Starting Profiling
« Reply #4 on: September 09, 2016, 02:45:27 pm »

A few personal observations:

Manually scanning sheet after sheet gets really boring.  You may want to dial down your first attempt (or 3) down to 1 sheet (which on US paper and minimal size) gets you to 728 or so patches.  When you get your feet wet, you can bump up the patch count.

I'm not sure as I'm at work right now but I think saving the workflow does *not* save the read data so I always save that separately.  Also, i1Profiler has a super annoying habit of forgetting stuff cm/in or other settings so do doublecheck.

I put a couple of tiny marks on the white surface of the scanning 'table' so I know where exactly to place the chart (left/right).  Let the charts dry 24-48 hours *covered* before scanning.  Make sure you use a utility that prints WITHOUT color management

As an aside, don't uncheck the box that includes scan data in the profile. That way, if you ever lose the intermediate files they can be recreated by just opening the icc profile. Once you are comfortable with the process using single page profiles you can expand to many more patch sets.
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Starting Profiling
« Reply #5 on: September 09, 2016, 02:55:15 pm »

A few personal observations:

Manually scanning sheet after sheet gets really boring.  You may want to dial down your first attempt (or 3) down to 1 sheet (which on US paper and minimal size) gets you to 728 or so patches.  When you get your feet wet, you can bump up the patch count.

I'm not sure as I'm at work right now but I think saving the workflow does *not* save the read data so I always save that separately.  Also, i1Profiler has a super annoying habit of forgetting stuff cm/in or other settings so do doublecheck.

I put a couple of tiny marks on the white surface of the scanning 'table' so I know where exactly to place the chart (left/right).  Let the charts dry 24-48 hours *covered* before scanning.  Make sure you use a utility that prints WITHOUT color management

Howard, if you save the workflow after you have read-in the data, it does save the data with the workflow. Save Workflow is the easiest way of saving everything that matters at once.

Good point on printing the targets with No Colour Management. For the Epson printers, this is easy - just use the Adobe Color Print Utility. For the new crop of Canon printers, it is not so straightforward. It depends on the printer model, the Photoshop version and operating system version. In case of uncertainty, best to discuss the specifics with Canon tech support.

I've been experimenting with dry-down time, and while your advice is conventional and correct, I have also found that much less drying time works just as well. While there is some comment that 10 minutes is enough, personally I wouldn't feel comfortable to go that far because there can be very subtle colour shifts as the glycols evaporate, but a minimum of a few hours may well suffice.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

howardm

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1984
Re: Starting Profiling
« Reply #6 on: September 09, 2016, 03:16:06 pm »

I think there was someone here on LuLa a few months ago that did a controlled test for dry-down dE (but I think it was done w/ Canon dyes, not pigments.  I dont recall the details but there was definitely a break in rate-of-change in dE after a few hours and overall, it was something like 0.4dE between fresh-from-printer vs. dried for a day or two.

HowardG

  • Full Member
  • ***
  • Offline Offline
  • Posts: 105
    • http://www.hgrillphotographic.com
Re: Starting Profiling
« Reply #7 on: September 09, 2016, 03:23:52 pm »

Andrew, Howard, Doug, and Mark.....

I can't thank you guys enough for taking the time to give me some guidance and advice. I have been 'playing' with the i1 for a week or so and on the phone with tech support at X-Rite because it appeared there was a problem with it and I have a replacement coming though it seems I may have gotten the first one intermittently working.

I have discovered several things in this quest of mine to learn more about color management. First, printing the targets through the Adobe CMM, the i1 Profiler software, and the Canon Print Pro PS plug in all give different sizes to the chart even though they are all supposedly printing at 100%. Odd. Though X-Rite says that while that would be an issue for the ISIS or the scanning table it is not problematic for the hand held spectro as long as all the patches are there and the device picked them up and didn't flash red (ie the ability to read during the scan isn't dependent on a very specific patch size).

I didn't know about saving the data in the workflow so I really appreciate that.  And I was very unsure about how best to decide what media setting is best. I was just planning to use the recommended for my reproduction of the canned profiles but when I had my 7900 there were a few Epson papers I really like but obviously they aren't making profiles or recommending media settings for the Canon printers (though you can sort of guess a handful of options that are most likely to be right based on the paper type).

Andrew, the chart you kindly placed will certainly help now that I know to pick the media that seems to separate the squares the best.

I am sure OBAs will be problematic as I learn, so I am starting with low OBA papers till I get it down (Hahnemuhle Museum Etching, William Turner, Photo Rag).

In terms of number of patches I do see the 1005 patch chart has a good number of neutrals which I suspect is a good thing to have.  I am more than willing to scan more now that I have played and see how to do it.  Andrew, I watched the icc profile video on your website and gather 1500-1800 is sort of the sweet spot and I presume if I want to increase the number of patches to that to look for charts where the patches contain a good number of neutrals.

Thanks again....I am sure I will have more questions!

Howard
www.howardgrill.com 
Logged

HowardG

  • Full Member
  • ***
  • Offline Offline
  • Posts: 105
    • http://www.hgrillphotographic.com
Re: Starting Profiling
« Reply #8 on: September 09, 2016, 03:30:16 pm »

One more Question already!

I have an Eizo CG222 (yes it is a few years old) and before this was using the Eizo's Color Navigator software to profile with an i1 Display that I had. Now I have the option of using the Color Navigator solution with the i1 Pro 2vs the large number of patches that can be read from the screen with the i1 Pro 2 and the XRite Profiler software. Which is likely to give a better display profile?

I tried setting the default to the i1 Pro 2 profile I made with the XRIte software but seem to be getting 'double profiling' with a hideous result. I am guessing this is because the Color Navigator insists on inserting its profile that I believe is hardware based and then my i1 Pro 2 profile is added.  There is a few seconds before it turns hideous so i am sure this is the problem.  I presume it could be fixed by uninstalling Color Navigator but just wondering if the non-manufacturers results (i1 Profiler software) would be expected to be superior in any way.

Thanks.

Howard
Logged

howardm

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1984
Re: Starting Profiling
« Reply #9 on: September 09, 2016, 03:33:42 pm »

HowardG

  • Full Member
  • ***
  • Offline Offline
  • Posts: 105
    • http://www.hgrillphotographic.com
Re: Starting Profiling
« Reply #10 on: September 09, 2016, 03:43:03 pm »

I tend to use Keith's media test image.  YMMV

http://www.northlight-images.co.uk/article_pages/media_test_images.html

Yes, thanks, I have gone through all of Keiths articles....they are a great resource.  The only thing I wasn't sure about is once you make the print with various media settings how do you assess which setting to use.  Is it the setting that gives the closest match on your monitor screen, the one that looks most colorful, the one that separates the squares the best as they get lightest and darkest?

Just not sure about that so any comments appreciated. The same would be true using the Canon Media Configuration Tool ink amounts...not sure how I would assess which is 'best'?

Howard
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Starting Profiling
« Reply #11 on: September 09, 2016, 03:55:49 pm »

I think there was someone here on LuLa a few months ago that did a controlled test for dry-down dE (but I think it was done w/ Canon dyes, not pigments.  I dont recall the details but there was definitely a break in rate-of-change in dE after a few hours and overall, it was something like 0.4dE between fresh-from-printer vs. dried for a day or two.

I don't recall it either. My comment is based on pigment inks. Regardless, no harm allowing longer dry-down times if feasible.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Starting Profiling
« Reply #12 on: September 09, 2016, 04:00:48 pm »

I like the Outback Photo Test Page http://www.outbackphoto.com/printinginsights/pi048/essay.html. It is comprehensive, each section was designed with a specific purpose in mind and they provide an instruction guide indicating how to interpret the results - what to look for and what deviations from satisfactory mean - very useful.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."
Pages: [1]   Go Up