Pages: 1 2 3 [4] 5 6   Go Down

Author Topic: Z3200 - Better ICC Color Profiles  (Read 2784 times)

Brad Paulson

  • Full Member
  • ***
  • Offline Offline
  • Posts: 172
Re: Z3200 - Better ICC Color Profiles
« Reply #60 on: December 06, 2017, 03:52:20 PM »

Apologies Mark.  I incorrectly thought you had supplied the files Doug was working with.  I withdraw all insinuations of lumpiness vis a vis you! 

I did like the music earlier this morning BTW.  Thanks for that.
Logged

Doug Gray

  • Sr. Member
  • ****
  • Online Online
  • Posts: 849
Re: Z3200 - Better ICC Color Profiles
« Reply #61 on: December 06, 2017, 09:41:35 PM »

Apologies Mark.  I incorrectly thought you had supplied the files Doug was working with.  I withdraw all insinuations of lumpiness vis a vis you! 

I did like the music earlier this morning BTW.  Thanks for that.

Brad,

I was the one referring to the profile's gamut surface with the adjective "lumpy". It undulates around the hue angles gamut boundary more than other profiles I've looked at. This does not have anything to do with whether a printer has banding. and is in no way an indication the profile is defective. It's a side effect of the addition of extra saturated inks the Z3200ps has. It simply has the gamut pushed out in places where the more saturated ink can be used instead of mixing CYM inks. It is harder to create printing drivers for these but HP is a highly technically competent company.
Logged

Brad Paulson

  • Full Member
  • ***
  • Offline Offline
  • Posts: 172
Re: Z3200 - Better ICC Color Profiles
« Reply #62 on: December 06, 2017, 10:16:33 PM »

Thanks for that detailed explanation Doug.  Makes sense now. 
Logged

Ernst Dinkla

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3534
Re: Z3200 - Better ICC Color Profiles
« Reply #63 on: December 07, 2017, 06:23:59 AM »

I should check it this weekend again but I got the impression that when HP APS does not recognize the puck correctly it still allows profile creation from already measured data files. If that is correct it might be possible to use Color Center for the measurement and bring the file to APS with some text changes .

I had computer issues and tried to reinstall HP APS (I kept the registration key etc) but it gave issues on the puck recognition.


Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
March 2017 update, 750+ inkjet media white spectral plots
Logged

Mark Lindquist

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1249
  • itís not about the photos we take - itís the ones we leave
Re: Z3200 - Better ICC Color Profiles
« Reply #64 on: December 07, 2017, 07:33:15 AM »

Kers has APS and has used it to make profiles.  I wonder though, Ernst, the APS uses Gretag MacBeth logarithms which are not the same as i1 that is in the current Z3200ps machines, yes?

Kers remarked he saw banding coming from the custom profiles made with APS.  There could be other causes for that as well, but curious to know your view on how the APS from Z3100 days relates to the current i1 spectrophtometer in the current Z3200ps machines?

Hope you can get your APS running again.

Mark

kers

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1668
    • Pieter Kers
Re: Z3200 - Better ICC Color Profiles
« Reply #65 on: December 07, 2017, 08:14:36 AM »

Kers has APS and has used it to make profiles.  I wonder though, Ernst, the APS uses Gretag MacBeth logarithms which are not the same as i1 that is in the current Z3200ps machines, yes?

Kers remarked he saw banding coming from the custom profiles made with APS.  There could be other causes for that as well, but curious to know your view on how the APS from Z3100 days relates to the current i1 spectrophtometer in the current Z3200ps machines?

Hope you can get your APS running again.

Mark

Hai Mark,

yes i saw some banding, but i think it is wise not to draw any conclusion upon this observation;
you really have to be with your head into color and into Z to fully understand what is going on. And then do a lot of testing.
I mean to say that it is very difficult to say at this moment what causes what and why.  This matter is just too complicated, with too many factors that influence the process.
- if i test one lens i already can spend hours on it to see what it does exactly- usually already complicated. (and to conclude that tests on the internet jump too fast to (wrong) conclusions)
so i stay at the side of this - very interesting - discussion, also because i do not have the knowledge nor the means ( spectrophotometer etc) necessary to find out what goes on exactly.
In this case i will be the engineer; i do not know exactly why it works , but if it does i am satisfied.

Looking forward to what the more experienced - like you - find out.

a good thing to keep the discussion to the Z3200 - - the Z3100 and the old APS make things only less clear.

cheers
Pieter





Logged
Pieter Kers
www.beeld.nu

Ernst Dinkla

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3534
Re: Z3200 - Better ICC Color Profiles
« Reply #66 on: December 07, 2017, 11:50:14 AM »

Kers has APS and has used it to make profiles.  I wonder though, Ernst, the APS uses Gretag MacBeth logarithms which are not the same as i1 that is in the current Z3200ps machines, yes?

Kers remarked he saw banding coming from the custom profiles made with APS.  There could be other causes for that as well, but curious to know your view on how the APS from Z3100 days relates to the current i1 spectrophtometer in the current Z3200ps machines?

Hope you can get your APS running again.

Mark

When I had only a Z3100 the APS was an optional thing and I got it. When the Z3200-PS arrived here there was an extra APS  license with it. There were some version changes over the 11 years I used it; v131, v140, v141, I have them archived.  What I recall though was a change in the headings of the created profiles in the switch from Z3100 to Z3200, as if the color engine / underlying profile creator, changed per machine. I must have written about this in the past. I did not observe much difference between the profile qualities made with either and Color Center improved a lot between the Z3100 and Z3200 that its last version is equal to APS so I did not pay much attention to that difference.



Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
March 2017 update, 750+ inkjet media white spectral plots
Logged

Mark Lindquist

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1249
  • itís not about the photos we take - itís the ones we leave
Re: Z3200 - Better ICC Color Profiles
« Reply #67 on: December 07, 2017, 11:52:23 AM »

Hai Mark,

yes i saw some banding, but i think it is wise not to draw any conclusion upon this observation;
you really have to be with your head into color and into Z to fully understand what is going on. And then do a lot of testing.
I mean to say that it is very difficult to say at this moment what causes what and why.  This matter is just too complicated, with too many factors that influence the process.
- if i test one lens i already can spend hours on it to see what it does exactly- usually already complicated. (and to conclude that tests on the internet jump too fast to (wrong) conclusions)
so i stay at the side of this - very interesting - discussion, also because i do not have the knowledge nor the means ( spectrophotometer etc) necessary to find out what goes on exactly.
In this case i will be the engineer; i do not know exactly why it works , but if it does i am satisfied.

Looking forward to what the more experienced - like you - find out.

a good thing to keep the discussion to the Z3200 - - the Z3100 and the old APS make things only less clear.

cheers
Pieter

Hi Pieter-

Yes, agreed.  Extremely complicated.
Mark McCormick and I are currently doing ICC profiles at the same time - he in Massachusetts and I in Florida.

We have successfully made 2 targets and have read them both:  1457 and 1877 from charts Mark made with i1Profiler.
I'm doing an 1877 patch target now, as I speak.  MHMG ran into a glitch when attempting to print a 2371 patch target - he's printing using the Mac Utility and I'm using the Windows Utility running on VMWare Fusion for Mac and mine is still reading so far.

We're learning a lot and it looks like we will be able to crack this nut soon.

Going to do an article and tutorial on Z3200.com and include profiles, etc.

Thanks bud - appreciate all you do.

Best -

Mark L

Mark Lindquist

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1249
  • itís not about the photos we take - itís the ones we leave
Re: Z3200 - Better ICC Color Profiles
« Reply #68 on: December 07, 2017, 12:00:26 PM »

When I had only a Z3100 the APS was an optional thing and I got it. When the Z3200-PS arrived here there was an extra APS  license with it. There were some version changes over the 11 years I used it; v131, v140, v141, I have them archived.  What I recall though was a change in the headings of the created profiles in the switch from Z3100 to Z3200, as if the color engine / underlying profile creator, changed per machine. I must have written about this in the past. I did not observe much difference between the profile qualities made with either and Color Center improved a lot between the Z3100 and Z3200 that its last version is equal to APS so I did not pay much attention to that difference.



Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
March 2017 update, 750+ inkjet media white spectral plots

Hi Ernst - thanks for that - it helps fill in puzzle pieces. Do you recall what was the last year they upgraded APS?

You also mentioned that the Software would still make an ICC profile without the dongle.  That's interesting.
Would like to get my hands on a copy and the upgrades.

So MHMG and I are making some good progress - will keep everyone posted with the results eventually.

Thanks Ernst -

Mark


Brad Paulson

  • Full Member
  • ***
  • Offline Offline
  • Posts: 172
Re: Z3200 - Better ICC Color Profiles
« Reply #69 on: December 07, 2017, 01:01:21 PM »


Going to do an article and tutorial on Z3200.com and include profiles, etc.


Please post a link here when the nut is cracked.  Donít want to miss it! 
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 12512
    • http://digitaldog.net/
Re: Z3200 - Better ICC Color Profiles
« Reply #70 on: December 07, 2017, 01:10:34 PM »

You also mentioned that the Software would still make an ICC profile without the dongle.
Note that some 'copy protection' is built directly into some measuring instruments. It IS the dongle. Copra's 'dongle' is a USB stick.
Logged
Andrew Rodney
Author ďColor Management for PhotographersĒ
http://digitaldog.net/

Mark Lindquist

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1249
  • itís not about the photos we take - itís the ones we leave
Re: Z3200 - Better ICC Color Profiles
« Reply #71 on: December 07, 2017, 01:12:38 PM »

Note that some 'copy protection' is built directly into some measuring instruments. It IS the dongle. Copra's 'dongle' is a USB stick.

Thanks for that reminder Andrew - good point. - Will be very careful.

Mark

Mark Lindquist

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1249
  • itís not about the photos we take - itís the ones we leave
Re: Z3200 - Better ICC Color Profiles
« Reply #72 on: December 07, 2017, 02:59:56 PM »

OK.  I've just started the 6000 patch target on my Z3200ps 44" printer.  Whoever thinks the HP Z3200ps printer is not a formidable printer with the ability to create profiles of all kinds and sizes, better think twice.

Not only is the Z3200ps Vivera Inkset still King of the Hill, the printer continues to prove itself absolutely the Swiss Army Knife of machines for the small to medium studio photographer/printer.

I ask myself, WHY would you want any other printer when everything else is huge, heavy, incredibly overbuilt and expensive to maintain when there is the Z3200ps still running on the brilliant design brought forth from the Barcelona team using 15 year old ink technology which has time and time again proved itself in head to head tests, still to this day, and all the other attributes that make this printer utterly a pleasure to own and use.

Granted, not everyone will make a 6000 patch target ICC profile, but it can be done.

I'm doing it.  Just because I can.  Just because this printer can.

Thanks MHMG for the Chart.

Off we go.  Will report back in.

Mark


MHMG

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1023
Re: Z3200 - Better ICC Color Profiles
« Reply #73 on: December 07, 2017, 03:53:40 PM »


I'm doing it.  Just because I can.  Just because this printer can.

Thanks MHMG for the Chart.

Off we go.  Will report back in.

Mark


I continue to have some glitches trying to run patch counts larger than 2000 on the Mac version of the HP Utility. This may be a software bug I can't circumvent with the Mac driver alone, but as Mark L. has already proven, the VMWare Fusion app opens up access to running the HP windows driver version on my Mac, so this avenue is a proven workaround if all else fails. That said, I've successfully auto printed and measured 1457, 1728, and 1877 targets on the Mac. Delightful just to sit back and let the HP Z3200 do its thing and save out a measurement file that can be dragged quickly onto BasicColor RGB Drop (or Argyll with iCCgen as a simple GUI to avoid terminal) and a nice profile built in no time. i1Profiler requires a little text cleanup of the HP generated data file as Doug Grey has shown us, but the HP Z-measured text file then ports easily into the i1Profiler software as well. It's all good! Any state of the art ICC profiling solution is likely to work.

The profile I built with the 1457 patch count for Moab Entrada Natural (my main go-to fine art matte paper) is remarkably excellent. I'd be perfectly happy just to live with the 1457 patch count printed and measured with utter simplicity on my Z3200.  i1Profiler does indeed max out the percentage of neutral/near neutral patches in these oddly chosen patch count files as reported by Ethan Hansen, et. al.. The HP printer utility sucks those patch target txt files in without any trouble, then creates an efficient layout of the target based on sheet/roll size selected and prints it. The 1457 patch file used 1 foot of paper on the 44 inch roll I had loaded in the printer, a reasonable economy given that I'd need 3 letter-sise sheets of paper to print this target from i1Profiler for measurement with my hand-held i1Pro2 spectrophotometer. No more having to go into adobe color printer utility or Apple CSU to print out "no color adjust" targets on multiple sheets of paper :)

Although I've admired the Vivera ink permanence since the beginning, I'm still a relative newcomer to the Z3200 owner's club. I've had my Z for about a year now, and I like it more and more as each day passes.

kind regards,
Mark
http://www.aardenburg-imaging.com
« Last Edit: December 07, 2017, 05:26:24 PM by MHMG »
Logged

Brad Paulson

  • Full Member
  • ***
  • Offline Offline
  • Posts: 172
Re: Z3200 - Better ICC Color Profiles
« Reply #74 on: December 07, 2017, 11:08:09 PM »

Doug, do you know offhand whether the Bills Balls print test would reveal the lumpiness caused by saturated inks used at gamut boundry that you referred to earlier, and whether that would be somewhat ameliorated by more patches than 1728? 

As background, I printed four samples in P and R out of LR and PS with what I believe an as well as can be done 1728 patch on a very smooth baryta coated almost gloss paper using DropRGB as the engine, Iím still seeing room for improvement with that test.  (I believe some of these results could be from imaginary OOG Lab colors that are presently insolvable.)

Attached is an iPad shot of what Iím looking at.  The pics on the other side appear excellent but just hoping to see all the Z can do.
« Last Edit: December 08, 2017, 01:55:19 AM by Brad P »
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 12512
    • http://digitaldog.net/
Re: Z3200 - Better ICC Color Profiles
« Reply #75 on: December 08, 2017, 10:41:07 AM »

Be somewhat careful with Bills Balls (not joking). They are synthetic and may contain 'imaginary' colors which is something interesting to use on profiles to see how they deal with them.
The idea behind the balls:


Quote
Andrew,
I intended this image to be used without a profile, and the answer about imaginary colors would completely depend on what RGB source profile you tagged it with.  All RGB numeric values do fall in the range of 0.0 to full-scale, so if you tagged the image with sRGB, then all colors on all balls would of course be within the gamut of sRGB. You can open the file in Photoshop and assign any RGB source profile to it, then look at the resulting soft-proof for a given printer profile.


Bill
Logged
Andrew Rodney
Author ďColor Management for PhotographersĒ
http://digitaldog.net/

Doug Gray

  • Sr. Member
  • ****
  • Online Online
  • Posts: 849
Re: Z3200 - Better ICC Color Profiles
« Reply #76 on: December 08, 2017, 12:18:03 PM »

Be somewhat careful with Bills Balls (not joking). They are synthetic and may contain 'imaginary' colors which is something interesting to use on profiles to see how they deal with them.
The idea behind the balls:
I agree with Andrew here. Exercise care with these. They (when in ProPhoto) are useful to get a sense of how imaginary as well as highly saturated colors are mapped to ones the printer can print. It's more problematic when used with printers with extra, highly saturated inks. These inks allow the gamut to stick out a bit in portions of the gamut where they provide enhanced color capability. The downside is that this also can reduce smoothness in things like grainger gradients and Bill's Balls. Especially when the source image is a huge space like ProPhoto. Once a requested color is outside of a printer's gamut there is no specification as to how these colors will be mapped.

This is one of the reasons soft proofing is so valuable. Especially when using ProPhoto RGB because the colors shown should, and in my experience do, reflect what will be printed quite well.  You will see the same banding soft proofing Bill's Balls that the actual prints show.

To see what they look like using them according to Bill (from Andrews comment), assign the printer profile to the Ball's RGB image. Then convert back to ProPhoto RGB using Relative Intent but with BPC unselected.  It's a little tricky to actually print them per Bill's recommendation. You need to print them without color management like one would print a profiling target sheet. See Adobe's ACPU.
Logged

Doug Gray

  • Sr. Member
  • ****
  • Online Online
  • Posts: 849
Re: Z3200 - Better ICC Color Profiles
« Reply #77 on: December 08, 2017, 01:11:24 PM »

Doug, do you know offhand whether the Bills Balls print test would reveal the lumpiness caused by saturated inks used at gamut boundry that you referred to earlier, and whether that would be somewhat ameliorated by more patches than 1728? 
No, More patches will usually produce a more colorimetrically accurate profile but will not materially change the rendering of Bill's Balls from ProPhoto. A portion of them are imaginary, and most of the rest, aside from the neutral colored ones, are out of gamut so will be rendered based on the profile software designer's chosen algorithm. It's pretty much arbitrary and you will see big differences between Argyll and I1Profiler amongst others. Even when using the same exact scan data to make profiles.

BTW, these differences are a principal reason for Andrew's video that "All Profiles are Not Created Equal."  Profile s/w should create the Colorimetric portions of profiles very closely as that is well defined. But even there, some s/w doesn't conform to the latest ICC specs, etiher V2 or V4. In particular the canned ones that came with my Canon 9500 II and Epson 9800 are defective but in different ways.

That said, perfectly well formed, "correct," profiles can and do have very different ways of doing Perceptual intent and especially how they map OOG colors in any Intent.

There's not a right or wrong for these as it's not a measureable attribute. My preference is that imaginary and OOG colors be printed with some similarity to the way they normally look in a wide gamut monitor. These colors are also OOG for the monitor but, unlike printers, monitors have well defined algorithms for how these colors are converted and what color results. Monitors that are similar to the Adobe RGB gamut will show Bill's Balls quite predictably.

Quote
As background, I printed four samples in P and R out of LR and PS with what I believe an as well as can be done 1728 patch on a very smooth baryta coated almost gloss paper using DropRGB as the engine, Iím still seeing room for improvement with that test.  (I believe some of these results could be from imaginary OOG Lab colors that are presently insolvable.)

Attached is an iPad shot of what Iím looking at.  The pics on the other side appear excellent but just hoping to see all the Z can do.

Doesn't concern me at all. I'd be surprised not to see it. Especially on a wide gamut printer with extra saturated inks. Has nothing to do with how well it prints in gamut colors and it's really something that the profile software determines and what the programmer considered important. You will see very different results with different s/w.
« Last Edit: December 08, 2017, 01:24:44 PM by Doug Gray »
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 12512
    • http://digitaldog.net/
Re: Z3200 - Better ICC Color Profiles
« Reply #78 on: December 08, 2017, 01:13:33 PM »

There's not a right or wrong for these as it's not a measureable attribute. My preference is that imaginary and OOG colors be printed with some similarity to the way they normally look in a wide gamut monitor.
Agreed. I've seen some profiles render Bill's balls to nearly all black and that's rather ugly. I've seen others rendering them so they appear blue. Much more visually pleasing.
Logged
Andrew Rodney
Author ďColor Management for PhotographersĒ
http://digitaldog.net/

Doug Gray

  • Sr. Member
  • ****
  • Online Online
  • Posts: 849
Re: Z3200 - Better ICC Color Profiles
« Reply #79 on: December 08, 2017, 01:26:44 PM »

Agreed. I've seen some profiles render Bill's balls to nearly all black and that's rather ugly. I've seen others rendering them so they appear blue. Much more visually pleasing.

Yup. My 9800 canned ones do that! But they are made incorrectly too. Oh well.  BTW, I added a reference to your video about how these vary in my earlier comment while you were posting this.
Logged
Pages: 1 2 3 [4] 5 6   Go Up