Pages: 1 [2]   Go Down

Author Topic: making printer profiles for the first time...  (Read 7616 times)

smthopr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 612
    • Bruce Alan Greene Cinematography
Re: making printer profiles for the first time...
« Reply #20 on: March 12, 2018, 06:21:08 pm »

This is extremely odd. M1 is much closer to M0 with the I1Pro 2 (and iSis for that matter) so you should see less difference between M0 and M1 than M0 and M2. No idea what you mean by "I did notice the the M1 profile does not allow a preview of paper white."  Is this some "feature" of the Fiery" software? Makes no sense to me.

Photoshop's view proof lets you select show paper white with all Intents and doesn't care about how a profile was created.
Well...  With the M1 profile from the EFI software, the "show paper white" box in photoshop is grey and can not be checked.  I did not try to make an M2 profile as my paper has a bit of OBAs.
Logged
Bruce Alan Greene
www.brucealangreene.com

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: making printer profiles for the first time...
« Reply #21 on: March 12, 2018, 06:25:04 pm »

Well...  With the M1 profile from the EFI software, the "show paper white" box in photoshop is grey and can not be checked.
For all Rendering Intents? Sounds like a simple matrix profile....
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: making printer profiles for the first time...
« Reply #22 on: March 12, 2018, 06:36:01 pm »

For all Rendering Intents? Sounds like a simple matrix profile....

Does any printer profiling s/w even have the option of making matrix profiles? They are completely inappropriate for printer profiles.

This is getting weirder and weirder.
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: making printer profiles for the first time...
« Reply #23 on: March 12, 2018, 06:38:30 pm »

Does any printer profiling s/w even have the option of making matrix profiles? They are completely inappropriate for printer profiles.
Yes but that's the only way I can get the same behavior as reported whereby the check box is grayed out.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

smthopr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 612
    • Bruce Alan Greene Cinematography
Re: making printer profiles for the first time...
« Reply #24 on: March 12, 2018, 07:42:26 pm »

Yes but that's the only way I can get the same behavior as reported whereby the check box is grayed out.
I don't think there was any option about matrix profiles in the software...

Perhaps EFI is not the best software, but it's all I have a license for.  I originally bought the spectro used to create i1 display pro offsets for creating accurate 3d display LUTs and it works well for that, so I got my money's worth :)  For me, the printer profiles are a bonus, but now that I'm making them, they seem quite good, but with a few quirks :) ;) :)
Logged
Bruce Alan Greene
www.brucealangreene.com

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: making printer profiles for the first time...
« Reply #25 on: March 12, 2018, 07:46:10 pm »

I don't think there was any option about matrix profiles in the software...
It's a major assumption and a stretch and without having access to the profile, that's the best I can do. I can't replicate what you report about this inability to check the box for paper simulation with any profile but a matrix profile. So maybe you can upload the profile?
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

GWGill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 608
  • Author of ArgyllCMS & ArgyllPRO ColorMeter
    • ArgyllCMS
Re: making printer profiles for the first time...
« Reply #26 on: March 12, 2018, 07:54:55 pm »

Does any printer profiling s/w even have the option of making matrix profiles? They are completely inappropriate for printer profiles.
Sure - if it's RGB then you can generate a matrix printer profile using ArgyllCMS if you choose the right option, but it's not the default. No, I wouldn't expect the result to be anything other than a curiosity.
Logged

smthopr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 612
    • Bruce Alan Greene Cinematography
Re: making printer profiles for the first time...
« Reply #27 on: March 12, 2018, 08:17:27 pm »

It's a major assumption and a stretch and without having access to the profile, that's the best I can do. I can't replicate what you report about this inability to check the box for paper simulation with any profile but a matrix profile. So maybe you can upload the profile?

I've tried attaching the profile here.  Hope it works :)
Logged
Bruce Alan Greene
www.brucealangreene.com

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: making printer profiles for the first time...
« Reply #28 on: March 12, 2018, 08:47:34 pm »

I've tried attaching the profile here.  Hope it works :)
It doesn't which is really, really odd! And when I try to look at some of the profile tags in the ColorSync Utility, it crashes (repeatedly). If I run Profile First Aid, it crashes. IF I delete your profile and run it again, it doesn't crash. Something here isn't Kosher. But heck, this is EFI, one messed up company....
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: making printer profiles for the first time...
« Reply #29 on: March 12, 2018, 09:15:39 pm »

It doesn't which is really, really odd! And when I try to look at some of the profile tags in the ColorSync Utility, it crashes (repeatedly). If I run Profile First Aid, it crashes. IF I delete your profile and run it again, it doesn't crash. Something here isn't Kosher. But heck, this is EFI, one messed up company....

It's a "standard" LUT printer profile operating out of LAB PCS. And that's about the only thing that makes any sense.

I'm able to examine the profile with Profile Inspector from color.org.  This is the most insane profile I've ever seen! There are 10 points on each axis for the AtoB1 tables and 62!!!! points on the BtoA1 tables. OK, that's pretty weird, but get this. The media white point tag has Y at 1.15.  That is simply crazy. 100% reflectivity is Y=1.00 (L*==100) and most media comes in between .85 and .93.

My guess is that Photoshop grays out the view paper white because it is outside any real life range.

You might want to consider making profiles with Argyll. It works. It's free.
« Last Edit: March 12, 2018, 10:55:11 pm by Doug Gray »
Logged

smthopr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 612
    • Bruce Alan Greene Cinematography
Re: making printer profiles for the first time...
« Reply #30 on: March 12, 2018, 11:39:16 pm »

It's a "standard" LUT printer profile operating out of LAB PCS. And that's about the only thing that makes any sense.

I'm able to examine the profile with Profile Inspector from color.org.  This is the most insane profile I've ever seen! There are 10 points on each axis for the AtoB1 tables and 62!!!! points on the BtoA1 tables. OK, that's pretty weird, but get this. The media white point tag has Y at 1.15.  That is simply crazy. 100% reflectivity is Y=1.00 (L*==100) and most media comes in between .85 and .93.

My guess is that Photoshop grays out the view paper white because it is outside any real life range.

You might want to consider making profiles with Argyll. It works. It's free.

Hmmm. seems pretty weird, even if I don't know what AtoB1 and BtoA1 are!

Crazy thing is of course, that the prints look pretty darn good and match the softproof very well.  (of course without the paper white preview, but I keep my display down to 60 white level when printing.)

I've looked at Argyll, and it looks, to me, like a big challenge as I'm not familiar with working with the command line interface.  So a pretty darn good learning curve here.  I do use DisplayCal though for display calibration for video and even with it's GUI, the documentation is confusing as .... heck.

Anyway, all is pretty strange as the EFI software has so few settings to choose from when making a profile, it's hard to believe it just doesn't make a correct, though simple, profile.
Logged
Bruce Alan Greene
www.brucealangreene.com

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: making printer profiles for the first time...
« Reply #31 on: March 13, 2018, 12:11:54 am »

Hmmm. seems pretty weird, even if I don't know what AtoB1 and BtoA1 are!

Crazy thing is of course, that the prints look pretty darn good and match the softproof very well.  (of course without the paper white preview, but I keep my display down to 60 white level when printing.)

I've looked at Argyll, and it looks, to me, like a big challenge as I'm not familiar with working with the command line interface.  So a pretty darn good learning curve here.  I do use DisplayCal though for display calibration for video and even with it's GUI, the documentation is confusing as .... heck.

Anyway, all is pretty strange as the EFI software has so few settings to choose from when making a profile, it's hard to believe it just doesn't make a correct, though simple, profile.

The media white point is of importance to Abs. Col. but not the other intents, which just go to full scale. Abs. Col. is used to show paper white. That is, normally Lab 100,0,0 doesn't print anything so is automatically at the media white point. Abs. Col. reads the media white point tag and corrects the requested printed color to match. Your profile is saying that the media white point is around L*=105 or so which, of course, isn't possible.

As for the BtoA1 tables, these are huge, really huge, and completely inconsistent with the reverse table sizes which are less than 1% as big byte wise. They are used to lookup and convert Lab values to the device RGB values for printing. That they are so large doesn't mean they don't work. It's just really odd.
« Last Edit: March 13, 2018, 12:35:17 am by Doug Gray »
Logged

GWGill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 608
  • Author of ArgyllCMS & ArgyllPRO ColorMeter
    • ArgyllCMS
Re: making printer profiles for the first time...
« Reply #32 on: March 13, 2018, 02:48:56 am »

There are 10 points on each axis for the AtoB1 tables and 62!!!! points on the BtoA1 tables.

11 on A2B0 and A2B1
31 on B2A0
62 on B2A1
16 on B2A2

Some of these seem a bit high for a 918 patch chart.

All the cLUTs are XYZ that make use of the XYZ matrix. This is a little unusual, since
many profilers opt for Lab PCS for printers, which trades a little linearity within
interpolation cells for better overall control of the conversion.
(I have heard of HP print profiles using the XYZ cLUT with matrix approach.)

Quote
The media white point tag has Y at 1.15.  That is simply crazy. 100% reflectivity is Y=1.00 (L*==100) and most media comes in between .85 and .93.
I've come across measurement values > 100% in some circumstances - typically high OBE containing paper, with an instrument that has drifted somewhat, but the CGATS data shows that there is something strange going on - the spectral data shows reflectivities up to 120% at long wavelengths.
Logged

smthopr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 612
    • Bruce Alan Greene Cinematography
Re: making printer profiles for the first time...
« Reply #33 on: March 13, 2018, 10:46:54 am »

11 on A2B0 and A2B1
31 on B2A0
62 on B2A1
16 on B2A2

Some of these seem a bit high for a 918 patch chart.

All the cLUTs are XYZ that make use of the XYZ matrix. This is a little unusual, since
many profilers opt for Lab PCS for printers, which trades a little linearity within
interpolation cells for better overall control of the conversion.
(I have heard of HP print profiles using the XYZ cLUT with matrix approach.)
I've come across measurement values > 100% in some circumstances - typically high OBE containing paper, with an instrument that has drifted somewhat, but the CGATS data shows that there is something strange going on - the spectral data shows reflectivities up to 120% at long wavelengths.

While I don't understand all this detail yet, I'm going to attach the M0 profile.  Is this one so strange as well?

And a stupid question:  I read the patches in a room with the shades closed and the lights off.  But there was still some leakage through the window and the computer displays were on. I did have the patches on the table near the displays.  Might those light sources have contaminated the measurements?  Should I always take the readings in the most dark possible room?

Logged
Bruce Alan Greene
www.brucealangreene.com

smthopr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 612
    • Bruce Alan Greene Cinematography
Re: making printer profiles for the first time...
« Reply #34 on: March 13, 2018, 03:50:29 pm »

I re-measured the target one more time being careful to minimize stray light on the target.  This profile seems a bit better and the "simulate paper white" now works.

Would any of you mind to take a quick peek and let me know if this looks like a normal profile?

Thanks
Logged
Bruce Alan Greene
www.brucealangreene.com

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: making printer profiles for the first time...
« Reply #35 on: March 13, 2018, 04:22:20 pm »

I re-measured the target one more time being careful to minimize stray light on the target.  This profile seems a bit better and the "simulate paper white" now works.

Would any of you mind to take a quick peek and let me know if this looks like a normal profile?

Thanks

Yep, Big differences and this one looks fine. Media white point Y=.86, which is pretty typical. Also works fine in Photoshop.

I also verified that it was the media white point XYZ values that were preventing Photoshop from showing paper white. I edited the numbers by multiplying each one by .7 and the profile worked fine in Photoshop. Very similar to your new one.

I had missed the header tag that they use XYZ as the PCS instead of LAB. I've never seen a printer profile that does that though it is completely legal. It also explains why there is such a large difference in 3DLUT points but are needed going from XYZ to device RGB. XYZ is a huge colorspace but, unlike ICCLAB, contains all physically realizable colors. ICCLAB does not due to a* and b* clipping.
Logged

smthopr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 612
    • Bruce Alan Greene Cinematography
Re: making printer profiles for the first time...
« Reply #36 on: March 13, 2018, 04:32:13 pm »

Yep, Big differences and this one looks fine. Media white point Y=.86, which is pretty typical. Also works fine in Photoshop.

I also verified that it was the media white point XYZ values that were preventing Photoshop from showing paper white. I edited the numbers by multiplying each one by .7 and the profile worked fine in Photoshop. Very similar to your new one.

I had missed the header tag that they use XYZ as the PCS instead of LAB. I've never seen a printer profile that does that though it is completely legal. It also explains why there is such a large difference in 3DLUT points but are needed going from XYZ to device RGB. XYZ is a huge colorspace but, unlike ICCLAB, contains all physically realizable colors. ICCLAB does not due to a* and b* clipping.
Thanks so much Doug!

I think that, maybe, stray light from the computer display effected the measurements.  This time I measured in the darkest room possible and turned the computer display away from the target print.

I made one print with the new profile and it looks correct :)  Now to go back and re-measure my two other papers...
Logged
Bruce Alan Greene
www.brucealangreene.com
Pages: 1 [2]   Go Up