Pages: 1 2 3 [4] 5 6 7   Go Down

Author Topic: Unexpected result after converting image profile to softproof profile  (Read 13598 times)

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1829
Re: Unexpected result after converting image profile to softproof profile
« Reply #60 on: March 24, 2018, 03:20:21 pm »

Mark,
This link lists a few places with the printer.
https://www.usa.canon.com/internet/portal/us/home/explore/product-showcases/printer-technology/dreamlabo-5000

The specs say it takes image inputs in either sRGB or Adobe RGB.

The 7th one in India was recently installed. Runs a bit over half a million USD.
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12510
    • http://www.markdsegal.com
Re: Unexpected result after converting image profile to softproof profile
« Reply #61 on: March 24, 2018, 03:24:00 pm »

Thanks Doug.

Yes - pricing with options goes as high as 700K I think. Needs quite a volume of business to justify that kind of investment.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 15547
    • http://www.digitaldog.net/
Re: Unexpected result after converting image profile to softproof profile
« Reply #62 on: March 24, 2018, 04:03:27 pm »

I can well understand that. The books one gets from Blurb and many other providers are produced on HP Indigo machines or similar and the gamut is not the widest, but many photos don't require huge gamut...
Keep in mind there are multiple Indigo press technologies, some with extended inks and gamut’s.
Logged
Andrew Rodney
Author “Color Management for Photographers"

Stephen Ray

  • Full Member
  • ***
  • Offline Offline
  • Posts: 173
Re: Unexpected result after converting image profile to softproof profile
« Reply #63 on: March 24, 2018, 04:04:00 pm »

So, what I'm saying is that this sRGB process can work well.  Yes, there are colors that my Epson can print that this lab can not, but really it's not that big of a deal.  The book looks fantastic and very much worth the $200 that I didn't spend myself!!!!

Yes, not that big of a deal for the greater population and a rather easy and rewarding workflow for most. In a more commercial photography or commercial printing field however, using identifiable stress-test colors such as some Pantone's and corporate colors, etc., the limits of gamut readily become noticeable and then the necessity of good color management is indispensable. Alas, there are still limits.
Logged

Stephen Ray

  • Full Member
  • ***
  • Offline Offline
  • Posts: 173
Re: Unexpected result after converting image profile to softproof profile
« Reply #64 on: March 24, 2018, 04:11:29 pm »

Rough data chunks?
Logged

smthopr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 545
    • Bruce Alan Greene Cinematography
Re: Unexpected result after converting image profile to softproof profile
« Reply #65 on: March 24, 2018, 05:07:30 pm »

I can well understand that. The books one gets from Blurb and many other providers are produced on HP Indigo machines or similar and the gamut is not the widest, but many photos don't require huge gamuts or can be squeezed in under softproofing satisfactorily, so the results come back looking OK; and the process you describe for preparing the photos is a good way to go. There is one relatively new machine on the market that I saw demonstrated about a year ago - the Canon DreamLabo 5000; it's an inkjet production printer for making photobooks, etc. and the output is stunning. Unfortunately, this machine is enormous and costs a fortune (https://www.youtube.com/watch?v=OfE5k0wClWA). There are extremely few in operation on the North American continent, so not obvious where to find firms using them, but it's the high-end of the industry and would deserve the best possible "pre-press" colour management of the photos being sent to it.

The company that made my book uses the fuji photographic paper process as they fold the sheets in half to make the book.  In this way one can, sort of, successfully span the fold with a photo that takes up facing pages.  And the facing pages lay flat to the viewer.  I don't think inkjet papers would survive the fold so well without extensive cracking.  The RC photo paper seems to do quite well with this.  I had just wished that the lab that they used also provided the true monochrome papers for b&w, but alas, only color prints of the b&w.  They look ok, but the color changes quite a bit with the viewing lightsource...

Logged
Bruce Alan Greene
www.brucealangreene.com

walter.sk

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1428
Re: Unexpected result after converting image profile to softproof profile
« Reply #66 on: March 24, 2018, 05:58:58 pm »

NO, don't do that. Pointless and only useful to clip colors you possibly produced and can output (but maybe NOT see on-screen).
See:

The benefits of wide gamut working spaces on printed output:

This three part, 32 minute video covers why a wide gamut RGB working space like ProPhoto RGB can produce superior quality output to print.
Thank you, Andrew.  The video was highly useful for me.  I highly recommend that anyone doing color management and softproofing view all three parts.  The presentation is meticulous in approach, with surprisingly good examples.  I was surprised that after at least 18 years of my using color management techniques on my prints that I still would have missed some of the not-so-subtle differences in color rendition that you pointed out.

Nowadays it is rare to have a person who is not only knowledgeable but able to present the material in such a clear fashion. 

How about running for president?   ;)
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1829
Re: Unexpected result after converting image profile to softproof profile
« Reply #67 on: March 24, 2018, 06:36:04 pm »

Rough data chunks?

I looked at the gamut taking slices through L*. It has some unusual shapes but nothing that suggests anything significantly wrong with it. The printer is relatively wide gamut and assigning the profile to an sRGB image will fail badly but processing it in accord with DryCreek's instructions should work fine and you get a large gamut.

Examining the profile, it has larger 3DLUT sizes (37 grid points - same as high quality I1P) than most of the Costco profiles and was likely made with a large patch set though spectral data isn't included in the profile.

What I find completely mind boggling is that the ink set is four CYMK but, rather than LM and LC it has "pink" and "Sky Blue."  This could be harder to smoothly combine than traditional ink sets. I'm guessing this has to do with getting longer print runs out of ink because mixing a bit of pink with sky blue could cover a lot of the sky areas that are common in retail photos. Most unusual.

Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1829
Re: Unexpected result after converting image profile to softproof profile
« Reply #68 on: March 24, 2018, 06:40:27 pm »

I highly recommend that anyone doing color management and softproofing view all three parts.  The presentation is meticulous in approach, with surprisingly good examples.

Yep. Best color management videos on the web. I've said this before and it remains true today. And I'm a nitpicker.
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1829
Re: Unexpected result after converting image profile to softproof profile
« Reply #69 on: March 24, 2018, 07:03:26 pm »

This might be of interest. To get an idea how close the different printers were to sRGB skipping the profile conversion for in gamut colors I picked 6 profiles randomly from the DryCreek database which are all close enough to sRGB that most non-picky consumers wouldn't notice unless they had prints made in different places and looked at them side by side. Even then, most aren't far off.

Norit-CA-Yucaipa-SwansonPhoto-Gloss_09-11-03.icc
Costco-CA-Alhambra-Photo Satin: 12-Feb-2018
Costco-CA-Carlsbad-Gls (Noritsu 3411): 04-Apr-2016
Costco-CA-Danville-Gls (Noritsu 3411): 19-Oct-2015
Fuji_Frontier5-sRGB_CA-HD_V3a
Isgo-Fuji-390-Lustre.icc

Assuming these profiles accurately describe their machines, I created "soft proof" images of the prints based on processing them straight as sRGB. This would be the same as tagging them with the printer profile and soft proofing in Photoshop using "Preserve Color Numbers." The image used was a standard KodakPhotoDisc in sRGB These were then converted colorimetrically to AdobeRGB. This was done because some of the printed colors are expanded and go outside sRGB and I didn't want to clip them.

The images are arranged on layers in a Tiff file so you can load it in Photoshop and see the differences that would occur from prints with each printer by just selecting the highest layer. The base layer is the original.

It's attached as a zip file.
Logged

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Unexpected result after converting image profile to softproof profile
« Reply #70 on: March 24, 2018, 11:26:33 pm »

Rough data chunks?

What color space did that Fuji Frontier print the color patches to build the profile? Was it "Printer Space" or "sRGB"?

This needs to be known if that profile is to be useful for soft proofing. And the Costco printer interface software is different than the Fuji Frontier DLxxx at say another Costco or a Walmart or Walgreens. A photo lab tech at Walgreens couldn't find the tabbed interface to locate selecting one or the other space. I had to get the store manager to find it and set it and her husband ran the local one hour photo now out of business.

And Walmart's photo lab tech said he couldn't find it either and he was moonlighting at Walmart to his primary job as a radiologist at our local hospital.

This not being able to choose the space the profile was built from may be what the OP is experiencing when he said the PRINT looked horrible...

Quote
While I did this with two files last week, and they were printed by Costco and came out fine after specifying "No Automatic Corrections," or whatever the actual term is, I repeated what I thought was the same procedure this week, with horrible results, looking way oversaturated and as if I somehow had double profiled them.

The question is, if I softproof an image in LR using the Costco profile, and adjust it to best resemble the original, and then convert the image to the Costco profile, is this actually causing a problem such as double profiling?
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 15547
    • http://www.digitaldog.net/
Re: Unexpected result after converting image profile to softproof profile
« Reply #71 on: March 24, 2018, 11:33:09 pm »

What color space did that Fuji Frontier print the color patches to build the profile? Was it "Printer Space" or "sRGB"?.
Fuji Frontier RGB color space. Download a profile, you'll see this fact! It WAS NOT sRGB. Or perhaps you can explain how neutering the color space to that and producing a profile could result in a LARGER color gamut. You can't as the question illustrates you need to study a bit more how profiles that define color spaces are created.
Quote
This needs to be known if that profile is to be useful for soft proofing.
That's wrong too.
Quote
And Walmart's photo lab tech said he couldn't find it either and he was moonlighting at Walmart to his primary job as a radiologist at our local hospital.
Better I guess than someone moonlighting as a poster here discussing color incorrectly (again).
Quote

This not being able to choose the space the profile was built from may be what the OP is experiencing when he said the PRINT looked horrible...
Nope:

“There are two ways to be fooled. One is to believe what isn't true; the other is to refuse to believe what is true.”
Søren Kierkegaard
« Last Edit: March 24, 2018, 11:54:06 pm by andrewrodney »
Logged
Andrew Rodney
Author “Color Management for Photographers"

Stephen Ray

  • Full Member
  • ***
  • Offline Offline
  • Posts: 173
Re: Unexpected result after converting image profile to softproof profile
« Reply #72 on: March 25, 2018, 04:04:08 am »

What color space did that Fuji Frontier print the color patches to build the profile? Was it "Printer Space" or "sRGB"? This needs to be known if that profile is to be useful for soft proofing.

Actually, either one is just as useful as the other providing the production file(s) is imaged the same way as the targets. No?

This not being able to choose the space the profile was built from may be what the OP is experiencing when he said the PRINT looked horrible...

Tim, reference Reply #2 post by the OP and maybe my post #54. The OP did not say the print looked horrible. He never had a print made. He was referring to the preview looked horrible.
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12510
    • http://www.markdsegal.com
Re: Unexpected result after converting image profile to softproof profile
« Reply #73 on: March 25, 2018, 09:22:45 am »


“There are two ways to be fooled. One is to believe what isn't true; the other is to refuse to believe what is true.”
Søren Kierkegaard

This is a classic! So true in all walks of life.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12510
    • http://www.markdsegal.com
Re: Unexpected result after converting image profile to softproof profile
« Reply #74 on: March 25, 2018, 09:46:44 am »

What color space did that Fuji Frontier print the color patches to build the profile? Was it "Printer Space" or "sRGB"?

This needs to be known if that profile is to be useful for soft proofing. And the Costco printer interface software is different than the Fuji Frontier DLxxx at say another Costco or a Walmart or Walgreens. A photo lab tech at Walgreens couldn't find the tabbed interface to locate selecting one or the other space. I had to get the store manager to find it and set it and her husband ran the local one hour photo now out of business.

And Walmart's photo lab tech said he couldn't find it either and he was moonlighting at Walmart to his primary job as a radiologist at our local hospital.

This not being able to choose the space the profile was built from may be what the OP is experiencing when he said the PRINT looked horrible...

Tim, there are several problems with this advice.

Firstly, on building profiles: let's go back to what a profile is supposed to do. It's main task is to characterize how a printer lays down colours in response to the patch values sent to the printer. It needs to do that without the distorting interference of any application managed colour (i.e. only the printer's own colour rendition matters without the original colour data being bent or limited by sRGB, aRGB or whatever). That is why we have applications that send unmanaged data to the printer. The operating system's colour management module needs this printer characterization data for doing the math that will send the printer the colour values needed to get the expected results from it. This is a shorthand way of saying a lot that you can further unpack by reading a book such as Fraser/Murphy/Bunting "Real World Color Management", or Andrew Rodney's "Color Management for Photographers" - both excellent resources.

Secondly, on consistency between machines and labs: the whole point of ICC colour management is to achieve inter-device consistency of results. If the profiles are built correctly as suggested just above and the kind of paper being used is the same or very close (especially as between gloss and matte), if you use the profile bespoke to each printer and the printers are maintained to spec, the outcomes should match pretty closely whatever the location of the lab, the name of the company or the location of the outlet doing the printing.

If you are trying to suggest that some of these photo printing departments aren't profiling their equipment properly or maintaining their equipment to do those two things above mentioned, you have a point - it is a possibility. But I think not necessarily the first place I would go looking for causes of the problems being discussed here, because on the whole, these companies do have adequate colour management resources and advice at their disposal - they need to in order to keep the customers happy, as the customers do not need to be imaging experts to know that blue skies should be blue and green grass green. Most memory colours, especially skin tones and the like, are well known to most people and it's easy to tell whether the prints are in the ball-park. They won't be masterpieces of printing accuracy, but I've been very favourably impressed with the quality of what a Costco photo printing department can produce, on more than one occasion. So I think the causes of the issues here are elsewhere.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 15547
    • http://www.digitaldog.net/
Re: Unexpected result after converting image profile to softproof profile
« Reply #75 on: March 25, 2018, 10:35:58 am »

Tim, there are several problems with this advice.
+1

I'll attempt to make this even easier for Tim to grasp.

We have three 'pure' color numbers (computers know only numbers) that represent red, green and blue: R255/G0/B0, R0/G255/B0 and lastly R0/G0/B255. There is NO color space, just three triplets of numbers. Three color patches are output on any printer. Each is measured and we end up with three Lab values (again numbers but numbers that define colors in a color space based on how the standard observer is said to see color). Until the numbers are output and measured, there is no color space. Just device values.

The numbers may represent sRGB if and when the output device can create sRGB. NO printer can do so nor does so. Some emissive devices may. We know this when we measure such devices and the measurements confirm that indeed, the numbers (along with other attributes of numbers like a gamma encoding or TRC, a white point) are specified.

What color space did that Fuji Frontier print the color patches to build the profile? Was it "Printer Space" or "sRGB"?
Understand now how the question leads others to suggest you buy Fraser and Murphy's book, read it then ask questions rather than try to answer them first?
Logged
Andrew Rodney
Author “Color Management for Photographers"

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Unexpected result after converting image profile to softproof profile
« Reply #76 on: March 25, 2018, 06:33:41 pm »

Tim, there are several problems with this advice.

If you are trying to suggest that some of these photo printing departments aren't profiling their equipment properly or maintaining their equipment to do those two things above mentioned, you have a point - it is a possibility. But I think not necessarily the first place I would go looking for causes of the problems being discussed here, because on the whole, these companies do have adequate colour management resources and advice at their disposal - they need to in order to keep the customers happy, as the customers do not need to be imaging experts to know that blue skies should be blue and green grass green. Most memory colours, especially skin tones and the like, are well known to most people and it's easy to tell whether the prints are in the ball-park. They won't be masterpieces of printing accuracy, but I've been very favourably impressed with the quality of what a Costco photo printing department can produce, on more than one occasion. So I think the causes of the issues here are elsewhere.
Didn't say that. I'll repeat for emphasis.

I know for a fact there are two settings set within the driver interface of Fuji Frontier dry lab inkjet printers THAT ARE USED AT THESE BIG BOX STORES including Walgreens and Walmart. Don't know about Costco. And I'm going to have to assume from what I've seen of these two printer driver interfaces at these stores, the differences may also be influenced by their Fuji model DLxxx. Walmart's driver is set in sRGB and the photo lab techs can't change it.

The Walgreen's tech can switch to "Printer Space" or "sRGB" space within the driver tabbed menu (very different driver interface from Walmart's). I did a print test printing an sRGB encoded image in both spaces (having the Walgreens manager set it while I watched her select each in the Fuji driver interface) and the level of saturation between the two prints was VERY PRONOUNCED meaning "Printer Space" has a much larger gamut than "sRGB". The European ICC printer profile whose gamut size is shown in the Colorsync Utility screengrab I had to convert the sRGB image to and print in "Printer Space" fixed the saturation issue.

Because these Fuji dry lab printers have varying gamut shapes according to old vs newer models (newer ones having a much larger gamut) it could be that a Costco photo lab tech may have created a profile by choosing one or the other color spaces and didn't choose the correct color space that the profile was built from.

The OP needs to clarify whether his HORRIBLE RESULTS was from the print or in a Lightroom soft proof using the Costco profile.
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 15547
    • http://www.digitaldog.net/
Re: Unexpected result after converting image profile to softproof profile
« Reply #77 on: March 25, 2018, 06:43:17 pm »

The Walgreen's tech can switch to "Printer Space" or "sRGB" space within the driver tabbed menu (very different driver interface from Walmart's).
Still confused it appears. The printer cannot produce sRGB; impossible. The printer can assume files FOR OUTPUT are in sRGB. The printer, depending on the front end can be fully color managed and accept something other than sRGB. The printer can convert from sRGB to an output color space or from something other than sRGB to the output color space. The output color space absolutely is not sRGB! You don't appear to understand the fundamental basics of color management; the difference between an input color space and an output color space. Or what the tech was actually doing. Hopefully Stephen can attempt to set you straight if possible; Mark and I have not been at all successful. Which isn't surprising but a bit sad.  :-[
Logged
Andrew Rodney
Author “Color Management for Photographers"

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 15547
    • http://www.digitaldog.net/
Re: Unexpected result after converting image profile to softproof profile
« Reply #78 on: March 25, 2018, 06:49:42 pm »

The OP needs to clarify whether his HORRIBLE RESULTS was from the print or in a Lightroom soft proof using the Costco profile.
Only for those who are not paying attention! He wrote:  I just uploaded the file to my wife's "album" and clicked on the image to view it.  It was horrible.
IF you believe this to imply it's the print, it is based on truthful hyperbole, massive assumptions and a lack of reading multiple posts here to be kind.
Stephen pointed this out many posts (#55) ago but you apparently missed the facts provided again: " It was horrible." Yes, because you're looking at an untagged file or the viewing software is not color managed.

"You must stick to your conviction, but be ready to abandon your assumptions". -Denis Waitley
« Last Edit: March 25, 2018, 06:52:58 pm by andrewrodney »
Logged
Andrew Rodney
Author “Color Management for Photographers"

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Unexpected result after converting image profile to softproof profile
« Reply #79 on: March 25, 2018, 08:57:04 pm »

In further reply to Mark this 2013 Photo.net thread I started concerning what the Fuji Frontier's color spaces do to sRGB data will show the saturation differences and how it will affect an ICC profile build...

https://www.photo.net/discuss/threads/how-are-you-all-liking-your-local-1-hour-photo-fuji-frontier-dry-labs.482851/

And here's the two prints I just shot for those who may have trouble with the PN page.

Logged
Pages: 1 2 3 [4] 5 6 7   Go Up