Pages: [1]   Go Down

Author Topic: Keeping gamma L* when sending images to a print lab with gamma settings at 2.2?  (Read 1068 times)

PhilippeRouquet

  • Newbie
  • *
  • Offline Offline
  • Posts: 9

My Eizo monitor CG277 is calibrated with your BasICColor Display software at prepress settings (5000k, 120cd and L* gamma). I work in Capture One Pro with either the embedded camera profile or ProStarRGB (L* gamma-based )as working color space. I process images with an ambiant Solux desktop lamp at 5000k.
I would like to send my images to this recommended print lab, Saal Digital, which uses identical monitor settings except for the gamma (http://www.saal-digital.co.uk/support/article/monitor-calibration/#sb).

My question is as follows:

For the soft-proofing stage with their ICC profiles, should I create another monitor profile with gamma 2.2 or can I stick to L* gamma? I know L* and 2.2 are very close but should I care at this stage?

When sending final jpg or tiff images with an embedded RGB color space for printing in their lab, should that color space be based on L* such as respectively L*-RGB and ProStarRGB to match my working gamma or should I convert to sRGB and ProPhoto RGB 2.2 or Adobe RGB to match their gamma (2.2)?

If I have to convert to sRGB or ProPhoto RGB 2.2 or Adobe RGB, should I have my monitor calibrated at gamma 2.2?

Many thanks in advance for your suggestions.
Philippe
Logged

BradFunkhouser

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 52
    • My Articles


For the soft-proofing stage with their ICC profiles, should I create another monitor profile with gamma 2.2 or can I stick to L* gamma? I know L* and 2.2 are very close but should I care at this stage?


It's my understanding that the calibration of a monitor's gamma is primarily going to affect how non-color managed applications show colors.  With color managed applications, the monitor's profile, built using the calibration, should provide accurate mappings back and forth between the monitor's RGB color space and Profile Connection Space (the device independent space that models human color perception).  The mappings in the profiles should in effect adjust to whichever calibration gamma you used, so theoretically it seems like you shouldn't see a difference in your soft proofing if you built a new display profile after calibrating with a slightly different gamma.

If the gamma changed significantly, then it would affect the gamut boundaries of the monitor's color space, which would in turn affect which of your soft proofing colors were in-gamut versus out-of-gamut for the different monitor profiles.  But even in this case, soft proofing colors that are in-gamut for both monitor profiles should theoretically still display the same.
« Last Edit: March 10, 2017, 12:08:51 PM by BradFunkhouser »
Logged

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2091

My Eizo monitor CG277 is calibrated with your BasICColor Display software at prepress settings (5000k, 120cd and L* gamma).

How different does your display look when calibrated to 2.2 gamma? Does it make a 21 step grayramp's shadows abruptly roll off into black or does it make the shadows look flat with milky looking blacks or what looks a bit like shadow flare. Perceptual Rendering Intent setting soft proofing treats shadows similarly. Log-C in cinema treats shadows similarly.

Since you're working in a color managed system it doesn't matter the color space gamma the image is edited and saved in since the preview is showing you what you should be seeing. As long as the printer renders according to a color managed workflow and honors the embedded 2.2 gamma of AdobeRGB or 1.8 of ProPhotoRGB then you should get WYSIWYG.

If in doubt do a test print.
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 591

The colorspace gamma of an image makes no difference when printing and it only makes a difference on a monitor if an application is not color managed.

Said another way, you can convert back and forth from gamma=1 to a different gamma and the displayed image, as well as the print of an image in a color managed app will not change.

However, the working space gamma does make a difference when you manipulate an image. Changing tone curves, hue and such in Photoshop will produce different results when the gamma's differ.

When printing an image from an RGB working space it is first converted to the XYZ colorspace, which is linear (gamma=1).  From there it is converted to ICC Lab, which is a clipped subset of CIELAB. ICC Lab is the native space (profile color space, aka, PCS) that is used by printer profiles. It is then rendered according to the selected Intent such as Relative or Perceptual. That is determined by the LUTs in the printer's device profile.
« Last Edit: March 10, 2017, 01:52:21 PM by Doug Gray »
Logged

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2091

Doug, I think the OP is referring to L* gamma as a type of non-standard display calibration "look" in how it displays a more gradual distribution of tone from black to white that looks more like Perceptual rendering especially in the shadow areas. IOW the shadows or a bit lifted and the highlights slightly darkened to bring out more definition viewing a 21step grayramp.

I don't know if the display profile contains a description of this different rate of tone distribution (contrast) in order to adjust for color managed previews editing in a 2.2 gamma working space.

At least this is how I understand what he's asking. I really don't understand the use of L* gamma in an already color managed standard workflow. Everybody's display should look the same or else what's the point of color managed previews.
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 591

Doug, I think the OP is referring to L* gamma as a type of non-standard display calibration "look" in how it displays a more gradual distribution of tone from black to white that looks more like Perceptual rendering especially in the shadow areas. IOW the shadows or a bit lifted and the highlights slightly darkened to bring out more definition viewing a 21step grayramp.

I don't know if the display profile contains a description of this different rate of tone distribution (contrast) in order to adjust for color managed previews editing in a 2.2 gamma working space.

At least this is how I understand what he's asking. I really don't understand the use of L* gamma in an already color managed standard workflow. Everybody's display should look the same or else what's the point of color managed previews.

I use a fairly similar setup to the OP with D50 but 100 nits instead of 120 also on an Eizo.  Eizo's can be profiled to different gammas as well as L*.  L*, which has a long, linear ramp to a gamma=3 will provide slightly less banding in a dark gray gradient when using 8 bits but the look of an image is otherwise not any different. I use a 10 bit path so I don't see any difference unless I go to gamma=1 when just a touch of banding shows up at the lowest luminance.
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 591

The gamma setting can make a big difference in how the various tools interact with colors. For instance here's a red/green adjacent patch set where the blur tool has been run back and forth over the patches. The image in sRGB (gamma approx. 2.2) develops a dark region where the red/green intersect while the same image in gamma=1 has a smooth transition through orange. The image was converted back to standard sRGB after blurring.


« Last Edit: March 11, 2017, 12:33:31 AM by Doug Gray »
Logged

PhilippeRouquet

  • Newbie
  • *
  • Offline Offline
  • Posts: 9

Many thanks to all contributors. The discussion got so technical, I had a bit of trouble to follow.
Basically, I do not have to change anything in a color managed workflow with an adequately calibrated monitor.
I've just received a reply from Saal Digital UK (German print lab with subsidiaries across Europe):
"For the monitor calibration it can leave the L* gamma setting, which is not much different to the 2.2 gamma.If you could sends us pictures or screenshots with the embedded sRGB or Adobe RGB profile that would be perfect."

I also found a page where they clearly indicate that they don't support ECI color spaces (L* based):
http://www.saal-digital.co.uk/support/article/srgb-adobe-rgb-1998-and-prophoto-rgb/#sb
They also indicate that they also support ProPhoto RGB (1.8) but unfortunately only in 8-bit tiff files.

I think I will stick to Adobe RGB in 8-bit tiff files for the moment.

In a later post, I will share my experience of using Saal Digital for fine art printing.

If anybody knows a good lab that accepts 16-bit tiff files in either ProStar RGB (ECI) or ProPhoto RGB, please let me know.
Best to all.
Philippe


« Last Edit: March 16, 2017, 06:25:40 AM by PhilippeRouquet »
Logged

Pictus

  • Full Member
  • ***
  • Offline Offline
  • Posts: 206
    • Retouching

The gamma setting can make a big difference in how the various tools interact with colors. For instance here's a red/green adjacent patch set where the blur tool has been run back and forth over the patches. The image in sRGB (gamma approx. 2.2) develops a dark region where the red/green intersect while the same image in gamma=1 has a smooth transition through orange. The image was converted back to standard sRGB after blurring.

Photoshop in gamma 1 also change the gradients

Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 591

Photoshop in gamma 1 also change the gradients



Nice demos of the blending impact of gamma. Interesting that the default blending setting is gamma=1 while the way colors are blended in filters like the blur tool and assorted blurring/resizing etc. are done, incorrectly, in the working space gamma.

Ideally, Adobe should provide an option to use gamma=1 instead of the working space gamma for those as well.
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 591

Many thanks to all contributors. The discussion got so technical, I had a bit of trouble to follow.
Basically, I do not have to change anything in a color managed workflow with an adequately calibrated monitor.
This is absolutely correct.
Quote
I've just received a reply from Saal Digital UK (German print lab with subsidiaries across Europe):
"For the monitor calibration it can leave the L* gamma setting, which is not much different to the 2.2 gamma.If you could sends us pictures or screenshots with the embedded sRGB or Adobe RGB profile that would be perfect."

I also found a page where they clearly indicate that they don't support ECI color spaces (L* based):
http://www.saal-digital.co.uk/support/article/srgb-adobe-rgb-1998-and-prophoto-rgb/#sb
They also indicate that they also support ProPhoto RGB (1.8) but unfortunately only in 8-bit tiff files.

I think I will stick to Adobe RGB in 8-bit tiff files for the moment.
Adobe RGB (1998) is a perfectly good space if you don't have colors that exceed it and most photographs don't. You can easily check by viewing a proof in Adobe RGB and selecting the out of gamut mask. Any colors that are out of gamut by even a tiny fraction will show the gray mask. It's actually overly critical and even colors in gamut but right on the Adobe RGB boundary will be masked as well. Proofing against a printer profile is different and you have to be out of gamut by about 6 dE before it gets masked.
Quote

In a later post, I will share my experience of using Saal Digital for fine art printing.

If anybody knows a good lab that accepts 16-bit tiff files in either ProStar RGB (ECI) or ProPhoto RGB, please let me know.
Best to all.
Philippe

Don't sweat the 16 bit issue. If you have colors that are outside Adobe RGB then use ProPhoto 8 bit tifs for printing if the vendor allows it. Few printers even purport to offer 16 bit printing to the device and it makes no observable difference printing photos. However, do use lossless tifs since jpgs are not only 8 bit but uses lossy compression so you don't even get true 8 bits.
Logged
Pages: [1]   Go Up