September 2014 - colorsync-users

Banding issues with jpegs
by Brian Aderer 13 Sep '14

13 Sep '14

Hi all Not sure if this went thru or not, sorry if its a repeat Im having a recurring issue with converting tiffs to jpegs thru photoshop. My workflow is: tango 48bit hdr scan -》sf hdr converter, downsample to 24 bits and export into j holmes d cam 3 space -》 pshop edit save as tiff -》downsample and export to srgb jpeg thru save for web. Looks fine on my calibrated monitors, 2.2 gamma set a little warm and a little flat and dark for editing for electronic output. On some monitors, especially mobile devices, Im getting severe banding in some images especially highlights such as the attached images. Can anyone suggest a solution? https://db.tt/qkfGaJb9 https://db.tt/VC4uiszr https://db.tt/VC4uiszr Thanks Brian

3 2

Using iProfiler When Data Set Has Been "White Point" Adjusted
by William Whitfield 12 Sep '14

12 Sep '14

Greetings, I have been testing the SubstrateRelCalc_v6_DRAFT4.xls file, and another excel file that does about the same thing, to adjust my white point. I then paste the new XYZ and Lab numbers into my original measurement file replacing the spectral data. However when I try to use this file with iProfiler it states this is an invalid measurement file. Before iProfiler I used Profilemaker to successfully create a profile predicting a curved target from just a linear run using an adjusted measurement file containing only lab numbers. Is there a way to use a lab / XYZ file in iProfiler, or is my only option to use Profilemaker? I only own Profilemaker and iProfiler. -Bill-

3 2

Re: Colorsync-users Digest, Vol 11, Issue 133
by Remington, David 11 Sep '14

11 Sep '14

I was referring to LR and PS on the same display. 100% view sounds familiar. Not very useful with large images. You are right I could use an instrument but DCM does work for this type of thing for a quick and simple comparison. The actual accuracy or lack of is less important than confining that what should be the same is not or vice versa. David Remington Manager, Digital Imaging and Photography Services Harvard Library Preservation D-70A Widener Library Cambridge, MA 02138 p 617-495-4701 f 617-495-0403 w http://imaging.harvard.edu/ On Sep 10, 2014, at 3:00 PM, colorsync-users-request(a)lists.apple.com wrote: > Message: 3 > Date: Wed, 10 Sep 2014 08:18:15 -0600 > From: Andrew Rodney <andrew(a)digitaldog.net> > To: "colorsync-users(a)lists.apple.com" > <colorsync-users(a)lists.apple.com> > Subject: Re: Colorsync-users Digest, Vol 11, Issue 132 > Message-ID: <A769430C-0CA0-4BC9-ACBF-C6849EBC3114(a)digitaldog.net> > Content-Type: text/plain; charset=us-ascii > > On Sep 10, 2014, at 6:21 AM, Remington, David <david_remington(a)harvard.edu> wrote: > >> Just the other day I was using Lightroom 5 on a dual monitor set-up and wondering why the image displayed on the second screen in develop mode is darker than it is when opened in photoshop > > If the display profiles defining the calbiration are not identical (and that's unlikely) wouldn't that throw the DigitalColor Meter analysis off? Also, in LR, you need to view the data in Develop at 1:1 (100%) while viewing the image that way in Photoshop. On the same display, they should appear the same. On differing displays, I suspect the calibration would be a huge factor here. Lastly, why not use a instrument? Presumably you used one to build the profile for both displays. > > I just don't as yet see anything that the DigitalColor Meter provides that is useful for color managed workflows. But I'm open to suggestions. > > Andrew Rodney > http://www.digitaldog.net/

2 1

Color accurate workflows (was in search of a D50 Editing colorspace)
by Peter Miles 11 Sep '14

11 Sep '14

of my previous post I wrote: > > Conclusions > ... absolute colorimetric from an RGB color space will work, using Adobe color management… > 1: Only if the source RGB space has a D50 white point. > I have been reflecting on the implications of the above as it pertains to the workflow from a profiled RGB flatbed scanner and moving color meaning from scanner colorspace to editing colorspace, or from scanner color space to printer colorspace. Going to the trouble of profiling a scanner using It8 target is a great idea. All going well, potentially the colors now encoded in the scans in my 'custom' profiled scanner space, I might well be near the original. I can not tell by looking at my monitor because the colors that are displayed on my monitor have been scaled the to my monitors white point. But none the less, potentially, there they are encoded in my custom scanner color space inside of photoshop. However unless the scanner space white point happens to be D50, any attempt to move the absolute color meaning of pixels out of the scanner colorspace using Adobe will fail. Because Adobe will first invoke a relative colorimetric transform from the scanner space to the profile connection space (D50) before it moves it to the destination colorspace. And in that process the absolute color meaning of the original scanner pixels will be lost. in the same way it gets lost when it's being scaled to my monitors white point. The implications of this to folk who attempt to maintain the integrity of digitally encoded colors of art works as they work on seem to me to be significant. At least to me anyway Regards Peter Miles

8 33

Photoshop Color Settings - Conversion Options
by Louis Dina 10 Sep '14

10 Sep '14

I have been through this before a number of years ago, but want to be sure I have this clear in my mind. I took a wide gamut RGB target with very saturated colors and brought it into Photoshop. It was untagged, so I tagged it with ProPhoto RGB to give me the widest possible gamut and the most saturated Lab numbers. Then, I converted it in PS to a high gamut RGB printer profile (glossy RC paper on an Epson 3880). I used Relative Colorimetric with BPC for all the following conversions. Before converting from ProPhoto RGB to the destination printer profile, I altered my Conversion Options in the PS Color Settings dialog box. 1. The first conversion from ProPhoto RGB was done with Color Settings Conversion Options set to Abs Colorimetric and BPC turned Off. 2. The second conversion from ProPhoto RGB was done with Color Settings Conversion Options set to Rel Colorimetric and BPC turned Off. 3. The third time, my conversion from ProPhoto RGB was done with Color Settings Conversion Options set to Rel Colorimetric with BPC turned On. The actual conversion was done from Edit > Convert to Profile > Rel Col and BPC. In each case, the Info Palette shows that the RGB numbers are identical, regardless of Color Settings Conversion Options. However, the Lab readout in the Info Palette is different for each conversion. So, if conversion options is set to AbsCol w/o BPC, I am assuming the Lab readout is showing what the actual Lab numbers would in the dried print if I read the print with my spectro (assuming a perfect conversion, of course). Is that correct? Assuming that is correct, I'm a little less clear on why I get different Lab numbers when Color Settings have RelCol Conversion Options, one with BPC turn On and one with BPC turned Off. In both these cases, white shows up 100L/0a/0b, which I feel confident is the color of paper white (which is less than 100/0/0). What surprised me, however, was that Black was different, one reading 7L/0a/0b, and the other reading 1L/0a/0b. I would have thought they'd be the same. I could use some clarification. Also, which Conversion Option makes the most sense to leave as the default setting in Color Settings? Thanks, Lou Dina

3 2

Re: Colorsync-users Digest, Vol 11, Issue 132
by Remington, David 10 Sep '14

10 Sep '14

Displaying the values sent to the monitor is exactly what makes the digital color meter useful for trouble shooting and finding problems and inconstancies in “color managed” applications. If color management is working correctly in an application or between applications the same values will be sent to the screen from different input sources and from the same source through different applications. e.g. comparing how a web browser and photoshop display a tagged image. Just the other day I was using Lightroom 5 on a dual monitor set-up and wondering why the image displayed on the second screen in develop mode is darker than it is when opened in photoshop. Just my eyes playing tricks on me?? NO! Displayed values are lower as confirmed through digital color meter. Now as to why it does that??? David Remington Manager, Digital Imaging and Photography Services Harvard Library Preservation D-70A Widener Library Cambridge, MA 02138 p 617-495-4701 f 617-495-0403 w http://imaging.harvard.edu/ On Sep 9, 2014, at 11:23 PM, colorsync-users-request(a)lists.apple.com wrote: > > Message: 3 > Date: Tue, 09 Sep 2014 14:29:20 -0700 > From: Ben Goren <ben(a)trumpetpower.com> > To: Andrew Rodney <andrew(a)digitaldog.net> > Cc: ColorSync List <colorsync-users(a)lists.apple.com> > Subject: Re: Color accurate workflows (was in search of a D50 Editing > colorspace) > Message-ID: <C9A663B8-129E-4B18-A4C1-B6C724D9FB45(a)trumpetpower.com> > Content-Type: text/plain; charset="utf-8" > > On Sep 9, 2014, at 2:20 PM, Andrew Rodney <andrew(a)digitaldog.net> wrote: > >> On Sep 9, 2014, at 3:08 PM, Peter Miles <P.Miles(a)massey.ac.nz> wrote: >> >>> 'Classic Color Meter' is by Ricci Adams >>> http://www.ricciadams.com/projects/classic-color-meter >>> Its a tool that I find to be a very useful addition to Photoshop's eye dropper tool. >>> I use it set to display color in Lab mostly. >> >> But unless I'm misunderstandings, it's not doing things differently than the Apple product which will not necessarily agree with a color managed app as Chris suggests. > > For me, the tool's only utility is for applications which lack their own numerical color readout. It's better than nothing and not all that bad, but it definitely has its limits. > > For example, the ACR profile editor, last I worked with it, lacked a color readout. The least-worst ACR profiles I was able to create I made with the use of the Adams Classic Color Meter. > > I can't imagine any advantage it might offer over the many different ways Photoshop offers to provide similar information. > > b& >

3 2

Paper Question
by Mike Stewart 09 Sep '14

09 Sep '14

Hello: I will try and simplify this. Let's say I have 20 different press sheet papers and I am measuring the *L*a*b values of the paper in M0 mode to see if they fit under Gracol2006, SWOP3c2, etc. Is there a rule of thumb as to the allowed variance; not a DeltaE variance; but a plus/minus variance? One paper may be L: 95, a: 0, b: -2 (Gracol2006) another may be L: 92 a: 1, b: -1 (could be SWOP3v2 or Gracol2006) So; if Gracol is L: 95, a: -0.03, b: -1.95 can I look at something as simple as L: 95 +/- 2, a: -0.03 +/1, b: --1.95 +/- 1 Thanks - Mike Stewart

3 3

In search of a D50 Editing colorspace
by Peter Miles 08 Sep '14

08 Sep '14

Lars Borg wrote.. Or please show me a case where the above doesn't hold true. It's just that on occasions I'm calculating RGB values for the AdobeRGB space from a spectral measurement from a real object using Spectrashop. For example the artist I previously mentioned has wanted in the past to print the colors of a colorchecker chart with an absolute match. If I calculate the RGB values for AdobeRGB with the observer illuminant as D65, then for the neutral samples I get approximately R=G=B. However when I print it absolute colorimetric from photoshop on our print set-up these neutrals comes out neural relative to D50 on the print. I.e is no longer an absolute match the observer illuminant I asked for. So I figured that it would be much easier for me to get absolute matches to the observer illuminant I ask for if my editing colorsapces started out as D50 in the first place. Regards Peter ________________________________ From: Lars Borg <borg(a)adobe.com> To: Colorsync List <colorsync-users(a)lists.apple.com> Sent: Thursday, 4 September 2014, 6:34 Subject: Re: In search of a D50 Editing colorspace A key point that many seem to have missed: With ICC profiles, ALL editing spaces are effectively D50, including sRGB. If you examine for example the sRGB profile, you'll find that it converts device values to D50 colorimetry. You'll also find that the chromaticities in the XYZ tags don't match the sRGB spec, as they've been chromatically adapted to D50. It has a non-D50 white point tag, but this is not used for relative colorimetric. It's very useful for finding the original D65 chromaticities, though. So in the search of a properly sized workspace profile, no need to limit your search to a D50 white point. _______________________________________________ Do not post admin requests to the list. They will be ignored. Colorsync-users mailing list (Colorsync-users(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/colorsync-users/peter_milesnz%40yah… This email sent to peter_milesnz(a)yahoo.co.nz

10 26

Re: rendering intents
by David 08 Sep '14

08 Sep '14

RPP 64 offers the beta but does not offer Profoto RGB. Cheers, david. Sent from my Verizon Wireless 4G LTE smartphone -------- Original message -------- From: Ben Goren <ben(a)trumpetpower.com> Date:09/08/2014 1:33 PM (GMT-08:00) To: Spinnaker Photo Imaging Center <spinnakerphotoimagingcenter(a)dnmillerphoto.com> Cc: ColorSync <colorsync-users(a)lists.apple.com> Subject: Re: rendering intents On Sep 8, 2014, at 12:58 PM, Spinnaker Photo Imaging Center <spinnakerphotoimagingcenter(a)dnmillerphoto.com> wrote: > Is printer space the custom .icc profile I create for my Epson 9900 and a given substrate? Yes. > And, > where would I use the printer space in my workflow? After all other processing is done, when you're ready to print and / or soft-proof. Save the finished image as a TIFF still in your working space, and then use Argyll to convert it to your printer space. > I start with sending CR2 image files directly to RPP 64, Raw Photo Processor 64. Then use Lab TIFF 16-bit and open in either Lightroom 4, or > Photoshop CS5. ProPhoto RGB becomes my RGB color working space in PS. LR is Melissa RGB. > > Does this make sense? Yes and no. LR might ProPhoto the lease insensible working space available to you. BetaRGB is, for the most part, still the best working space to use otherwise. RegardlessRPP 64, be sure to save the image with your desired working space directly from RPP rather than use Photoshop to convert it from something else. For example, if you're using ProPhoto, save it as ProPhoto directly from RPP and keep the embedded profile in Photoshop. > I mention RPP 64 since it uses Argyll, if I am correct. I don't believe that's the case, but I could be worng. It certainly doesn't call Argyll as a separate set of executables (that I'm aware of), though it might incorporate some of Graeme's code. Iliah may be able to comment further if it matters. Cheers, b&

2 1

Re: rendering intents (was: In search of a D50...)
by Rich Wagner 08 Sep '14

08 Sep '14

This is exactly the point I tried to make in the bug report. The user *is* being lied to. The UI creates confusion, because in these circumstances users are expecting the app to be doing exactly what it says, and the app is intentionally doing something else without informing them. There are plenty of instances where the UI prevents someone from doing something that is not appropriate or possible. Assign a CMYK profile to an RGB image? Convert a 16-bit image in sRGB to indexed color? Or an 8-bit Lab image to indexed color? Take a 32-bit image and Save for Web as a JPG? The list goes on and on, and I fail to see the difference in this case. I still feel that if a rendering intent is not appropriate, it should be greyed out. The astute user will ask why and will learn something. The less astute user will just do what is allowed, and will realize that certain transformations are not allowed in certain circumstances. I fail to see the “confusion” that is created. To the contrary, it prevents users from thinking that certain transformations are allowed when they are not, “because Photoshop can do it.” —Rich Wagner On Sep 5, 2014, at 12:00 PM, colorsync-users-request(a)lists.apple.com wrote: > On Sep 5, 2014, at 11:45 AM, Steve Upton <upton(a)chromix.com> wrote: > >>> Further, some users will complain that they're being denied access to >>> rendering intents.:-) >> >> But you see, that’s the exact problem I find with users, only in reverse. They complain that they’re granted access to intents that don’t exist. > > I think this is a classic UI issue. We should not be shown a Rendering Intent we can't use. I don't see why in this context, the RI's that don't apply are not grayed out. > User's can't complain. They should not be granted access to selecting something that isn't going to do what they select, they are kind of being lied to. > > Andrew Rodney > http://www.digitaldog.net/ >

7 8