June 2013 - colorsync-users

Re: Colorimetric Accuracy in the Field
by Rich Wagner 05 Jun '13

05 Jun '13

On Tue, 04 Jun 2013 13:18:48 -0400 Iliah Borg wrote: >>> 3. Use dcpTool. (http://dcptool.sourceforge.net/Introduction.html) This command-line driven software can decompile binary format DNG profiles into XMP, compile XMP files into new compiled DNG profiles, make a profile "invariate" (an invariate profile won’t cause changes in tint when you make adjustments to exposure settings), and untwist a profile. (An invariate profile won’t change color when you adjust exposure settings, but still has hue twists embedded within it. Untwisting a profile removes all hue twists completely. See http://dcptool.sourceforge.net/Hue Twists.html) > > > With dcpTool you can reduce the "DNG profile" (DNG Camera Profile, DCP) to a simple matrix (ColorMatrix), however in ACR/LR white balance is still applied using ColorMatrix interpolation or extrapolation. Unless ad hock dcp is calculated for the light in the scene the result of interpolation/extrapolation is unpredictable in terms of colour. Iliah, Of course, DNG profiles can be created from a single illuminant with both the ColorChecker Passport software and the DNG Profile Editor, where both matrices end up identical. dcpTool can do a lot more than just reduce a DNG profile to a simple color matrix. TheHueSatDelta tables in DNG profiles have third dimension, value, that in the past was almost always set to 1, but that Adobe now uses to implement the “hue twists” in their new generation profiles. There is also a LookTable, which is exactly the same as a HueSatDelta Table, but is intended to be applied later in the processing pipe, “after any exposure compensation and/or fill light stages, but before any tone curve stage” that can also apply hue twists. dcpTool can make a profile Invariate. Making a profile ivariate shifts any hue twists to before the basic exposure adjustments stage, so they are applied immediately. Therefore an invariate profile won’t cause changes in tint when you make adjustments to exposure settings. This is a way to modify the standard Adobe profiles if you want to keep the twists, but just prevent colors from changing when you adjust Carmera Raw or Lightroom’s exposure controls. Making a profile ivariate moves any LookTable to the HueSatDelta table. Specifically, if a LookTable exists, and no HueSatDelta tables exist, dcpTool simply copies the LookTable to the HueSatDelta tables, and the removes the original LookTable. If both a LookTable and HueSatDelta tables exist, the tables are mathematically combined, and the combination becomes the new HueSatDelta table. Again, the original LookTable is then removed. If no LookTable exists, then the profile is already invariate. dcpTool can also Untwist a profile entirely. Untwisting a profile removes all hue twists completely. An untwisted profile will always keep colors true, regardless of original in-camera exposure, or changes to exposure controls in Camera Raw or Lightroom. Untwisting a profile removes all intensity (V in HSV) dependencies in both the LookTable and any HueSatDelta tables. Effectively, the twist at a single point in the twist table is chosen, and applied to all intensities. The twist chosen is that corresponding to a skin tone that is accurately exposed, thus preserving the profile's skin tone. I'm particularly looking for feedback from people who have been using this tool, as appropriately modified DNG profiles would appear to effectively deal with many of the complaints about LR/ACR and colorimetric accuracy. Thanks, --Rich Wagner

3 4

Re: Colorimetric Accuracy in the Field
by Rich Wagner 05 Jun '13

05 Jun '13

On Wed, 05 Jun 2013 10:04:22 -0700, Ben Goren <ben(a)trumpetpower.com> wrote: > On Jun 5, 2013, at 9:55 AM, Andrew Rodney <andrew(a)digitaldog.net> wrote: > >> Bruce Fraser's Lab values from the target are based on Lab values supplied by Chris Murphy** which he states in the piece and those values differ a bit from Lindbloom. > > But do those values match those of the actual ColorChecker Bruce Fraser photographed? > > As I wrote, the quality and consistency of the ColorChecker is commendable, but it *IS* a mass-produced item and differences of a few DE aren't at all uncommon. And those differences can easily become significant with the age and condition of the ColorChecker. > > I'll bet you a nice dinner at a restaurant of your choice in the Phoenix metropolitan area that the N9.5 patch on Bruce Fraser's personal ColorChecker is at least a few DE from the L*=96.14 a*=0.00 b*=0.00 he assumed it was. > > b& This issue is easy to deal with. If you're concerned about the accuracy of the reference measurements, you can replace the built-in ColorChecker reference measurements in the ColorChecker Passport software with spectral measurements of your own CC Passport. Just make a copy of the (Mac) app and save it as a backup, open the application package (option-click), look for the Reference folder. There's a file inside called ColorChecker24_spectral.txt. Replace it with your own spectral measurements from your Passport, close and save. You unfortunately can't do this with the DNG Profile Editor as the measurements are contained in compiled code. --Rich

1 0

Re: Publishing vs Curating
by Glenn Andrews 05 Jun '13

05 Jun '13

Stanley, That is in absolutely fantastic story, and there is a good deal of truth to the observation that contemporary viewers are so used to seeing clean and bright color that anything else tends to look like a mistake. It has happened to me, and with originals, not reproductions. A few years ago I was visiting the Ansel Adams collection at LACMA. A little voice in my head kept saying, "Gee, shouldn't he bump up the contrast a little?" Clearly my eye had been completely corrupted by all the commercial advertising campaigns I had worked on over the years where there is no such thing as too much. I ad to take a moment to clear my head of the countless bight shiny images that advertisers had used to get my attention over the years and recalibrate to a different viewing experience. On the subject of colorimetric accuracy in digital photography, I tend to agree that this does not..cannot...have the same meaning in creative scene photography as it would in archival reproduction or static product photography. After all, creative photographers are striving for a "look", not a reproduction. For these uses, I would suggest that the proper way to think about a camera profile would be similar to the way photographers traditionally related to known film types. Kodachrome, Ektachrome, and all the rest were characterized by the manufacturers relative to known and controlled lighting and development conditions..and this was simply the starting point once photographers went into the field and started shooting under completely variable lighting, exposure and development conditions. If a profile is understood and used in the same way that knowledge of film charactistics was used for lifetimes, I think that the knotty issues would tend to dissolve away to a large degree. Glenn Andrews, Certified G7 Expert Color Clarity 818 886-1037 O 818 263-7789 M 818 975-5239 F glenncolorguy(a)yahoo.com www.colorclarity.net http://colorclarity.blogspot.com ------------------------------ Message: 2 Date: Tue, 04 Jun 2013 15:48:56 -0700 From: Stanley Smith <ssmith(a)getty.edu> To: José Ángel Bueno García <jbueno61(a)gmail.com> Cc: Jeffrey Stevensen <jeffstev(a)maine.rr.com>, colorsync-users <colorsync-users(a)lists.apple.com> Subject: Re: Colorimetric Accuracy in the Field Message-ID: <51AE0C68020000B60012453E(a)mail.getty.edu> Content-Type: text/plain; charset="utf-8" Well, the obvious goal is dE2000 =0.0. -- a goal that is equally obviously not attainable. Sure, we reproduce our paintings in books, on displays at a press check, and projected on unpredictable machines for lectures. I would gently suggest that given the truly bad color memory we all have, that if the color presented on any of these platforms is not "accurate" to the degree that everyone here seems to be striving for, it won't make a damn bit of difference to the viewer. These differences are really only noticeable when you compare the reproduction with the original in a controlled setting, and I never see visitors to the Getty walking around the gallery with the catalog open-- eyes darting from the book to the painting (I'm somewhat embarrassed to note that I do this all the time, and it is often a sobering experience). Of course we need to be close, and we do strive for that, but we have no control over the viewing conditions of a person leafing through one of our catalogs-- a fact that renders extreme color accuracy standards moot. One more thing-- Just because curators strive for accuracy in reproduction, it doesn't mean that publishing professionals always do. I've told this story more than once: We had a big Paul Strand show a few years back, and I did what I mentioned above-- walked around the gallery with the book open. Strand printed very dark and flat-- he loved the murky print. Our reproductions were shockingly different, but looked much like other reproductions of his work that I had seen in the past-- normal contrast and more open shadows. When I asked our publisher why there was such a difference the answer was very revealing: They couldn't sell any books if they had faithfully reproduced the original prints-- it would have looked like a mistake. In Strand's defense, I believe he was used to throwing a lot of light on a print when evaluating-- not a lighting condition that is often reproduced by a casual viewer. Stanley Smith J. Paul Getty Museum Stanley Smith Head of Collection Information and Access J. Paul Getty Museum 1200 Getty Center Drive, Suite 1000 Los Angeles, CA 90049-1687 (310) 440-7286

1 0

Re: Colorimetric Accuracy in the Field
by Rich Wagner 04 Jun '13

04 Jun '13

Ben, I agree with many others that this has been an interesting and informative discussion. You have mentioned your dissatisfaction with DNG profiles in terms of accuracy. There are currently (at least) three ways to make custom DNG profiles. 1. Use the ColorChecker Passport software. This builds new profiles de novo (without requiring a "base profile"), but gives zero control over the TRC or anything else. (You can edit the app package and replace the CC spectral measurements with your own, but that's about it.) 2. Use the Adobe DNG Profile Editor. This software requires the use of a DNG "base profile" and it stacks adjustments on top of that in what is called a "recipe" file (in XMP), which can then be incorporated into a new (compiled) DNG profile. There is more control than with the ColorChecker software, but it is visually oriented, rather than "by the numbers." It is good for making "looks" but not colorimetrically accurate results. 3. Use dcpTool. (http://dcptool.sourceforge.net/Introduction.html) This command-line driven software can decompile binary format DNG profiles into XMP, compile XMP files into new compiled DNG profiles, make a profile "invariate" (an invariate profile won’t cause changes in tint when you make adjustments to exposure settings), and untwist a profile. (An invariate profile won’t change color when you adjust exposure settings, but still has hue twists embedded within it. Untwisting a profile removes all hue twists completely. See http://dcptool.sourceforge.net/Hue%20Twists.html) Do you have any experience with the latter? Is so, I'd be interested to hear what your experiences are. The same is true for anyone else out there, of course. Thanks, --Rich Wagner

2 1

Re: ICC "print-centric"
by Chris Cox 03 Jun '13

03 Jun '13

While the ICC does have more representatives from printing than other areas - they are very much interested in displays, cameras, cinema projectors, spectral rendering, and other areas. There is a lot of active research and work going on with the ICC and related standards groups that you may not be aware of... Hollywood is not innovating, so much as catching up with the ICC. Chris On 6/3/13 6:52 AM, "colorsync-users-request(a)lists.apple.com" <colorsync-users-request(a)lists.apple.com> wrote: > The ICC has become a club for the printing industry, they are not > interested in displays except for proofing, not interested in cameras, and > not interested in consumers.

3 2

RPP raw photo processor 64
by Millers' Photography L.L.C. 03 Jun '13

03 Jun '13

Before digitalizing client's original art, the camera/lens/sensor/lighting has been profiled. The NEC PA271w has been profiled. I do the exercise with gray card for both exposure and white balance. Image files look fine on the NEC PA271w display. Therefore when opening dng image files in PS CS5's RAW, the display looks very much like the original art on the NEC PA271w!. I just don't know how to open an image file in RPP64 and successfully have a good image on the NEC PA271w. Oh, the Epson Pro 9900 is also custom profiled. But my real trouble and concern is how to use RPP 64 as simply as with PS CS5 raw. All suggestions are invited. Needed! ! ! David B Miller, Pharm. D. member Millers' Photography L.L.C. dba Spinnaker Photo Imaging Center Bellingham, WA www.spinnakerphotoimagingcenter.com 360 739 2826

13 59

Re: Question about tone remapping with Relative Colorimetric
by Rich Wagner 02 Jun '13

02 Jun '13

On Jun 1, 2013, at 6:07 PM, Terence Wyse <wyseconsul(a)mac.com> wrote: >> and as Scott eluded > > ALLUDED.....grammar was never one of my strong soots. :-) Oh, I thought he eluted... that would have explained the washed-out color. ;-)) --Rich

1 0

Re: Question about tone remapping with Relative Colorimetric Rendering
by Louis Dina 02 Jun '13

02 Jun '13

Scott/Terry: Thanks for the answers. I figured converting to Relative Colorimetric with BPC turned OFF for a given paper would give me the numbers of what I'd get on paper, but it doesn't. I suspect the Info Palette shows the converted numbers that are being passed along to the printer driver. Then the driver does its thing. Absolute Colorimetric does the job and gives me the numbers I was looking for. It does keep all the in-gamut tones the same and forces the OOG tones to the border...thanks. Just to help me understand better, I printed the 0-255 RGB 11-step wedge to a matte fine art paper using first Perceptual, then Relative Colorimetric with BPC. This paper has a Dmax of about 15L* and a paper white of 97L*. The RC print keeps the midtone at the original file value, but scales everything else toward the middle from both ends. I had always read that RC intent kept all the in-gamut colors and tones unchanged, but apparently not, unless I am missing something. Abs Colorimetric keeps in-gamut tones where they are. The Perceptual print scales everything and lightens the print overall. That makes sense, since we lose a lot more values at the dark end than we do at the light end, so dark values move more. Appreciate the help. Lou > ------------------------------ > > Message: 8 > Date: Sat, 01 Jun 2013 18:07:56 -0400 > From: Terence Wyse <wyseconsul(a)mac.com> > To: Louis Dina <lou(a)loudina.com> > Cc: Colorsync Users List <colorsync-users(a)lists.apple.com> > Subject: Re: Question about tone remapping with Relative Colorimetric > Rendering > Message-ID: <9FB5C7AA-1D83-4ED7-9D80-B0B5BC7597A0(a)mac.com> > Content-Type: text/plain; charset=us-ascii > > > On Jun 1, 2013, at 10:35 AM, Louis Dina <lou(a)loudina.com> wrote: > > > RC Intent has always > > been described as accurately reproducing in-gamut colors, then clipping > all > > out-of-gamut colors and tones, bringing them back to the edge of the > > destination gamut. > > All bets are off when you turn on black point compensation....BPC scales > the black point of the source to the black point of the destination instead > of clipping.....it also has a subtle effect on colors as well. > > If you want *accuracy*, use absolute colorimetric.....and as Scott eluded, > have your color settings prefs set to absolute as well if you want a > accurate readout in the info palette. > > Terry > > >

1 0

Question about tone remapping with Relative Colorimetric Rendering
by Louis Dina 01 Jun '13

01 Jun '13

I'm trying to understand how tones are remapped when going from file to print, using an RGB driver. I'll use an example so my question is clear. I created a black to white gradient in Photoshop, Adobe RGB. 11 step posterization. I purposely chose to keep color out of this test, since I wanted to keep it simple to understand how the luminance values get mapped. That gives RGB numbers, from dark to light, of 0 25 51 76 102 128 153 178 204 229 255. The info palette shows L* values of 0 6 20 32 43 54 64 73 83 91 100. I *converted* the file to a custom RGB profile (using the epson printer driver), Relative Colorimetric, BPC on, of a weak Dmax matte fine art paper. This paper achieves about 18L* for Dmax and has a white of about 96L*. I chose a weak paper on purpose to amplify the effect. The info palette shows changes in the two darkest patches ONLY. Black goes from L* of 0 to 2, and the next lighter patch goes from 6 to 8. All other patches remain the same. The RGB numbers change too reflecting profile tables to adjust brightness and neutrality. I believe these are the new RGB numbers that will be sent to the printer driver, correct? They obviously aren't the achieved values in the print, because the paper can achieve only 18L* black and 96L* white. Is there a way to know what the final L* values in the print will be using Photoshop, or in looking at the profile? If so, how? I assume this information must be in the profile. (I built this printer profile using i1Profiler). OK, switch to the original RGB numbers (before conversion) and go to Proof Preview instead of Convert to Profile. If I proof preview using RC intent, BPC turned On, the file display shows the two darkest patches brightening up (same as the conversion). However, since I am not converting, the numbers in the info palette are the original numbers. Still in Proof Preview, RC, BPC On, if I turn on Simulate Black Ink, the entire tonal scale lightens up, most notably, from blacks to the midtones, but even a little bit in the 1/4 tones. Obviously, the profile knows this, but it doesn't show me the numbers. If I then turn on Simulate Paper White, the display reflects the dingy yellow-white of the paper (96L/0a/3b or so). Now, the light end of the tonal scale is scaled back, giving a more accurate view of the real tonal range of the final print. Unfortunately, the numbers in the info palette still show the original numbers. Is there a way to see the final numbers I would read in the print if using a spectro (assuming a perfect profile, of course)? Sorry for the long email, but I want it to be clear. RC Intent has always been described as accurately reproducing in-gamut colors, then clipping all out-of-gamut colors and tones, bringing them back to the edge of the destination gamut. Obviously, the whitest white this paper can produce is 96L/0a/3b, and the blackest black is 18L*/0a/0b. Here's the real question...Will all the grayscale tones in my step wedge from 18L to 96L reproduce accurately on the final paper using RC Intent? Or will they be scaled somehow? The Proof Preview shows them ALL changing. And will my two black patches (original numbers of 0L and 6L) both end up being 18L? Again, the proof preview says No. I'd like to believe the midtones will stay locked, so 50L in my info palette will read 50L in the print, but I am not convinced this is the case. Thanks for any responses, and sorry again about the length of this email. I hope it was clear. Lou Dina

3 3

Re: Courses and training - need suggestions or recommendations.
by David Wollmann 01 Jun '13

01 Jun '13

I have seen some changes but don't recall which pages. I use this book for a college course, 2nd edition. At one point one of the students had purchased the first edition by mistake. I can't really say if the changes were significant or not. It did throw off that student when I was referencing certain pages during a lecture one week of class. Here is a link for an eBook & PDF version; http://www.peachpit.com/store/product.aspx?isbn=0321267222 Chris Murphy would have more information if he is still on the list. David Wollmann > Date: Fri, 31 May 2013 09:57:29 -0400 > From: "Pylant, Brian" > >>> This book will help. >>> Real World Color Management (2nd Edition) [Paperback] >> >> Does anyone know if the second edition contains significant new material? I >> still have the first edition... > > Anyone?

3 2