June 2019 - colorsync-users

RE: Exposure Value
by graxx＠videotron.ca 24 Jun '19

24 Jun '19

"Styrofoam cup", Ben? Would you have a link to some kind of image, anywhere? To help "visualize" what you mean by that? Ilah's web site is fascinating 😊 / Roger -----Original Message----- From: Ben Goren <ben(a)trumpetpower.com> Sent: Monday, June 24, 2019 1:16 PM To: Robin Myers <robin(a)rmimaging.com> Cc: Roger Breton <graxx(a)videotron.ca>; 'colorsync-users?lists.apple.com' List <colorsync-users(a)lists.apple.com> Subject: Re: Exposure Value On Jun 22, 2019, at 10:31 AM, Robin Myers via colorsync-users <colorsync-users(a)lists.apple.com> wrote: > Perfect reflecting diffuser 100 > Fluorilon FWS-99 2009 99.9 > Spectralon SCS-99 99.7 > Kodak Gray Card 1977 (3 specimen average) 95.0 > ColorChecker Passport 2013-12 Neutral 9.5 52.2 > ColorChecker Passport 2013-12 Neutral 5.0 95.2 Spectralon is great stuff, but damned expensive. Expect to pay as much for a Spectralon target as a Canon L series lens. There’re two homebrew-style options that are typically plenty “good enough” in photographic settings. The first, if you want 99%+ reflectance, as close to “pure white” as you’re going to get without spending lots of money: PFTE (Teflon) thread seal tape. Not very durable, and it’s got basically an eggshell surface that’s a bit more specular than ideal (Spectralon is close to ideally Lambertian)...but it’s dirt cheap and you can get it anywhere. Use a bit of creativity and you can figure out how to wrap several layers around a card to make a target. If it gets dirty or damaged or lost...no biggie; all you’ve invested is your time, so just make another. The second is a styrofoam coffee cup (or packing material, etc.) Its total reflectance is a lot less...in the 80% range. But it’s much more spectrally flat than the photographic tools (ColorChecker, gray card, etc.). Outdoors or in mixed lighting, its near-cylindrical shape lets you sample the illuminant from all angles, which can be creatively useful. And, with the right combination of lens and cup, you can fit the cup over the lens and get a near-ideal uniform sampling of all the light in the scene — much like those hundreds-of-dollars custom white balance tools that fit over the lenses do, but, again, with near-perfect spectral uniformity. (Put the cup over the lens, take a picture, and use it for your in-camera custom white balance. It’s probably pretty close to where you want to meter, too; experiment some and you could get that dialed in perfectly and consistently.) Coffee cups are pretty reliable spectrally; the more creative shapes you might find in a craft store (spheres, cubes, etc.) are more hit-or-miss. Visually compare with some thread seal tape (if you don’t have a spectrometer) if you’re unsure. Last note: read everything Iliah has on the Raw Photo Processor / RawDigger Web site and accompanying blog. You couldn’t ask for a better crash course on camera exposure, metering, development, etc., etc., etc. Cheers, b&

1 0

Re: Printer profile issue
by Don Hutcheson 24 Jun '19

24 Jun '19

Randy, Can you be more specific? How are you determining the dE? What target did you use? What pre-linearization did you use, G7 or something else? How did you set your ink limits? Profile editing is a desperate last resort these days, that just sweeps the dirt under the rug. Much better to solve the source problem. Don ************************** Don Hutcheson President HutchColor, LLC 908-500-0341 don(a)hutchcolor.com ************************** > Date: Sat, 22 Jun 2019 18:35:36 +0000 (UTC) > From: "zbet66(a)yahoo.com" <zbet66(a)yahoo.com> > To: colorsync-users(a)lists.apple.com > Subject: Printer Profile Issue > Message-ID: <852842756.464303.1561228536392(a)mail.yahoo.com> > Content-Type: text/plain; CHARSET=US-ASCII > > I've tried both Onyx and Xrite i1 Profiler to profile an Epson solvent printer. The profile comes out fine as far as tone and gray. The Delta E's for Reds and Magentas are over six. > I also tried more sampling with the Isis spectro. The results are still the same with either profiling application. > The only solution I can come up with is to purchase a profile editor. The EFI editor is a bit expensive and comes with a profile maker which I do not need. > Does anyone have another possible solution to this? > Randy ZauchaManaged Color >

1 0

Epson Doubleweight Matte
by graxx＠videotron.ca 24 Jun '19

24 Jun '19

Finally got hold of a fresh sample of Epson Doubleweight Matte paper. Here is a link to its measurements, in Excel. https://1drv.ms/x/s!AkD78CVR1NBqkLVN8r7LgHiGqxRYZg In M1 mode, I believe it's going to be sufficient to carry my humble color separations experiments : CIE Lab D50/2 (KonicaMinolta FD-7) 93.86 2.53 -8.01 Thank you all for your kind and patient help / Roger P.S. "Funny", this paper appears to have a slightly "reddish" tint. Of course, it does not appear "reddish" to me, because of chromatic adaptation, no matter what lighting I observe it under.

2 1

RE: Fogra 51 and 52
by graxx＠videotron.ca 24 Jun '19

24 Jun '19

Dear Rafik, At some point, I will buy the idea that "some" data had to be brought in to build the Fogra51 profile, and it only makes sense that the data was the result of averaging a number of measurements from popular M1 instruments such as the ones you (Hanno) enumerated below : > "The original M1 measurement data for FOGRA51 (6 test prints, 6 > printers, 4 papers; all measured by 4 devices - i1pro2, fd7, exact, > spectrodens)..." This development reminds me of the way the original SWOP TR-001:1995 dataset was arrived at, out of an average of 12 sheets, 4 measured by X-Rite, 4 measured by Kodak (not 100% sure?) and 4 measured by Gretag (SPM-100). It was an average of "selected sheets", ones that were closest in terms of SWOP's target solid ink densities and dot gain on Champion TextWeb #5 coated paper. Now, is Fogra51 a "straight" averaging of M1 measurements? It's possible that, through averaging, some of the "outliers" fall into place and the resulting data becomes "smoother", mathematically. It's very possible. In my own practice and those of many users I know, averaging a number of sheets helps catch the "centroid" of the system behavior. But it's possible and likely that the data was also smoothed on top of being averaged. *And* what I find more likely is that the averaged and smoothed data was also "edited" synthetically, using advanced tools, we, mere mortal color users don't have access to, in order to "align" with existing data for historical or other considerations. Yet, I gather that the data was extensively tested "visually" between participating proofing vendors and printers, and that "everyone" was satisfied with the results. (I got the points you raised on your reply on the ECI List and will go through them) I confess this is all speculations on my part as I did not participate in Fogra51 (I wish...) I have a link to a PDF which dokuments Fogra51 development but it is in German... (which makes me regret not having taking more than the first German language course during my undergraduate studies -- I did take three courses in Italian, though...) / Roger -----Original Message----- From: Refik Telhan <rtelhan(a)icloud.com> Sent: Sunday, June 9, 2019 10:30 PM To: colorsync-users(a)lists.apple.com Cc: graxx(a)videotron.ca Subject: Re: Fogra 51 and 52 Dear Roger, The following quotation is from Mr. Hanno Hoffstadt's mail dated 19.02.2019 which he had sent as part of the thread "[ECI-EN] Issues related with the switch from FOGRA39/47 to FOGRA51/52": "The original M1 measurement data for FOGRA51 (6 test prints, 6 printers, 4 papers; all measured by 4 devices - i1pro2, fd7, exact, spectrodens)..." Best, Refik -------------------------------------------------------- Refik Telhan, EE B.Sc. Light and Color Management Consultancy Aydogdu Sokak 12A, Tarabya Mahallesi Sariyer, 34457, Istanbul, Turkey Mobile: + (90) (532) 426 21 87 -------------------------------------------------------- On 09.06.2019 00:32, "colorsync-users on behalf of Roger Breton via colorsync-users" <colorsync-users-bounces+rtelhan=icloud.com(a)lists.apple.com on behalf of colorsync-users(a)lists.apple.com> wrote: Do we know what instrumentation was used to measure Fogra 51 and 52? Just "M1"? Best / Roger _______________________________________________ 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/rtelhan%40icloud.com This email sent to rtelhan(a)icloud.com

4 11

RE: Printer Profile Issues
by graxx＠videotron.ca 23 Jun '19

23 Jun '19

46L, 57a and -9b is way too restricted for solid Magenta? GRACoL2006 Magenta is 46 74 -2. The a* value needs to be pushed a whole lot more. / Roger -----Original Message----- From: colorsync-users <colorsync-users-bounces+graxx=videotron.ca(a)lists.apple.com> On Behalf Of zbet66--- via colorsync-users Sent: Sunday, June 23, 2019 2:32 PM To: colorsync-users(a)lists.apple.com Subject: Printer Profile Issues The printer is a ten color Epson Surecolor 80600 so I am sure it has enough gamut. I run a custom 24 patch color check with Colorthink spreadsheet and I have matched these colors on printers with smaller gamuts like color copiers. The Magenta is 46L, 57a and -9b. The ink restrictions were done by measuring the Onyx target so you are correct, I should look there. Also, I will plot the problem colors in Chromix to see if they indeed do fall within the gamut of the resulting profile. The problem could well be in the ink restrictions!Thanks, I will report back. Randy Zaucha _______________________________________________ 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/graxx%40videotron.ca This email sent to graxx(a)videotron.ca

1 0

Printer Profile Issues
by zbet66＠yahoo.com 23 Jun '19

23 Jun '19

The printer is a ten color Epson Surecolor 80600 so I am sure it has enough gamut. I run a custom 24 patch color check with Colorthink spreadsheet and I have matched these colors on printers with smaller gamuts like color copiers. The Magenta is 46L, 57a and -9b. The ink restrictions were done by measuring the Onyx target so you are correct, I should look there. Also, I will plot the problem colors in Chromix to see if they indeed do fall within the gamut of the resulting profile. The problem could well be in the ink restrictions!Thanks, I will report back. Randy Zaucha

1 0

Printer Profile Issue
by zbet66＠yahoo.com 23 Jun '19

23 Jun '19

I've tried both Onyx and Xrite i1 Profiler to profile an Epson solvent printer. The profile comes out fine as far as tone and gray. The Delta E's for Reds and Magentas are over six. I also tried more sampling with the Isis spectro. The results are still the same with either profiling application. The only solution I can come up with is to purchase a profile editor. The EFI editor is a bit expensive and comes with a profile maker which I do not need. Does anyone have another possible solution to this? Randy ZauchaManaged Color

2 1

RE: Exposure Value -- Follow Up
by graxx＠videotron.ca 23 Jun '19

23 Jun '19

I had the chance to take a few more pictures today with my camera, under direct sunlight. I fear I have egg on my face for I could not find much wrong doing by my camera this time ☹ I need to further test with "live" subjects. But I remain suspicious... This is a link to my analysis. https://1drv.ms/b/s!AkD78CVR1NBqkLVMSaCo3oELVlKP0A Thank you for your patience and kind help. / Roger Breton www.graxx.ca

2 1

RE: Printer Profile Issue
by graxx＠videotron.ca 22 Jun '19

22 Jun '19

If you care to send the characterization data my way, I could create you a profile using ColorToolbox. Have you tried Argyll? I would try to avoid editing the profile first. The other thing I would look at is the "gamut" of the device in its characterized state, it's possible that the Reds and Magentas are not "saturated" enough for the colors you want to match? I ran into a similar issue, lately, and as soon as I loosened the ink restriction on Yellow and Magenta, all of a sudden my problem was solved. Best / Roger -----Original Message----- From: colorsync-users <colorsync-users-bounces+graxx=videotron.ca(a)lists.apple.com> On Behalf Of zbet66--- via colorsync-users Sent: Saturday, June 22, 2019 2:36 PM To: colorsync-users(a)lists.apple.com Subject: Printer Profile Issue I've tried both Onyx and Xrite i1 Profiler to profile an Epson solvent printer. The profile comes out fine as far as tone and gray. The Delta E's for Reds and Magentas are over six. I also tried more sampling with the Isis spectro. The results are still the same with either profiling application. The only solution I can come up with is to purchase a profile editor. The EFI editor is a bit expensive and comes with a profile maker which I do not need. Does anyone have another possible solution to this? Randy ZauchaManaged Color _______________________________________________ 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/graxx%40videotron.ca This email sent to graxx(a)videotron.ca

2 1

RE: Exposure Value
by graxx＠videotron.ca 22 Jun '19

22 Jun '19

Thank you so much Robin for taking the time to explain this new SNI concept. Is this something that can simply be evaluated using an i1Pro? With Spectrashop? I’ll head to your web site next to pursue the links you kindly provided. I would not have suspected that the N9.5 patch on the ColorChecker chart to be less “spectrally neutral” than the N8 or N5 patches. I have the Munsell Grays swatch books in both Matte and Glossy finish, it will be interesting to see what kind of SNI these are. I also have some “gray” NCS swatches… I guess it would also work on my calibrated BCRA tiles? Best / Roger From: Robin Myers <robin(a)rmimaging.com> Sent: Saturday, June 22, 2019 1:31 PM To: Roger Breton <graxx(a)videotron.ca> Cc: 'colorsync-users?lists.apple.com' List <colorsync-users(a)lists.apple.com> Subject: Re: Exposure Value Hello Mr. Breton, When you are looking for a new gray card, may I suggest that the SNI value calculated in SpectraShop may be of use. The SNI is the Spectral Neutrality Index, a value which I have developed to evaluate neutral references. This value is an indicator of a specimen's spectral uniformity. The scale is 0-100, with 100 being a spectrum that is monotonic in value. Here are some example SNI values. Perfect reflecting diffuser 100 Fluorilon FWS-99 2009 99.9 Spectralon SCS-99 99.7 Kodak Gray Card 1977 (3 specimen average) 95.0 ColorChecker Passport 2013-12 Neutral 9.5 52.2 ColorChecker Passport 2013-12 Neutral 5.0 95.2 The SNI formula is based solely on the spectral values, no perception is involved. This allows for material evaluation alone. Notice that the white patch (Neutral 9.5) of the CC Passport has a much lower SNI than the Neutral 5 patch from the same chart. This is due to the high absorbance of violet wavelengths by the titanium dioxide colorant used for the white and also the slightly yellow color of the same titanium dioxide. This can be readily observed by placing a Fluorilon or Spectralon 99% reflecting tile (scintered PTFE) next to any white material containing titanium dioxide (TiO2). The TiO2 material will appear slightly yellow since the human vision system will adapt to the tile’s white, allowing the true color of the TiO2 to be observed. The SNI value was created specifically to answer questions about which material is more spectrally neutral. I find it very useful for checking gray reference materials such as gray cards. These articles on my website may also be of use. Neutral References for Digital Camera Calibration http://chromaxion.com/information/neutral_references.html Gray or White Card for Neutral Balancing? http://chromaxion.com/information/gray_or_white.html Regards, Robin Myers Robin Myers Imaging robin(a)rmimaging.com <mailto:robin@rmimaging.com> robin(a)chromaxion.com <mailto:robin@chromaxion.com> 925-519-4122 On 21 Jun 2019, at 19:14, Roger Breton via colorsync-users <colorsync-users(a)lists.apple.com <mailto:colorsync-users@lists.apple.com> > wrote: Dear Ilah, Thank you so much for applying your vast technical expertise to my humble problem. I realize, again, that there is so much to learn in digital photography, I feel ashamed to have neglected digging into this fascinating part of image processing for so long (It took me more than a few CPU cycles of my brain to figure Exposure Value, today)... For sure, I need to invest in a "new" gray card, I confess that this one has seen its share of "tales from the trenches". Having said that, I'm relieved to read in your analysis that my camera "light meter" isn't playing tricks on me or need to go back to the manufacturer under warranty. I *will* reshoot, for sure, as you suggested. It's an exercise worth every effort and probably investigate using one of your suggested software.. For now, I am not sure what to conclude from my experience? Tomorrow, I will probably experiment with shooting A) RAW alone and shooting B) RAW + JPEG. In "theory", there should not be any difference between the two modes of shooting but if there should be one, then this is going to prove interesting. Have you had a chance to take a look at my Excel sheet at all? To see whether I correctly applied the metering equations from measured Luminance and Illuminance? I think the calculations are sound. One last thing I have not tried and I *will* try is, to shoot with my antique D100, in RAW, to compare with shooting the same scene with the D810, in RAW, to test any potential differences between the two systems, in terms of exposure. One thing I wonder, as I was thinking about the experimental setup, something completely absurd (?), was, what that, could there be any difference in the underlying capture of the pixels between the two cameras? Suppose, on the D810, at 1/200s, F/5.6, ISO400, the histogram looks skewed to the left, suggesting some degree of under-exposure (see my earlier links)? Suppose further, on the D100, at the same 1/200s, F/5.6, ISO400 settings, the histogram looks "normal", with pixels distributed throughout the tonal range, from 0 to 255, then what? Until I'm convinced that there is nothing wrong with this camera, I figure I need to further test it, until I'm fully satisfied that there is nothing "wrong" with it. Best regards / Roger -----Original Message—— <snipped for brevity>

1 0