[kaseki]

Having just skimmed all of, and read much of, the DisplayCAL requirements and operation description at DisplayCAL?Open Source Display Calibration and Characterization powered by ArgyllCMS, my present view with respect to ever getting a display exactly calibrated parallels Dante's observation: "Abandon every hope, who enter here."

In other words, any expectation that one monitor's display of a photograph will exactly match what other calibrated monitors -- nearby or on the Internet -- present as the same photograph will need a healthy dose of realism. There are just too many adjustments that can be made to cope with various monitor technological limitations to expect the same photometric behavior over the entire monitor model space, color space, gray scale, and room environment.

---------- Post added 15th Feb 2019 at 18:28 ----------

With respect to the original quest, x-rite service has provided me with a link to the color data for the color checker Passport. This is actually accessible among all the downloads that the service tab on the color checker gets to. It is D50 based. I need to figure out what scheme I'm going to use to build an actual image of these colors. The image of the passport that correlates the color block number designations with the tabulated data does not seem to have the true colors, but I was toying with Gimp so there is no guarantee that I measured correctly. I have asked if there is a D65 version.

[Class A]

But, I will note that even I didn't trust it to work, especially when I got a new monitor, which ended up calibrating to a green tint when dealing with the white point.

I'm glad I'm not the only one.

My solution in that case was to leave the monitor at its default and then let the calibrator work without the adjustment.

It sounds like you went through calibration and profiling but chose the "native white point" option, i.e., opted out of targeting a specific white point such as D65.

I did that as well, but additionally subsequently manually adjusted the monitor's white point (using a service menu) so that it visually matched the white point of another monitor that I was able to calibrate without getting a tint. Ideally, I would have organised a perfect light source and a colour chart to have an even better white point reference.

After I completed the calibration, I basically printed a photo through 2 or 3 online printers and checked the images vs. what I saw on the monitor (just a visual confirmation).

That's obviously not as systematic as what kaseki is attempting but as long as you can use your monitor to get the prints you want, it seems that the native white point of your monitor was in pretty good shape. Having said that, there are many displays that are way too blue out of the box. So it normally isn't a good idea to trust the manufacturer with respect to the white point. Having a claim of "factory calibration" would be very nice to have in order to trust the white point at least to some extent.

[kaseki]

...With respect to the original quest, x-rite service has provided me with a link to the color data for the color checker Passport. This is actually accessible among all the downloads that the service tab on the color checker gets to. It is D50 based. I need to figure out what scheme I'm going to use to build an actual image of these colors. The image of the passport that correlates the color block number designations with the tabulated data does not seem to have the true colors, but I was toying with Gimp so there is no guarantee that I measured correctly. I have asked if there is a D65 version.

OK, I have examined the image of the Passport Color Checker's colors in x-rite's web site and found them relatively distant and noisy relative to x-rite's specified D50 RGB values. So I took their image and used Gimp to replace every block with a color patch that had the specified RGB values. I also grabbed the background "frame" and adjusted that to be a constant color and free of noise.

Note that the black patch in particular, specified to be 52/52/52, is in the physical passport blacker or grayer than the background frame depending on light direction. I used a color picker to determine the frame color in their image (several places with marginal agreement) and then just set it all to 42/42/42.

x-rite's web site image is a jpeg, brought into Gimp and modified yields an xcf file, and default exported yields a png file. This is included below. I can export at other file types, I expect, so let me know your druthers.

[kaseki]

JPG file exported from Gimp file:

Attached Images

 

[PDL]

@Kaseki
You are making this way to hard. Go here:
X-Rite: +Colorimetric values for ColorChecker Family of Targets

[kaseki]

@Kaseki
You are making this way to hard. Go here:
X-Rite: +Colorimetric values for ColorChecker Family of Targets

That is the data I was given by x-rite support. Note the differences between the upper image of the patch pattern at that link and the colors of the png and the jpg that I provided above built with the values given by x-rite in the table.

[PDL]

That is the data I was given by x-rite support. Note the differences between the upper image of the patch pattern at that link and the colors of the png and the jpg that I provided above built with the values given by x-rite in the table.

I don't think that the colors in the link are really the colors on the chart. If they were, then why would you buy the real chart? The page is a description of the values not a direct copy of those values.

Again, you are overthinking this.

[emalvick]

Going back to my earlier post, the white point on many monitors is not great. Calibrating a white point by setting it through my calibration software, actually made it worse through one of my monitors. But, calibrating with the default white point isn't bad. The calibrator will notice that the colors are off and adjust the profile to get a correct white point. Is this perfect or exact? no.

But you do have to ask yourself what is the end game. For my case, it was making sure that what I got on the screen matched what I got from a printer. And when I say match, I mean that when I look at the image it reasonably matches. I'm sure if I hold my photos up to the monitor I can see differences, but prior to calibrating I was off by a lot, especially in the skies and yellows, and now I am not.

The other issue with regard to ambient light is something I try not to worry too much about. I work at different times of day and the lighting in the room I work in can vary during daylight hours and evening hours. My x-rite calibrator is supposed to be able to adjust for ambient light, but I've always found it annoying, much like I find the auto-brightness settings on my phone quite annoying. I quit using it, and have never had any issues. If I have any issue now, it's that my eye glasses have the anti-glare coating that makes things a bit yellow. I don't think I can calibrate for that short of buying new glasses, but man the glasses eliminate a lot of headaches I used to get processing photos. On the plus side, I've only truly noticed a problem in images that are very light or black and white images that have a lot of lighter colors. I used to put a slight tint to my black and white images (a very light sepia type cast), but I've quit that, and now it is fine. And, I actually like the images better now anyway.

Ultimately, it comes down to how exact you want to be, to what end you are going through this, and whether it is worth the effort. Of course that logic holds for many things, so I am not knocking it. I've purchased expensive glass for the marginal returns you get on your dollar because it was worth it to me.

[Class A]

Calibrating a white point by setting it through my calibration software, actually made it worse through one of my monitors.

But, calibrating with the default white point isn't bad. The calibrator will notice that the colors are off and adjust the profile to get a correct white point.

Are you talking about the difference between making adjustments to the white point in a pre-phase of the calibration using monitor hardware controls versus skipping that phase?

In any event, either you allow the calibration software to influence the white point or you don't.

If the monitor's native white point deviates significantly from the white point target (e.g., D65) then this will require heavy adjustments.
Unless the respective calibration curves are uploaded to the monitor to be realised by the monitor's internal hardware-LUT, they will be realised by the computer's graphics card. As the latter approach is bit-depth limited (due to the limited channel between computer and monitor), heavy adjustments will result in loss of tonal values and hence in banding and/or colourisation of B&W gradients/photos.

Hence one should either select a monitor profile that produces a white point close to the target or adjust the monitor's white point using the monitor controls in the respective pre-phase of the calibration process.

One can test the quality of one's monitor calibration by using the very useful monitor test pages by Han-Kwang Nienhuys.

[emalvick]

Are you talking about the difference between making adjustments to the white point in a pre-phase of the calibration using monitor hardware controls versus skipping that phase?

In any event, either you allow the calibration software to influence the white point or you don't.

Right. I am essentially talking about skipping the adjustments to the white point pre-calibration. I think in the case of one of my monitors, the software has to make more adjustments if I try to adjust the white point prior to calibration. Honestly, it isn't a great monitor, and I don't use it for editing; I mostly throw the library over on that monitor (i.e. thumbnails for browsing).

Either way, the calibration software does make its own adjustments in the calibration phase itself, and everything pretty much comes out right in the end. For instance, an image in photoshop, straddling the two monitors will be consistent across the two. Of course, how off you are at different stages and how much calibration is needed can impact things significantly, but I've not seen problems through links like you provided and prints (which worry me most).

For my own personal needs, I have bigger problems with my eyes than the fact my monitor may be off from perfect by 0.01%.

[kaseki]

Since Mike bought an Aputure MX LED, I thought I would buy the LS-mini20c for comparison and because I knew that it could achieve 120 nits of reflected luminosity from the x-rite color checker (white squares).

the LS-mini20c "light" has these salient characteristics CRI >= 96, TLCI >= 97 and color temperature adjustable between 3200K and 6500K. Beam zoom and barn-door modification are included. Initial random impressions are as follows:

• The box containing the light and its support parts could have been designed by an origami master. It has an incredibly efficient layout providing protection to and from all of the included parts.
• The light itself provides a lot of capability in a small package. It borders on "cute."
• The zoom function and barn doors appear useful and operate as expected.
• The supplied product manual could stand some further elaboration. In particular, neither the manual or the website made it clear that the dial on the side that controls intensity and color temperature works like a rotary dimmer, except that the two push states are not variable on and off, but variable temperature and variable intensity depending on push condition. The rotary motion vs. dial change seems backwards to my expectation.
• Zoom, temperature, and intensity dials are quantized -- they jump to the next indicated value as their controlling rotary motion is changed. Zoom is mechanically continuous, intensity and temperature may be quantized in fact as well as dial readout. This is to be determined.
• While there is only one way the various parts can actually work together, a general identification of what goes with what would save a bit of time.
• It wasn't until I charged the battery that I discovered that there is a green tell-tale that illuminates when charging is complete, replacing the red when charging begins. Should be noted in the manual. Charge times should be listed.
• Battery is marked Li-ion. Best lifetime charging technique could stand to be in the manual also.
• The battery charger is intended to hang from an electrical receptacle. Given common US duplex receptacle covers and noting that the heavy battery itself is a cantilever to the electrical contacts, to assure the connection is secure the upper receptacle of a vertical duplex receptacle is the better choice to use.
• The battery took about a day to charge. Without any use for 48 hours, I started charging it again. It presented the red tell-tale and sometime after an hour but before two hours changed to green. It may take a bit of use to get the battery to settle down.

I have only tested the light as an illuminator for a brief time using the mains adapter. The appearance of my reconstruction of the x-rite color checker viewed in Darktable and in Gimp on my factory calibrated BenQ SW2700PT (operating I think in sRGB) vs. the "light" illuminated color checker revealed that the color patches "2" and "9" (brown is #1) look a tad off in the pinkish sense. (The passport is slightly redder.) I'll have to look into that further. It may just be a deficiency in the light's spectrum but I want to reconfirm the color coordinates. Also, I am not certain that my replaced monitor ICC is not being used under Linux. Much to look at here. Last, please note that the color checker patches are "painted" with a surface texture such that their bi-directional reflectivity function is not going to yield the same appearance as the patch images shown on an IPS monitor screen.

Not withstanding the complaint nits above (standard pun disclaimer), I think this light will be useful for actual small object photography as well as for playing with calibrations.