Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
02-09-2019, 11:47 AM   #1
Loyal Site Supporter
Loyal Site Supporter




Join Date: Dec 2017
Photos: Gallery
Posts: 527
Correct illumination for use with a color target

Given a calibrated IPS monitor operating at a specific color temperature, e.g. 6500K, displaying an image of a color target, e.g. x-rite Passport color checker, and a physical Passport color checker in one's hand, any attempted comparison would seem to me to require a 6500K source illuminating the Passport. Such a source would exclude tungsten based lamps and most arc lamps (except high pressure ones where the spectrum merges into a quasi black body distribution), but should include a set of LEDs that together have the same spectral output as those in the monitor that illuminate the LCD panel. The light source at some modest distance should illuminate the Passport with an illuminance such that the reflected luminous intensity from the white square is the same (nits) as the monitor is set up to generate, e.g., 120 nt.

Short of another calibrated monitor emitting white light imaged onto the Passport, does anyone know of a photographic "flashlight" or modest size and cost modeling source that might approximate these desired properties.

Thanks

02-09-2019, 02:40 PM   #2
Site Supporter
Site Supporter




Join Date: Jan 2013
Location: Hampshire, UK
Posts: 728
Would Rotolight NEO II LED come close? It's top end is 6300K, so slightly below your 6500K example, but it is variable. It's one of the many flash and continuous hybrids that are appearing that is getting my notice. You probably know the range already ...
02-09-2019, 03:04 PM   #3
Pentaxian




Join Date: Mar 2017
Location: Otago, New Zealand
Posts: 422
I'd be more inclined to go with strobe and lighting gels than led's personally, though I have a feeling that even without the colour correcting gel, setting white balance carefully and using software to compensate would get you there also.

If you are using LED's keep in mind that many of the cheaper ones have distinct spectral spikes which can skew results.
02-09-2019, 04:07 PM   #4
Pentaxian




Join Date: Jun 2008
Location: Idaho
Photos: Gallery
Posts: 773
Not sure what you're after. If you have a color target, the illumination for that target is irrelevant if it is corrected by the system which captures the image. The binary values for each target swatch should end up almost the same if that system does its job under different illumination CTs, as long as those are continuous spectral sources or very high CRE sources which emulate given color temperatures (e'g', 3200K, 4500K, or 6500K).

When that binary record is displayed on your monitor, the appearance of your color target will have the appearance of being illuminated by a 6500K source if the monitor is adjusted to that value. Likewise, the monitor could be adjusted to 4000K and the image should approximate the original target if it had been illuminated by a 4000K source.

Essentially, your camera tries to take a standardized record of a target, compensating for different color temperatures (assuming it is set accordingly or has a good auto white balance). The reproduction of that digital record then depends on your display and its CT setting. Hence you can make a scene taken in cool shadows look like a sun lighted scene with software which adjusts the red/blue ratios. This doesn't change the monitor CT value but changes the recorded values of a target.

It's more important to have your monitor adjusted properly when making a print (to the CT for which the print will be viewed) since that print could be viewed under differing lighting and may not look like what your monitor showed. But then again, this depends on how accurately your printer is. Ideally, the printer should been calibrated to your monitor for specified viewing conditions.

To make a long story short, if your target was shot with 6500K light, it should appear nearly identical when that image is displayed on a 6500K monitor (in a 6500K environment). That may be what your goal is.


Last edited by Bob 256; 02-09-2019 at 04:15 PM.
02-09-2019, 09:00 PM   #5
Loyal Site Supporter
Loyal Site Supporter




Join Date: Dec 2017
Photos: Gallery
Posts: 527
Original Poster
Thank you all for your insights. My thought was that if the monitor were calibrated using the usual technique of photometric measurement via sensor (such as x-rite makes) and cal software, it should be possible to check the result by an uncorrelated approach -- comparing using one's eyes a monitor image of a test target taken by the target's manufacturer and published for this purpose to the in-hand physical test target -- and leaving my camera and flash out of the experiment.

Barry E: I needed clues such as yours to look up. While NEO 2 may be overkill with the added flash, I may be able to extract a few search terms to help look further.

Brooke Meyer: Hauling my PC and monitor across town to the nearest Sherwin Williams might be a tad inconvenient. However, I will check them out for an understanding of just what lighting sources they use. I don't recall such a setup last time I was at my nearest SW, but maybe I overlooked it. Note that fluorescent lamps have the same issue as typical "white" (phosphor coated near-UV) LEDs, a spectrally varying output instead of a black body output. Hence the CRI /= 100.

I'll report what I find.

kas
02-10-2019, 02:45 AM - 1 Like   #6
Pentaxian
swanlefitte's Avatar

Join Date: May 2015
Location: New Orleans
Photos: Gallery
Posts: 2,245
Just thinking. What if the room is dark and you take one swatch on screen and one white then compare if the color shifts between the physical swatch illuminated by both? Shouldn't the same color bouncing off the same color not produce color shift but just intensify it?
02-10-2019, 10:35 AM   #7
Loyal Site Supporter
Loyal Site Supporter




Join Date: Dec 2017
Photos: Gallery
Posts: 527
Original Poster
QuoteOriginally posted by swanlefitte Quote
Just thinking. What if the room is dark and you take one swatch on screen and one white then compare if the color shifts between the physical swatch illuminated by both? Shouldn't the same color bouncing off the same color not produce color shift but just intensify it?
Nice idea! The only issue I can see at the moment is that if the monitor is not truly white in a large white patch, it will make the white patch on the physical test target (x-rite) being illuminated off-white and thereby look the same. A color bouncing off of a color could not be used, I believe. The white part of the screen would have to be the entire test target illumination.

See also the thread at: Suitable colour chart lighting for creation of dual-illuminant profiles? - PentaxForums.com

---------- Post added 10th Feb 2019 at 12:49 ----------

QuoteOriginally posted by Brooke Meyer Quote
...
Maybe a better idea for you is to find out what lighting is preferred in Art Museums. Those people are picky about lighting! Museum Lighting in the 21st Century
Thank you for that link. It is a bit out of date re recent developments, and not really oriented toward the 6500K area of the spectrum, but informative none-the-less. It may be of note that I used CREE architectural lights in my kitchen renovation ca 2008-2012. CREE claimed that their lights adjust themselves to retain their color temperature. At the time I checked the spectrum that they sent me for apparent adequacy, and tested a lamp against stone and tile to be sure that the important color carried through the design was retained under the CREE lighting.
02-10-2019, 03:27 PM - 2 Likes   #8
PDL
Pentaxian




Join Date: Oct 2006
Location: Woodinville, WA USA
Photos: Gallery
Posts: 1,549
The whole idea about using a color chart is to create custom color correction for a given light source.
For instance, shoot in a environment where you have mixed light sources. (I have shot in areas where there is filtered daylight, fluorescent, Halogen, Tungsten and Mercury Vapor lights in the same room at the same time)
In your RAW image converter open up the color chart, correct for exposure (histogram as spread out as possible), use the eyedropper on the white balance tool to select the middle grey square. That sets your white balance for the images shot in that environment.

You do not need a certain color temperature light source to shoot a target. You have it backwards. You use the color chart to assist with setting the color temperature. You will need additional software to create a custom icc profile for those lighting conditions.

Since you are using a ColorChecker Passport you should be following:
https://www.pentaxforums.com/forums/125-flashes-lighting-studio/382892-suita...-profiles.html
and
https://www.pentaxforums.com/forums/32-digital-processing-software-printing/...e-one-pro.html

02-10-2019, 03:36 PM   #9
Loyal Site Supporter
Loyal Site Supporter




Join Date: Dec 2017
Photos: Gallery
Posts: 527
Original Poster
QuoteOriginally posted by PDL Quote
The whole idea about using a color chart is to create custom color correction for a given light source.
For instance, shoot in a environment where you have mixed light sources. (I have shot in areas where there is filtered daylight, fluorescent, Halogen, Tungsten and Mercury Vapor lights in the same room at the same time)
In your RAW image converter open up the color chart, correct for exposure (histogram as spread out as possible), use the eyedropper on the white balance tool to select the middle grey square. That sets your white balance for the images shot in that environment.

You do not need a certain color temperature light source to shoot a target. You have it backwards. You use the color chart to assist with setting the color temperature. You will need additional software to create a custom icc profile for those lighting conditions. ...
Thank you. A very succinct explanation that I hope to internalize somewhere on the path between reading and doing.
02-14-2019, 06:44 AM - 1 Like   #10
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 10,056
QuoteOriginally posted by kaseki Quote
...but should include a set of LEDs that together have the same spectral output as those in the monitor that illuminate the LCD panel.
No, depending on what you are after, that shouldn't do the trick.

If you want to check the monitor's performance by using the colour checker as a reference, you need to light the colour checker with a source that is compatible with the one that was used to capture the colour checker you are displaying on your monitor. If you are displaying a simulated reference colour chart then it should state what virtual light source is implied (will typically be daylight, I'm assuming).

QuoteOriginally posted by kaseki Quote
A color bouncing off of a color could not be used, I believe.
Yes, I agree. That would give you a squared spectrum rendition of the colour (which isn't the same as an amplified rendition of the colour).

QuoteOriginally posted by PDL Quote
The whole idea about using a color chart is to create custom color correction for a given light source.
Yes, but I understood the OP as wanting to use a colour chart to validate the monitor.

With the correct illuminant (I understand regular daylight would do; at the right time of the day, unmodified by colour walls, etc. of course), the monitor's rendition of a colour checker should look very similar to the real colour checker, provided the monitor is properly calibrated (and profiled).
02-14-2019, 12:33 PM   #11
Loyal Site Supporter
Loyal Site Supporter




Join Date: Dec 2017
Photos: Gallery
Posts: 527
Original Poster
QuoteOriginally posted by Class A Quote
No, depending on what you are after, that shouldn't do the trick.

If you want to check the monitor's performance by using the colour checker as a reference, you need to light the colour checker with a source that is compatible with the one that was used to capture the colour checker you are displaying on your monitor. If you are displaying a simulated reference colour chart then it should state what virtual light source is implied (will typically be daylight, I'm assuming).

Yes, I agree. That would give you a squared spectrum rendition of the colour (which isn't the same as an amplified rendition of the colour).

Yes, but I understood the OP as wanting to use a colour chart to validate the monitor.

With the correct illuminant (I understand regular daylight would do; at the right time of the day, unmodified by colour walls, etc. of course), the monitor's rendition of a colour checker should look very similar to the real colour checker, provided the monitor is properly calibrated (and profiled).
Thank you again. I wanted to validate the monitor and any effect that the presentation software might induce, which might vary between Raw Therapee, Darktable, Firefox, or one of the many Linux image viewing apps such as gThumb.
With the correct illuminant (I understand regular daylight would do; at the right time of the day, unmodified by colour walls, etc. of course), the monitor's rendition of a colour checker should look very similar to the real colour checker, provided the monitor is properly calibrated (and profiled).
This is the crux of the matter. I haven't looked, but I assume that x-rite has images of their color checker taken at different color temperatures and then white balanced. If not, then let's assume that we can take such an image. In order to compare a monitor view of that image to an in-hand eye-ball view of the/a color checker, the in-hand color checker needs to be illuminated by a source matching that used to generate the image being displayed. (Y/N?) I'd like to be able to do this without dragging the PC and monitor out into the sunlight.
02-14-2019, 01:05 PM   #12
Pentaxian
emalvick's Avatar

Join Date: Feb 2008
Location: Davis, CA
Photos: Gallery
Posts: 1,615
If you are calibrating a monitor, I can't see where you would need the color checker chart?

I have an older x-rite calibrator, and the software basically requires you to put the sensor over the screen, and the software uses its own set of colors to calibrate the monitor. It removes all the variables you are concerned about.

The passport is more for profiling your camera and other devices. Again, I think the methods are somewhat straight forward, because one of those squares on the passport is a neutral gray (like a graycard), and the software will know what the color temperature was when you took the shot. But maybe, I am oversimplifying this. I've not profiled my cameras before, I just looked closely into it for a while. I'm generally happy with having my monitor calibrated, so never went through with buying the color-checker.
02-15-2019, 02:59 AM - 1 Like   #13
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 10,056
QuoteOriginally posted by emalvick Quote
If you are calibrating a monitor, I can't see where you would need the color checker chart?
I sympathise with the OP's pursuit because a hardware calibrator cannot always be fully trusted.

First, there is inter instrument disagreement (which appears to be higher among the Spyder products than the X-Rite products).

Second, different backlight technologies require different compensation for the particular filters used in colorimeters. Only a few calibrators (like the X-Rite i1 display Pro) support the interrogation of actual filter characteristics and the software has to be able to account for the data. If the calibration software isn't tailored to the calibration device and the particular panel used in the monitor your calibrating, results won't be optimal.

Third, even spectrophotometers (the most expensive type of hardware calibrators) are not immune to certain backlight technologies. There are some CCFL panels with spectral spikes that can interfere with the limited resolution of a spectrophotometers yielding incorrect results.

Fourth, your eyes may not be well aligned with a CIE standard observer and hence calibration results may not always produce a correct perceptual result.

I have a monitor that regularly is calibrated to produce a green tint, with a number of calibration devices. Obviously, the devices see the monitor output as "more magenta" than my eyes do and hence compensate in a direction that is not needed. Before you conclude I'm colour blind, I have very good colour vision, even in the blues which males can be bad at. I don't have the same tint problem with other monitors either. I did notice, though, that depending on the software you use and the particular calibration settings you choose, sometimes you can calibrate a number of monitors (I have four displays I use all the time) so that they pretty much (but not entirely) look the same, but sometimes the monitors display pretty different renderings easily recognisable by surprisingly inconsistent white points.

In summary, I definitely recommend hardware calibration but it is not a guarantee for getting an entirely correct result.
02-15-2019, 09:09 AM   #14
Loyal Site Supporter
Loyal Site Supporter




Join Date: Dec 2017
Photos: Gallery
Posts: 527
Original Poster
Thank you Class A. I was going to respond to emalvick by invoking cases I've read about where monitor calibrations were performed but the users found the results unsatisfying. I was going to appeal to the idea of having an independent method of checking the result of the calibration. But your message is much more thorough and grounded.
02-15-2019, 02:46 PM   #15
Pentaxian
emalvick's Avatar

Join Date: Feb 2008
Location: Davis, CA
Photos: Gallery
Posts: 1,615
I can appreciate the reasoning.

I am hardly an expert, but back to Class A's post, I see where I have the X-Rite i1 display Pro, which I guess is for the better. I know back when I got it, I did research to the extant I could. 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.

My solution in that case was to leave the monitor at its default and then let the calibrator work without the adjustment. 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). Anyway, I think I got lost that you were trying to verify your calibrator, which is what I essentially did in a layman's way. I was happy and never thought much about it since.

I know that isn't ideal, and I suspect that if I would have had problems, I would have had to dig in deeper like you are doing here.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
checker, color, colors, display, e.g, idea, image, lamps, light, lights, monitor, passport, patch, photograph, photography, photoshop, respect, screen, service, source, space, spectrum, swatch, target, test
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Architecture illumination.... dcmsox2004 Post Your Photos! 3 07-22-2018 06:09 AM
Architecture glare and illumination zztopd Post Your Photos! 7 02-10-2014 02:54 PM
New ACR Lens Correct vs K-5 in Camera Correct emblemex Pentax K-5 12 12-08-2010 12:20 PM
Night Grand Illumination - Colonial Williamsburg cdurfor Post Your Photos! 7 12-06-2010 11:37 AM
Falls Illumination Edicon Post Your Photos! 10 08-05-2009 05:44 AM



All times are GMT -7. The time now is 05:10 AM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top