Kelvin Temp System: do you use it? Why?

neuroanatomist

I post too Much on Here!!
CR Pro
Jul 21, 2010
24,647
2,167
Zeidora said:
For critical shots, I do a custom white balance, particularly, if there are no good white/black/neutral grey points to do a soft white balance in photo editing software. The RAW adjustments are all good and well for mood shots, but for color accurate reproduction it is better to make adjustments before shooting, and not adjust the RAW files further. Sometimes there is no option for including a color checker (e.g., compound microscope images).
I trust you're aware that an accurate white balance isn't the same as a full color profile with a ColorChecker. If you're setting a custom WB in the field for shooting JPG, that's ideal. If you're shooting RAW, it's a waste of time that could be spent shooting, IMO. "If there are no good white/black/neutral grey points," in the picture, well...you shot the exposure reference for the custom WB, just use that in post (also, custom WB limits what you can use for that reference to something large and gray, for critical work I prefer to include a more comprehensive reference like a ColorChecker or SpyderCube). As I keep saying, if you're shooting RAW, there no difference in terms of when you apply the WB, except whether you choose to spend the time while shooting vs. at more leisure in post.

For your microscopy, since what you want to balance is the light source, you can usually use the same no-slide image for flat-field correction and to set WB. If you really require a neutral gray WB reference, you could use the Applied Image IAM-4 (a density step wedge in microscope slide format; I use one for microdensitometry calibration).
 

retroreflection

EOS 90D
May 19, 2015
124
5
Describing a light source by Kelvin Temp assumes the light source behaves as a black body radiator, correct?
Terribly grey radiators exist, and LEDs and fluorescent lights generate white light in a manner totally separated from black body emissions (the trick to their energy efficiency).
Manufacturers of these lighting systems claim to conform to color temperatures, but shaving a few nickels here and there is the key to everyday low prices.

It seems to me that a blanket utilization of Kelvin Temp (assuming jpeg shooting for speedy delivery) is setting yourself up for disappointment in some lighting situations. A more open mind may be in order. A custom white balance does not impose conformance to the black body curve which just might be the reality you need to see.
 

Zeidora

EOS RP
Feb 15, 2015
667
10
neuroanatomist said:
Zeidora said:
For critical shots, I do a custom white balance, particularly, if there are no good white/black/neutral grey points to do a soft white balance in photo editing software. The RAW adjustments are all good and well for mood shots, but for color accurate reproduction it is better to make adjustments before shooting, and not adjust the RAW files further. Sometimes there is no option for including a color checker (e.g., compound microscope images).
I trust you're aware that an accurate white balance isn't the same as a full color profile with a ColorChecker. If you're setting a custom WB in the field for shooting JPG, that's ideal. If you're shooting RAW, it's a waste of time that could be spent shooting, IMO. "If there are no good white/black/neutral grey points," in the picture, well...you shot the exposure reference for the custom WB, just use that in post (also, custom WB limits what you can use for that reference to something large and gray, for critical work I prefer to include a more comprehensive reference like a ColorChecker or SpyderCube). As I keep saying, if you're shooting RAW, there no difference in terms of when you apply the WB, except whether you choose to spend the time while shooting vs. at more leisure in post.

For your microscopy, since what you want to balance is the light source, you can usually use the same no-slide image for flat-field correction and to set WB. If you really require a neutral gray WB reference, you could use the Applied Image IAM-4 (a density step wedge in microscope slide format; I use one for microdensitometry calibration).
Agree on WB and profiling being distinct. That was not OPs question, so I assume pre-profiled camera-lens system.

Re RAW adjustment vs. pre-shoot adjustment, you are missing the point. Consider you shoot an all-over picture of orange flowers. There are loads of shades of orange, and the hue will be affected by the K setting. So what is the true color? Unless you pre-set the K value (or do custom WB) before you shoot, there is no way of getting an accurate color hue. [There is also the tint question, which is a function of exposure, again not what OP asked].

You could shoot a white balance/grey card/color checker first, then take picture of flower, then adjust K of first image in RAW and transfer that value to the flower image. I wonder what is more cumbersome, though.

Re grey wedges, fine for greyscale calibration, but not the same as a color-checker, or similar. Re balance light source vs. WB of camera, the old photo setting (3200K) on microscope incandescent light sources (plus 80 series filter for daylight film), is a bit dated, particularly if you consider LED sources these days. WRT WB/K/profiling, there is no difference between dLM and dSLR. Axiocam HRc with Zen works quite nicely.
 

mb66energy

EOS R
Dec 18, 2011
1,444
322
Germany
www.MichaelBockhorst.de
Dylan777 said:
Hi guys,
I have been shooting with AWB for years. I recently pushing myself manually adjusting Kelvin Temp in white balance. It does add another steps, however, I really like the results I'm getting :)

Do you use it? why?

Best,
Dylan
I shoot raw since 1997 so white balance is a thing for post processing. As physicist I like to have as much reliable science in my life. Kelvin temperatures are reliable for our sun, incandescent lights, very good LED lamps.

During exposure I always use the sunlight setting (5200 K) to have reliable and stable matching between scene and display.

Mostly I set K values in postprocessing for
(1) artificial light sources as "first approach" - works well with incandescent lamps and LED lamps with CRI of 95%. Works well e.g. with a slide projector as light source.
(2) photos taken in the morning or afternoon.

Ad (2): AWB gives better whites but usually I like the mood of the yellowish light while it disturbs the differentiation in greens: always a little bet smeared yellowish-green palette. By tuning the K setting between daylight and the "white K setting" I get what I want: The right mood but very good color differentiation. Example: If 3600K gives pure white, I use e.g. 4400K as a optimum setting.

Best - Michael
 

neuroanatomist

I post too Much on Here!!
CR Pro
Jul 21, 2010
24,647
2,167
Zeidora said:
Re RAW adjustment vs. pre-shoot adjustment, you are missing the point. Consider you shoot an all-over picture of orange flowers. There are loads of shades of orange, and the hue will be affected by the K setting. So what is the true color? Unless you pre-set the K value (or do custom WB) before you shoot, there is no way of getting an accurate color hue. [There is also the tint question, which is a function of exposure, again not what OP asked].

You could shoot a white balance/grey card/color checker first, then take picture of flower, then adjust K of first image in RAW and transfer that value to the flower image. I wonder what is more cumbersome, though.
Sorry, but the highlighted statements above indicate that it's you who is missing the point, rather egregiously. Those statements are true if shooting JPG, but not for RAW. The in-camera WB setting (auto, cloudy, tungsten, Kelvin, custom WB, whatever) has absolutely no effect on the RAW image data. In that regard, it's the same as the settings for picture style, ALO, high ISO NR, color space (sRGB vs AdobeRGB), etc. None of those affect the RAW image, all they do is set metadata flags that tell your RAW converter what values to display when the file is opened. You can set your in-camera WB to 2500 K or 10000 K and just leave it there. Your JPGs will mostly look awful, but for the RAW image you can set the color temp to whatever you want in post, and the resulting image will be exactly the same as if that Kelvin temp was set in-camera before the shot was taken. No difference.

To reiterate – if you're shooting RAW, there is no difference whether you apply a WB setting before the shot is taken or when you're post-processing the image. That's why it's a RAW image.

To re-reiterate, the only in-camera setting that affects the RAW image data is long exposure NR, which subtracts a dark frame from the RAW image prior to writing it to the card. No other in-camera setting – none of them, including Kelvin or Custom WB – affect the RAW image, only the metadata (meaning those settings can be altered at will in post, with no 'penalty' and nothing lost). Again, that's the whole point of shooting RAW.

As for 'cumbersome', you just select your WB reference image and all the files to which you want it applied, then use the WB eyedropper tool on the reference image. Or, if you know the color temp you want, just select a batch of images and type in that value. Either way, it's a couple of mouse clicks and done.


Zeidora said:
Re grey wedges, fine for greyscale calibration, but not the same as a color-checker, or similar. Re balance light source vs. WB of camera, the old photo setting (3200K) on microscope incandescent light sources (plus 80 series filter for daylight film), is a bit dated, particularly if you consider LED sources these days. WRT WB/K/profiling, there is no difference between dLM and dSLR. Axiocam HRc with Zen works quite nicely.
My point had nothing to do with balancing the light source. I was suggesting the step wedge for a neutral gray (i.e., reduced intensity) WB reference, as opposed to using the white light source directly.
 

greger

7D
Jan 1, 2013
259
1
I use AWB for everything except flash. I don't shoot indoors much, but would use flourescent if needed. Shooting in Raw I rarely have to adjust colour in post. Adding a blue filter in PS has enhanced a couple of pictures when I tried it.
 

YuengLinger

Sufficiently Pixilated
CR Pro
Dec 20, 2012
3,144
1,404
USA
My dermatologist said something about Kelvin the last time he "burned" off a wart with liquid nitrogen.

Ouch.

SERIOUSLY though...when I use my strobes (Einsteins) I set to 5600K so I get better looking, more consistent flesh tones on the little preview thingamajigger on the back of the camera. Otherwise I get swings, where some look right, others too warm. Setting to K works for me in this situation.

Otherwise, AWB seems great on the 5DIII--even under fluorescent lights.
 

YuengLinger

Sufficiently Pixilated
CR Pro
Dec 20, 2012
3,144
1,404
USA
rfdesigner said:
nc0b said:
I definitely use Kelvin when shooting ballroom dancing under incandescent lighting. No I don't shoot raw, as these pictures are often used on a web site or for scrap book archives of the dance organization. I have no interest in the added complication in workflow of shooting raw. Ballrooms often have fancy chandeliers with dozens of tiny bulbs, and a setting of something around 3200 K works very well. The skin tones are warm and pleasant, but not annoyingly yellow or approaching orange if AWB is used.

I may get flamed for shooting JPG, but my output with a 70-200mm f/2.8 II and a 6D fulfill the needs of the Colorado dance organization. Compared to everyone else's results with their compact cameras, the dances I shoot using Kelvin turn out excellent.
I might suggest that successfully shooting JPG for paying clients shows a degree of professionalism.. you can get it close enough in-camera, just like we all had to do with film.
Professionals I know tend to use the best tools available. But real men don't eat quiche, right?

Please remember, when talking about film, y'all had that plastic strip...I think it was a negative? Didn't it have quite a bit of leeway for working with prints? Dodgin' and burnin' and all that? JPGs limit what can be done with images, digital and print. RAWs are the soft generation's negatives, I guess.
 

3kramd5

EOS R6
Mar 2, 2012
3,084
405
YuengLinger said:
rfdesigner said:
nc0b said:
I definitely use Kelvin when shooting ballroom dancing under incandescent lighting. No I don't shoot raw, as these pictures are often used on a web site or for scrap book archives of the dance organization. I have no interest in the added complication in workflow of shooting raw. Ballrooms often have fancy chandeliers with dozens of tiny bulbs, and a setting of something around 3200 K works very well. The skin tones are warm and pleasant, but not annoyingly yellow or approaching orange if AWB is used.

I may get flamed for shooting JPG, but my output with a 70-200mm f/2.8 II and a 6D fulfill the needs of the Colorado dance organization. Compared to everyone else's results with their compact cameras, the dances I shoot using Kelvin turn out excellent.
I might suggest that successfully shooting JPG for paying clients shows a degree of professionalism.. you can get it close enough in-camera, just like we all had to do with film.
Professionals I know tend to use the best tools available. But real men don't eat quiche, right?

Please remember, when talking about film, y'all had that plastic strip...I think it was a negative? Didn't it have quite a bit of leeway for working with prints? Dodgin' and burnin' and all that? JPGs limit what can be done with images, digital and print. RAWs are the soft generation's negatives, I guess.
A hallmark of professionalism is use of appropriate tools, they need not be the best. If for example a client prioritizes quick-turnaround, adding steps post-capture is inappropriate.
 

YuengLinger

Sufficiently Pixilated
CR Pro
Dec 20, 2012
3,144
1,404
USA
3kramd5 said:
YuengLinger said:
rfdesigner said:
nc0b said:
I definitely use Kelvin when shooting ballroom dancing under incandescent lighting. No I don't shoot raw, as these pictures are often used on a web site or for scrap book archives of the dance organization. I have no interest in the added complication in workflow of shooting raw. Ballrooms often have fancy chandeliers with dozens of tiny bulbs, and a setting of something around 3200 K works very well. The skin tones are warm and pleasant, but not annoyingly yellow or approaching orange if AWB is used.

I may get flamed for shooting JPG, but my output with a 70-200mm f/2.8 II and a 6D fulfill the needs of the Colorado dance organization. Compared to everyone else's results with their compact cameras, the dances I shoot using Kelvin turn out excellent.
I might suggest that successfully shooting JPG for paying clients shows a degree of professionalism.. you can get it close enough in-camera, just like we all had to do with film.
Professionals I know tend to use the best tools available. But real men don't eat quiche, right?

Please remember, when talking about film, y'all had that plastic strip...I think it was a negative? Didn't it have quite a bit of leeway for working with prints? Dodgin' and burnin' and all that? JPGs limit what can be done with images, digital and print. RAWs are the soft generation's negatives, I guess.
A hallmark of professionalism is use of appropriate tools, they need not be the best. If for example a client prioritizes quick-turnaround, adding steps post-capture is inappropriate.
Then we are kind of getting mixed up. If comparing to film days, there was one hour photo, or polaroid...

You are right, JPG has an important place in a few situations these days, and, getting it right consistently for JPG shows great camera skill. And that's where proper color temps are critical.

But lower quality images, even snapshots, have long been acceptable for "night on the town" work, direct flash and all. And I understand that photojournalism standards lean towards JPG. Like everything in photography, trade-offs, so, for less flexibility and lower standards generally when speed or hoped for integrity is the priority, JPG.