White balance "Setting" picker doesn't differentiate between "As shot" and "Daylight"

I’ve only noticed this very recently. I’m not 100% sure it’s all photos, but it seems that picking “Daylight” gives the same results as “As shot”.

For example I have a photo I’m working on now which has 5270, -4 as the “As shot” values and picking daylight leaves both numbers unchanged. If I pick “Cloudy” it goes to 6000, 0 and then I pick “Daylight” and it goes back to 5270, -4.

That depends on how PL interprets you camera setting.

I always have my camera WB set to 5600°K / 0 (confirmed by ExifTool) and yet PL insists on showing me 5510°K / -3.

Check what temperature your camera WB is set to. If it is set to auto, this might explain a lot.

Doesn’t picking “Daylight” in the WB panel in PL4 always give you the same numbers, no matter what the photo?

In fact, according to the manual.

So in my mind, when I select “Daylight” I expect to see 5200, 0 and not 5270, -4, which is what my camera decided the WB should be at the time.

There’s a very interesting article on Wikipedia

But, even without that, from many years of shooting large format film, I can tell you that daylight colour film is always marked at 5600°K.

“Daylight” also depends on where you are in the world and what time of day it is.

Here’s a series of images I took for teaching WB. All were taken by measuring the light with a Minolta IIIF colour meter and setting the camera to that reading.

With the light behind me 5150°K

Against the light 8580°K

There appears to be quite a difference in rendering if you look at the whole image.

But, if you look at a crop of just the seat back…

With Crop Contra Crop

… it is very difficult to tell which is which.

The PL manual is both wrong and right at the same time. 5200°K is at the lower/warmer end of “daylight” but could be accounted for if the manual writer lives in the middle of Paris when the air pollution affects the light colour :wink:

If you are using auto WB on your camera, my guess is that PL will try to match whatever that temperature is at the time of shooting.

1 Like

In PL daylight: 5200,0. That should be the value when you choose daylight. It’s hard coded.


1 Like

Not for me, using a Canon 90D with the camera set to ‘Auto White Balance’.

In PL (3) and using ‘As Shot’, the temperature and tint varies from image to image, which is what I’d expect because using AWB will mean the camera will write a different temp. and tint to the EXIF on each shot and in turn PL reads those varying data and uses them.

If though I set PL to use ‘Daylight’, the temperature is always 5200K and the tint is always zero. Again, that’s what I’d expect. PL is ignoring the (variable / AWB) temp. and tint data in the EXIF and applying its own, fixed, values.


I’ve never seen selecting Daylight result in anything other than 5200, 0.


It’s not a question of the theory of WB or any interpretation. My understanding is that the WB “names” are a form of preset, and therefore that this should not happen.


That’s wrong. It should be 5200,0. I’m on windows. No interpretation, it’s hard coded.


It might be “hard coded” from PL’s point of view but 5200°K is not the normally accepted temperature for daylight.

If a camera has been set to automatic WB, the image will already have been adjusted, in camera, to what the camera thinks daylight should look like. What that temperature is will depend on the camera manufacturer. Then PL has its own idea of what daylight is.

Capture d’écran 2021-09-05 à 09.13.05

According to this chart, daylight can be anything between that produced by a low sun (5000°K) and that produced by the sun being directly overhead (5500 - 6000°K). The generally accepted definition of daylight for photographic colour integrity has been 5600°K for many decades.

My guess is that DxO, along with some camera manufacturers, have chosen 5200°K because it produces a slightly warm rendering, which psychologically makes people feel “happier”.

The truth is, you are free to choose whatever temperature suits the mood you wish to convey in any particular image.

But to return to @zkarj 's original subject, if the “as shot” automatic WB temperature is near enough to what DxO calls daylight, then you won’t see any obvious change. The difference between 5200 and 5270 isn’t worth worrying about.

That table goes further. It also mentions 6500 daylight overcast. So the daylight range is between 5000 and 6500.
I think you misread the post of @zkarj. PL doesn’t show the right value, whatever that value might be. If PL uses 5200,0 for daylight, than that value must be shown when daylight has been selected. It doesn’t.
Beside that, calculating can be done only with a fixed number. As long as it is between 5000 and 6500 for daylight according to that article. If you don’t like the result you can to correct that.

Checked CaptureNx2
Daylight is divided in 3 subgroups: direct sunlight 5200, clouded 6000, shadow 8000.


I’ve just checked on my Mac mini Big Sur DPL 4 and choosing Daylight invariably put to 5200K and 0, even if from camera it’s 5248k -4.

I think we’re still at cross purposes here.

The files in question are RAW files and thus the camera’s interpretation of the light at the time is just a pair of numbers recorded into the metadata at the time the file was written. By default, PL uses these numbers on the RAW data to come up with an interpretation in the RGB space and it marks this behaviour by showing those numbers and listing “As shot” in the drop down list. The user is then free to choose whatever values they wish according to better knowledge than the camera and/or artistic desire.

The WB palette in PL is a pair of sliders/numeric entries to set those values numerically, or a short list of “standard” values as determined by DxO which can be used as a form of preset. In testing, I can choose “Cloudy” from the preset list on any number of photos and it always sets the numbers to 6000, 0. Without fail. If I pick “Shady” I get 8000, 0 without fail. I would expect, per the manual, to get 5200, 0 without fail when picking “Daylight”, as these numbers have nothing to do with the camera or the photograph in question. They are pre-chosen values baked into the software.

But that’s not what happens. Sometimes I get 5200, 0 and sometimes I get the very same numbers that the camera chose.

I think you’re right and I apologise for having missed the point.

If only that were true.

Here is how I set my camera…

Here is what ExifTool reads the white balance…

exiftool -whitebalance _JNA0002.NEF
White Balance                   : 5600K

And here is what PL4 shows…

Capture d’écran 2021-09-05 à 10.30.29

So, yes, there is definitely something weird going on with how PL “interprets” the white balance.

@StevenL This has been consistently wrong since, at least, PL3 (I don’t have PL2 on my machine). Surely, this could do with investigating?

I don’t think any converter shows the same numbers as in the camera when choosing ‘as shot’, except for the brands own converter.


Indeed. I just tried opening the same image in Affinity Photo

Capture d’écran 2021-09-05 à 11.58.51

Capture One

Capture d’écran 2021-09-05 à 12.03.58

On1 PhotoRaw

Capture d’écran 2021-09-05 à 12.08.16

:roll_eyes: :crazy_face:

The ‘as shot’ or ‘camera’ is always calculated by the converter based on the image itself.
I’ve been trying to find out how that happens.
Google on ‘raw demosaic algorithm’ and look for the more technical sources. Also with ‘raw color temperature algorithm’.There’re more ways leading to Rome. Dutch expression. :grinning:


[5200 is colder than 5600 but the difference is small

This is where it gets confusing :nerd_face:

If you think it the sense of physical temperature, you are right.

But in the sense of colour temperature, it is the inverse. Lower colour temperatures appear warmer than higher ones.

Do you remember this physics lessons at school when the teacher told you to choose the blue part of the flame because it was hotter than the yellow part?

I think that’s also a mistake. The statement 5200 is colder than 5600 is right. The way how it is used in photography is wrong.
You watch an image with color temperature 7000. What do you see??? You don’t see an image with 7000K, you see an image that’s corrected for 7000K. If you change the coler temperature to 10000 you’ll see the color is getting reddish. This is NOT for 10000 is more reddish but for the overcorrection you use.

You use a color temperature of 5600 fixed. Let’s say 5600 is neutral white, no correction. You’ll get the image without correction, more as you see it yourself. I believe PL is using 5400 for that purpose, being the initial value. Not sure, but it is around there.

I could touch white hot iron for a moment, I couldn’t touch red hot iron. White hot is so hot that the moisture in your hands is evaporates so quick that it creates a vacuum between your hand and the iron. A molding lesson.