Setting WB by temp?

FrankC wrote in post #18898221
The problem with setting a fixed WB outdoors is that the light will vary - sometimes slowly, sometimes quickly if there are lots of clouds etc.

My solution is to use Auto White Balance, and tweak afterwards in Lightroom.
However, I appreciate that this may not be so practical if you're shooting large volumes of images in JPEG.

With fixed lighting (i.e. indoors) I use a whibal or other grey card for a test image in whatever light setup is being used. I still stick with AWB, though.

Through testing, I have determine that AWB alters WB setting even when there is NO CHANGE to the lighting!... a bunch of shots in rapid succession in the space of a few seconds in a cloudless sky, yet WB setting varied considerably. They could have had a single WB value be used for all shots made within a 30 second window, yet AWB chose values scattered all over the place, with values from 4424K thru 5895K ... 60% of the AWB values had deviated from the ideal neutral gray card value of 5000K of the lighting conditions. While 40% of the AWB values were 5001K, 60% were not, and yet there were no strong colors predominant in any of the shots to unduly influence AWB value in the wrong direction!

https://photography-on-the.net …p?p=12259856&po​stcount=11

When strong colors were present in the scene in the same sunlight conditions, the error rate for AWB increased to 80% error! (Two posts later in the same thread above you can read my post, showing this result.)

I just read through your thread there.

I think my interpretation of what AWB does is slightly different to yours - to me it measures what is lit i.e. the scene.
It isn't an independent assessment of the actual light itself - but a combination of this and the scene.
You can better measure the colour temperature using a colour meter or a grey card or one of the "dome" devices.
But the natural light varies a bit anyway - even on cloudless days.

I shoot events and weddings - the light and circumstances vary so much that it's impractical to set a fixed white balance for most of the day.
I find using AWB works well (not perfectly) most of the time, and gives me a good preview on the back of the camera. Setting a fixed WB would only
work for a few images, and not make the subsequent processing any easier. For studio work, I occasionally set a fixed WB, but usually don't bother. I have a
reference image, and set the WB of all the other images the same in post-processing.

So, I don't care if the WB varies a bit from image to image using AWB. I shoot in raw, and batch correct in Lightroom - I don't correct each one individually.
I don't necessarily aim for 100% accurate colour either in general photos - I aim for what looks right to me, and is "pleasing".
If I needed to get truly accurate colour then the best answer is something like the X-rite ColorChecker.

FrankC wrote in post #18898446
I just read through your thread there.

I think my interpretation of what AWB does is slightly different to yours - to me it measures what is lit i.e. the scene.
It isn't an independent assessment of the actual light itself - but a combination of this and the scene.
You can better measure the colour temperature using a colour meter or a grey card or one of the "dome" devices.
But the natural light varies a bit anyway - even on cloudless days.

I shoot events and weddings - the light and circumstances vary so much that it's impractical to set a fixed white balance for most of the day.
I find using AWB works well (not perfectly) most of the time, and gives me a good preview on the back of the camera. Setting a fixed WB would only
work for a few images, and not make the subsequent processing any easier. For studio work, I occasionally set a fixed WB, but usually don't bother. I have a
reference image, and set the WB of all the other images the same in post-processing.

So, I don't care if the WB varies a bit from image to image using AWB. I shoot in raw, and batch correct in Lightroom - I don't correct each one individually.
I don't necessarily aim for 100% accurate colour either in general photos - I aim for what looks right to me, and is "pleasing".
If I needed to get truly accurate colour then the best answer is something like the X-rite ColorChecker.

But if you are in sunlight with cloudless sky, the light does NOT change from 1 minute to the other, it changes over the course of the daylight period from dawn to dusk. From 12pm to 1pm there is virtually no change to the light striking the scene. So a professional would set Daylight on the camera for shooting, or 5200K via the RAW convertor eyedropper sample of an 18% gray card during post. He would NOT be getting variable values during that time period, not the range with values from 4424K thru 5895K as were experienced in my linked post nor whatever range of values you might be getting. The scene content does NOT alter the WB of the sunlight which falls on the scene, nor the WB value measured by an incident color temperature meter being used during those 60 minutes (at least not unless we are shooting in the greenishness of a tree filtered forest, or the color bias caused by a nearby wall painted some dominant color).

As for not using a 'set WB' during wedding shoot, why not?!

• Daylight for all shots between dawn and dusk will capture the natural variation that the human eye is accustomed to seeing over the normal progression of the day. (We are not trying to get a perfect 'neutral' all during those hours as we might if shooting textiles for a client manufacturer who insists upon perfect color balance no matter what time of day the shot is made, for the typical wedding client...the bride knows her gown looks 'warmer' as we approach sunset.) (this is assuming sunlight cloudless sky all day, for the purposes of this discussion...leaving out open shade or clouds passing)
  
• For indoor incandescent, we could use 2900K (or perhaps 3200K if we deliberately want the photos to be a bit warmer so the brain knows 'inside shot' when it sees the photo.
  
• The time when a fixed WB does not necessarily work indoors is if there is a mix of daylight from the window and incandescent from the light fixtures indoors. If you choose a custom WB value, only ONE PART of the scene will be 'neutral'...while another part is warm and another part is cool because the balance of the light in different parts of the scene is dependent upon proximity to the window.
  

I accept that light changes will be minimal between 12 & 1pm in the circumstances you describe. I find the AWB gives me a good enough representation of the light in these circumstances. However, the camera is not an incident colour meter - so it is affected by what is being lit i.e. a combination of what you're photographing and the light falling on it. It doesn't treat the lighting and scene separately - all it sees is the combination.
As you say, an incident colour light meter should give pretty similar colour values during that period - but a camera won't necessarily.

Regarding wedding shoots - the lighting and scenes changes rapidly. The AWB gives me a much better representation of my photo (on the back of the camera) than setting a fixed WB would. I then adjust to taste in Lightroom. However, I find that the AWB usually gives me the best starting point.

Also - the above only works for raw - if I was shooting jpeg professionally, then I would operate differently.
I also use a carry-around camera (G7X II) and the jpegs from that using AWB are usually good anyway (though I actually shoot in both jpeg and raw - even for my holiday shots).

FrankC wrote in post #18898600
I accept that light changes will be minimal between 12 & 1pm in the circumstances you describe. I find the AWB gives me a good enough representation of the light in these circumstances. However, the camera is not an incident colour meter - so it is affected by what is being lit i.e. a combination of what you're photographing and the light falling on it. It doesn't treat the lighting and scene separately - all it sees is the combination.
As you say, an incident colour light meter should give pretty similar colour values during that period - but a camera won't necessarily.

And so you precisely state the reason WHY one should use Daylight WB rather than use AWB...in that situation the camera will NOT vary if set to Daylight, but it WILL vary when set to AWB.

Your argument for AWB in changing indoor light is more reasonable...use a Variable setting when the WB varies constantly.
But use a FIXED WB when the lighting is virtually fixed. Especially if the presence of strong colors throw off the AWB badly.

Wilt wrote in post #18898709
And so you precisely state the reason WHY one should use Daylight WB rather than use AWB...in that situation the camera will NOT vary if set to Daylight, but it WILL vary when set to AWB.

But - I don't mind it varying - generally I prefer it, because it gives me a better preview (on the back of the camera). It takes me no extra time to adjust in LR.
I might consider it if I was in a fixed position, the subject was unchanging and in a fixed position, I was shooting for a significant time and the light wasn't changing.
However, this represents very little of my photography in practice.

Overall, I would have to disagree with you that a professional photographer would set a fixed white balance.
A professional photographer knows what the options are and what the affect of them is. They then use those which suit what they want and need to achieve. In many cases there will be a number of equivalent ways to achieve the same result. In which case, the options chosen are just personal preference.

I did a lot of experimenting when I started photography and settled on methods which work for me.
However, it's always worth revisiting - so I'll try out the Custom/Fixed White Balance method again to see if my mind has changed !

You guys are making this waaaaaay harder than necessary

2 best ways I have found to get good color and no White balance problems

Shoot K....Never let the camera tell you what the WB is

1-Go to live view, adjust K temp so what you see in your viewfinder is what you eye sees for color.....This is the most accurate way to get whats actually on the scene and what your eye sees for what your trying to photograph. Pretty easy to train eye to see color temps

2nd way I do it

Shoot K

2-Set all day light pics to 5200K. Set all tungsten pics to 3600K and gel flash with 1/4 orange...Fluorescent..​.set wb to 4200 K use a gel flash- I use a lighter than 1/4 orange that has a little more yellower cast....once you import images then you can adjust to taste or go to Auto WB in post and they will be very close

I prefer 2nd method because I do gel and it gets my skin tones on the $$$$

I never let the camera decide what the color is at capture process- I control the capture color process...I have never used a grey card in 40 yrs...why?? because all grey cards are not created equal

I always control the color of the light that I introduce on subjects. Gel is huge for accurate colors

Photo-vision calibration targets are pretty good but not always right

I know K works especially outside. Its never wrong. The light that you introduce will have biggest impact on what you are doing

1st shot is straight off camera capture in total mixed light room. Loads on tungsten, colored lights from DJ..all kids of stuff going on

2nd shot was sunset shot at 5200 K and warmed to 6300K in post