This was one sunny moment in the 36 hours or so we spent in Dijon (beautiful, but a desert (single S intended) for coeliacs). Thanks, sleasyjet for nearly screwing up our New Year celebrations!
Hopefully I have applied Joanna’s techniques - at least the ones I have read about - right.
J’aurais eu moins froid le jour ou tu a pris la tienne, on dirait. Les monuments, l’architecture nous ont plu. 'y a que le probleme gluten qui nous empecherait d’y retourner.
It must have been warmer when @BenThon took his shot of the ducal palace. We loved the variety of architectural styles. Would happily go back if we could find somewhere to eat GF.
Well, there’s nothing wrong with your rendering, no matter how you went about it.
Here’s my version (barely different), using my habitual approach…
… and here is the DOP with my VC added…
D6701829.ARW.dop (32,7 Ko)
I think my main differences were setting the colour temperature to 5600/0, as the sky and clouds seemed a bit magenta. And I tend to use bigger areas for the Smart Lighting markers. Oh, and I tend to avoid using the Selective Tonality sliders, so I worked more with the RGB curve, which also raised the saturation slightly.
Although it’s all a matter of personal taste.
You are probably right about the white balance. I didn’t ever think about the light temperature until I started with digital, and I still forget to reset it nearly all the time - that’s one of the main reasons I started shooting RAW.
Is there some formula for averaging the wb between several different observations? The white van in the left foreground suggests 8636/2, the louvre door in the centre next to the statue 5475/-1, and the white van on the right (in the side street) gives 5284/-10…
Having tried all three, I think I prefer the original camera setting of cloudy - it was a very windy day, and the blue sky was fleeting, but with the tint set to zero…
How do you get a local WB? WB is global.
Or did you try several surfaces?
George
@George, he’s not getting a local WB, he’s simply reporting what the WB pipette gives when he clicks on different ‘white’ objects in the scene.
I realized that later.
George
My main experience with colour temperature started with LF (large format) film, where it is possible to define the temperature for each sheet of film. My film of choice is Velvia 100 transparency film, which is balanced for 5600°K.
So, every shot I take has to be balanced to that temperature in order to avoid horrendous casts, by using colour correction filters that are various degrees of blue or orange, in order to offset any cast produced by the ambient light.
Needless to say, for film, I use a colour meter that not only takes a reading of the light, it also tells me which filter(s) to use.
I made a small series of digital example images that demonstrates how the direction you are shooting in affects either the filter requirement for film or the WB for digital.
To start with, I shot with the light coming from behind me and, with AWB, got this…
I measured the temperature of the light with my colour meter and it gave me 5150°K, so I used PL to set that temperature and got…
Not too much difference but still noticeable.
Then I turned the chair around , so the light was coming from in front of me and, with AWB, got this…
I measured the temperature of the light with my colour meter and, this time, it gave me 8580°K, so, again, I used PL to set that temperature and got…
Now, the measured and corrected shots look very different but, if I take a crop out of both shots, and place them side by side…
… you can see they are identical.
In your image, Mike, the light is coming from behind and to the right and the majority is in shadow so, if anything, the image is going to be fairly cool and, as you found, the white van gives a “bluish” temperature. But, if you measure the sunlit building in the distance, you get a warm rendering. Both of which are not “right” - which is why. I stuck with 5600°K, so most of the image looks “right”.
My normal practice for digital WB is to always shoot at “daylight film” temperature (5600°K) and then decide what I want to be pure white, if anything, in PL and use the pipette on that. But that is not always the case because I may want to evoke a certain atmosphere of warmth or coolness and, since I am not shooting for a catalogue, where absolute colour matching is important, having the camera always taking at a fixed temperature and tint (5600/0) allows me to adjust the ambiance in PL more easily.
Finally, don’t rely on PL for absolute colour temperatures, My Nikon D850, set to 5600/0 usually comes into PL at 5643/-3, but that has changed as PL has been updated.
PL’s idea of “daylight” is 5200, which I’m guessing is to match digital screens but is way too cool for my taste. And, in the end, how white your whites are is all a matter of personal taste, not fact.
By a strange co-incidence I am trying to get my head around Chevreul’s law of simultaneous contrast atm… Not in the original; my son gave me Dan Margolis’ translation/expansion at Christmas. Slow progress because I didn’t get any physics at school.
Watch the pixel values before and after when using the pipet. You’ll see that PL is trying to correct to a neutral grey value.
You can use the pipet only on a gray, even values of the channels, surface. Gray in the real world. If that surface isn’t gray on your image, then it’s likely illuminated by a non white light source. The pipet is trying to correct that.
Gray is defined as even channel values, that includes black and white.
George
Thanks, George.
I did know that was the idea - it’s just that there isn’t always a grey, and neutral enough, pixel to be found in the picture,
Which is why I wondered about a way of averaging the whites…
In the meantime, I am going to give 5600/0 as default for daylight a whirl. I’ve often found the standard Sony setting (5400, IIRC) a little cool.
Cheers
Mike M
PL uses 5400 as the temp where no correction is done. Even daylight dosn’t have to be white, it changes all over the world and time. Just chose a satisfying temp.
George
I agree. There’s no way to have PhotoLab automatically average between two sets of values for white balance - but it’s a simple matter to choose numbers for temperature and tint that fall in between two sets (or in the range of three or more sets) and see if you like the result. If not, adjust further. I do this regularly with mixed lighting - it only takes a few seconds.
The reason most daylight films were set to 5600° Kelvin was that noon day sun shine at midday on a clear sky, was considered to be 5600° Kelvin. And as most films were sold in the area between roughly 47° north to 52° North. That was why 5600° was settled on.
If you had a window facing the north. Light coming into that was considered to be 6000° Kelvin. This meant that anybody outside of those latitudes needed to use filters and if I remember correctly they were hard to get hold of, and very expensive in those early days.
Like @Joanna. I have always set my camera’s to 5600° Kelvin. it makes it much easier for corrections in whatever processing program you use. If I had a colour meter. It might be a different story.
