RustyBug 
Offline
• • • • • • •
Upload & Sell: On
|
hanay78 wrote:
Dear @RustyBug@ thank you very very much for your answer!!!!!! 
I have checked with suncalc. That day, when I took the photograph, the sun was 6.5 degrees under the horizon. That is, the moment represents the interface between the civil and the nautical twilight.
This is a moment in which I like a lot to shoot, since it represents the interface between seeing something and not seeing anything  Also, in terms of public lighting, it is already working, bulbs are hot delivering full light power, and landscape is still partially visible in a dim light. I suppose that the later, for many photographers, could be considered a boring light. I like it, but I find it very challenging
I do not want to correct for "good" lighting. I hope I know how to do it. Aesthetically, I do not like that at all. I want to present the subject in soft dim lighting. I cannot visually share the feeling of cool colors in twilight in general. But maybe this is me, or better said, that I cannot interpret what I see at twilight in terms of cool colors.
I can state my intentions as: I want to represent the beauty of a place with the delicacy of the dim soft colors of the twilight. For me this is a rich palette, and not forcefully a cold one thanks to westerly sky lighting, reflection of this in clouds, and public lighting. In terms of L, in Lab, I want things to be seen good enough but in the threshold of dimness. This is what I perceive in the field. That is also were I struggle with the equipment and my own incapacity, that gives me, with different equipment, very different results IMHO.
From what I read, in the field, my perception make a kind of HDR treatment. It produces, in the conditions in which I took these photographs, no shadow clipping in the field with my eyes. It produces deep shadows with texture, something I would recognize as zones 1 to 2. This is not in my photographs, in which deep shadows are much darker. I prefer aesthetically darker punchier deep shadows.
I am discussing not only this photograph but in general how to systematize the post-processing of these images, where I struggle systematically.
To reproduce a realistic dim light scene, in the threshold of civil and nautic twilight, what is in your opinion, the adequate L value of white things illuminated with natural light?
...Show more →
Excellent discussion ... 
Jumping to the "adequate L value" of white things illuminated with natural light ... somewhere between zone 1 and zone 9. 
Imo, where you decide to land your whites is part of the key regarding the mood you are trying to convey. The premise that there is some "systematically" approach (i.e. mechanical / objective) to achieving a subjective / emotional response is one that kinda needs to be "kicked to the curb" (imo).
That said, the "math" of light is exponential, not linear. Because the human perception is "adaptive" to this exponential relationship, the key here is not to have a specific landing point. Rather (imo) to have the relative difference(s) that are indicative of a given time of day / illumination. Conceptually, the math is kinda like this ...
Daylight EV15 (i.e. Sunny 16) is 2^15
If this level of illumination is striking a "100% specular" surface, the return will be the same 2^15 value ... scaled to 0-255, this is 255.
If this level of illumination is striking a "white" surface, with a reflective value of 90-95%, then 255 *.9 = 230. If you take a look at color checker values, you'll see where the whites are in the 220-235 range-ish, for "correct" exposure.
Now, let's go to twilight.
Twilight EV10 - EV12 (we'll use EV10)
2^10 ... or 5 stops less than EV15.
If we use the Zone system of roughly 10 stops, then each stop of light on the 0-255 scale is ballpark 25 points. So, a reduction of 5 stops is ballpark a reduction of 125 RGB. So, 230-125 is around 105.
Thus, the same 90% reflectivity (i.e. white) when illuminated by EV15 vs. EV10 will have a roughly 5 stop reflectivity return difference.
So, if I'm looking at an "L" value for white during twilight, then it is going to be around 105/255 is around .42.
That said, bear in mind that the over values incur a similar reduction in return. i.e. a middle value of 128 (L50) in EV15 lighting, will return (naturally) a value much lower in EV10 lighting. NOTE: I "converted" an exponential to a linear when I said approximately 25 = 1 stop. From an EV15 perspective, that holds, but as the reduction to EV10 is exponential, the corresponding reduction will also be exponentially calculated. Point being, that if I use 25*5=125, then times "middle" 50%, that 5 stops isn't 125 points, but closer to 75 points, so 128 - 75 lands middle around 50. Now, we have a spread of "white" > "middle" around 105 > 50 (vs. EV15 of 230 > 128). 10% blacks follow similarly.
FYI, this is the first time I've "calculated" where whites would land, naturally in twilight. But, conceptually I've accepted that I know it is significantly less than when illuminated by less light, it will have a significantly lower reflective return (and knowing that our eye / brain accommodates relatively vs. absolute).
So, where we are "told" of a lot of things about what the values "should" be ... the vast majority of those so called rules of what it should be are correlated to "normal" lighting. As we move into the realm of reduced light ... then, yeah natural light return is going to be significantly reduced, and trying to follow the guidance for "normal" light response with greatly reduced light response can lead us into a situation of trying to fit a square peg into a round hole.
Now ... while I laid out some math, DO NOT feel compelled that the values should follow this systematically. But, just understand that the exponential return value of reflectance means that "white" will have a lower value (naturally, speaking) in EV10 than EV15, and it might be "much lower" than folks think they would be. So, if you want to present it naturally, then it's likely going to be "less than" what folks tend to think that "white" should be.
Ultimately extreme example would be a white sheet of paper in a dark room. With 0EV reflecting off the 90% reflective surface, the return would be 0 ... naturally, speaking. So, if we are photographing a white paper in a dark room, the decision of where the photographer desires to land the white paper, a natural presentation would be a value of 0. Yet, if the desire is to show the white paper white, then raising the exposure (i.e. moving it from its natural EV0) to EV15 would bring it back up to the 230 range. So, when I stated Zone 1 - Zone 9, that wasn't just a "whatever you want" answer. Rather, it was that the Mood / Time of day (i.e. variable amount of EV light) will result in a varied Zone for landing things, depending on where you want to convey time of day. Also, the reflectance value of your subject will impact that. A high sheen, white (98% reflectivity) will return more light than a dull, diffuse surface, illuminated with a soft, diffuse light, even if both are "white".
Long, I know ... but understand how light reflectance naturally occurs, is a piece of the puzzle if we are striving to present a natural look. ETTR and other aspects of "setting white points", etc. are mostly guidance for achieving a "correct" (i.e. correct for EV15 / natural daylight) ... but, that does not hold the same for reduced natural light, if you want to convey the natural light levels (vs. adjust ^ EV15 levels).
So, don't be afraid of the dark, if you're wanting to convey the dark. How dark (vs. light and the relative exponential relationship) you want to convey is up to you. 
There's more ... but, I'll rest for a moment.
|
