Im not disputing that. so explain the relevance/measure of scene luminance without the histogram.
June 16, 2023
577
-
-
Scene luminance is basically the light coming from the scene and I gave examples of both the RGB and luminosity histograms for a given image earlier.
I explained what the histograms are and represent in the earlier post and in the post I linked to.
-
your not answering my question. i will put it even more simply what advantage is knowing scene luminosity measure .(even though no one can even put a value on it) compared to just using the EVF considering it shows the most detail from any form of measure . so WYSIWYG is basically the ultimate luminosity meter.
-
The histogram in the EVF is based on the rgb values in the image data generated from the raw data.
It isn't a histogram of the raw data generated by the light hitting the sensor.
-
the raw histogram can not show fill flash levels 😁 in fact its useless in the field compared to the EVF
-
When you design a sensing system you start with a transducer, the device that turns whatever you're sensing into something you can measure - usually electrical voltage. Quite often what you want to measure is not the thing that you're sensing directly but something that you can derive from something you can measure more easily. For instance, ammeters usually work by measuring the voltage across a resistor, not the current(amps) directly. Then you'll generally have some signal processing - which conditions the signal (where 'signal' means the 'message' that you're trying to receive), does any processing that is necessary to extract any derived measurement from the data that you actually have , and converts it into a form that you want to display or log. Whilst the data from a histogram (and indeed the final image) is derived from the scene luminance, it's not directly measuring the scene luminance. A number of processing operations have been done which convert it into a different form, which is essentially a recipe for rendering a scene. Think of that as like painting by numbers - a list of paint colours and a drawing of the scene with a number in each patch saying which colour should go there. When yo've painted the scene you know something about the distribution of luminosity across the scene, but from the painting by itself, you couldn't make a good estimate of what the original scene luminance was.
-
Thanks Bob, i like to call the resistor a shunt resistor 😊
in fact when im setting up amp meters in yachts i actually use the negative 400 amp battery cable and spike a wire in and use a 10millamp meter to read the amps. -
You're going off on a tangent because nowhere did I suggest a raw histogram can show fill flash levels.
The in camera histogram is at best an approximation of the raw histogram because the in camera histogram is based on the rgb values processed from the raw data, as I stated earlier.
For a given scene luminance the in camera histogram will vary depending on variables like the metering mode, the WB setting, the camera's 'picture style' settings set by the user etc.
-
And don't forget the most important setting - ISO.
-
.....🙂
-
You'd use a millivolt meter to measure the current going through a shunt resistor. But your milliamp meter will act as a shunt to the shunt if you know its resistance. It itself is a milli (or micro) volt meter measuring the voltage across a resistor. If you shunt it the current will divide between the two resistors in proportion to the resistance.
-
Ah, to the root of the problem.
This thread is all over exposure as a single setting, which it most assuredly is. One setting to capture all the relevant reflected (and sometimes incident) light in the scene, the intensity of which varies significantly as splayed on each individual "pixel" of the sensor. Yeow, one setting...
Each and every scene presents a unique situation with regard to light. The job of "one-exposure-to-rule-them-all" is to put as much light as possible within the ability of the sensor to resolve it, or at least the light one cares about. For most scenes, the old "trying to put two gallons of water in a one-gallon bottle" saw. A lot of folk meter the exposure to put what they're interested in (the subject) at around middle gray, and let the shadows and/or highlights go where they may. Others (like me) try to put the highlights at the top of the sensor's resolvability, and let everything else, including the subject, go to the darker sensor regions. Both require attention in post, just require different tools to mess with. If one is pushing highlights to oblivion, they need a good highlight reconstruction tool. If one is pushing everything down into resolvability, I find that requires a good sensor to provide workable shadows and finely-controlled tone curves. Pick yer poison...
Messing with the light of the scene can change the exposure needed. Now, some messing with light is about having less of the scene in darkness, a lot of movie and studio lighting is about shadow filling, which may not change the eventual exposure the photographer wants. But, it's hard to direct such fill at just the pesky shadows, so it can also further illuminate the well-lit portions and thus cause an exposure adjustment, the one-setting-to-rule-them-all...
Me, I want tools that help me put the majority of the scene's light in the sensor's ability to resolve. A histogram can help with that, but only if there's a lot of brighter light in the scene to make the top of the histogram spike. If only a very small part of the scene has such light, the leetle histograms in the EVF may not show it such that you can see it. So, to Don's point, I'd prefer a "blinkie" mechanism that puts an indication of "brighter-than-sensor-can-resolve" indication on each pixel thusly affected. Current such mechanisms aren't based on the raw data, so not helpful...
-
The problem with this article is that it allow things like "Light meters operate through a known constant The value all light meters average for is middle grey". Not only this excludes evaluative / matrix metering, highlight priority metering, etc., it isn't correct for raw even for average or spot-metering. There is no "middle grey" in raw.
In the article they talk about raw ("A well exposed camera RAW capture..."), but use grossly distorted raw data (rendered JPEGs) to illustrate it.
And when they say, "The correct exposure is emotionally subjective", it seems they've dropped their hands in frustration. Exposure for raw has very little to do with the look of the final image, as it is with a negative, where it can be printed in a myriad different ways; while it is that look that causes emotions.
Don't you recognize nonsense when you see it? - "The camera / meter will read the luminance values of the sand, determine the darkest and lightest values, then average them to ‘plot’ middle grey." That's not how it works.
Where?
-
When I do that with my Camera, it looks like the Milky Way. Camera is on its way to Nikon
-
I read it through once again, but did not find. Maybe it is so hidden I can not find it.
The other article capturetheatlas.com/how-to-read-a-histogram-in-photography/ starts with exposure levels and also continues with them.
-
Maybe Donald in good faith will take a screenshot from that article and post it here for us to see "Exposure levels".
We would be a better community if we were to quit batting obscure one-liners back and forth, without references from the literature.
We would be a better community if we were to explain stuff more understandably to Donald so that he doesn't respond with aggressively defense tactics designed draw attention away from a lack of technical knowledge. The more Donald understands about basic photography, the better contributions he will make to this site.
The cognoscenti of this site should realize that an overly-technical response to a simple question can be counter-productive and can cause a piqued response.
The current trend of Donald-bashing does not reflect well on us as a community. Let's help him to realize that it's not a good idea to bite the hand that feeds ...
-
As you well know, one can have the same exposure and different raw histograms - with a flick of ISO setting. So exposure levels, present in the article or not, is some sort of an oversimplification, if not a straight error.
-
OK, it was stupid by my side to pick some words. I'll stop.
