Yes, thanks, but that simulated increase of a scene’s illumination results in the image data being amended as they describe the image and its pixels, and it is this change that I’m trying to find the technically correct expression for.
1 Like
the most precise description can be found here: https://github.com/darktable-org/darktable/blob/139f720a6dd7abc0c2ec702672e57e029bb4de10/src/iop/exposure.c#L552
1 Like
That line doesn’t say anything different (except for the black level element) that Tamas_Papp stated in post 2, which is anyhow a well understood function.
As I have so far not yet ventured into github and the source code, the link was interesting anyhow. Thanks!
If we try to look at the issue from another angle:
The three RGB values of a pixel are amended (proportionally) by the exposure function. Both before and after the “exposure correction” they add up to a common attribute of the pixel. What is that attribute called?
EDIT: By “add up” I simply refer to an abstract function rather than just summing up the values, as i.a. difference in spectral energy might need to be taken into account at the physical level.
1 Like
Exposure is what some try to offer to pay me in when using my photos.
9 Likes
Whenever the ratios of the R, G and B components to each other are intact, the chromaticity is preserved.
1 Like
Yes, chromaticity is preserved, but my question is for that component that isn’t preserved.
But since “chromaticity” is an “objective specification of the quality of a color regardless of its luminance”, I would think luminance is a possible answer.
“Luminance” is however, explained in Wikipedia as “a photometric measure of the luminous intensity per unit area of light traveling in a given direction. It describes the amount of light that passes through, is emitted from, or is reflected from a particular area, and falls within a given solid angle” (taken from the CIE definition), and that makes me wonder if it is the optimal term.
EDIT: Looking further in the Wikipedia article on Luminance we also find the following Description:
“Luminance is often used to characterize emission or reflection from flat, diffuse surfaces. Luminance levels indicate how much luminous power could be detected by the human eye looking at a particular surface from a particular angle of view. Luminance is thus an indicator of how bright the surface will appear. In this case, the solid angle of interest is the solid angle subtended by the eye’s pupil. Luminance is used in the video industry to characterize the brightness of displays.”
Elsewhere also stated that “luminance is a measure of light not over time (like luminosity) but over an area.”
So likely the best candidate, I guess.
Thanks!
I don’t think luminance would be a good alternative term. You can increase the luminance by just adding some amount of achromatic signal (1, 1, 1) to the original signal, and that is not what exposure does. Obviously, the gain applied by the exposure control also gains the luminance, but labeling the slider as luminance gives the wrong impression - even if the chromaticity is preserved, the gain still also gains the “chromatic” part of the signal, and not just luminance alone.
I can’t think of any better term than the existing ones right now. Gain would probably be the most accurate.
2 Likes
But is the fact that it generally is possible to increase luminance also by adding achromatic signal, necessarily a contradiction of classifying the change to an image actually performed by the exposure module (which keeps the chromaticity constant) as a change in luminance?
And please don’t disregard that I’m not talking about a change in the label of the slider. I get the feeling that if one link the term I’m searching for to a slider label, there’s a tendency to perceive a need for a 1:1 relationship between the label and the named effect. But that several methods can result in a certain type of effect, perhaps quantitatively different, doesn’t change the fact that each one of them might lead to this type of effect in their own ways.
Oh really? What about this definition of exposure:
the condition of being presented to view or made known
Perhaps that’s what I’m doing to my shadow information? Or vice versa in mt blown out highlights 
In all seriousness, it must surely be called “exposure” because that’s precisely what it’s simulating. With ISO invariance becoming more and more common people have already been arguing that your digital camera ISO ought to be relabeled “Gain” or “Applied Gain” as that’s all its doing, and for logical reasons that amplification performed by your ISO dial (and the exposure slider in darktable) is scaled so that any number of whole or part stops up and down correspond precisely with adjustments made to the shutter or aperture.
If you want to know more about why “stops” of light (EV) are defined the way they are you can refer to the Weber-Fechner Law.
There are at least 10 different sliders I use routinely to make my photos brighter or darker in darktable one way or another (there are probably just as many more ways to do it lying around in depreciated modules), and despite having quite a bit of experience with all of them the only one that performs entirely predictably is the humble - albeit the least accurately named? - exposure slider.
quantitatively and visually/perceptively different.
1 Like
I can see that you haven’t read the thread, including that your joke is old – and that this isn’t about the name of the slider, but rather what the term is for the technical attribute that changes when you’re making photos brighter or darker – with whatever slider you do it.
Works on black point shifting it up and down by an “offset” so not like exposure…
Yes, thanks, I discovered that while exposure multiplies a factor to the RGB values, global offset just adds a constant RGB value to all pixel values. RTFM … 
-)
Playing with a gray ramp for those sorts of things and looking at the waveform is a good way to visualize slider effects and the same for color esp using the vectorscope…
Yeah, I downloaded the Sweep_sRGB_Linear_Half_Zip file you helpfully posted some time ago, and has begun to use it for such purposes. Thanks!
1 Like
I think you’re talking to me, I believe made the only joke in the thread…
My somewhat tounge-in-cheek suggestion to change the slider name to ‘exponent’ is a more literal description of the operation, what might be wrong with that? That’s the mathematical operation behind the slider, no? And, what the slider affects is the brightness of the rendition, not the light energy collected at the scene. The implications to rendition noise are markedly different between the two, so why not differentiate?
Well, one has to recognize colloquialisms in order to communicate effectively. ‘exposure’ applied in post is one of them, has nothing to do with how the sensor was exposed to light at the scene, simply a post-manipulation that looks the same in a rendition. But most folk recognize the effect for that name, so we’re stuck with it. To be clear, I"m not going to change the name in rawproc, for this reason. But, the precise communicator in me would like to…
Your thread title, ‘“Exposure” and terms’ does call into question the name…
Even more simple initially you can just examine a crop of a colorchecker as a first start… shown below for the main colors… but you can do the same for tonal sections as well and even put markers on them to track values
No, not to you, Glenn.
(It was black_daveth who came up with the same jocular double meaning of exposure as lhutton had already used in the thread.)
Yes, I have with hindsight realized that my OP could have been more optimally worded, since so many are arguing about the exposure label, which I like you believe
so I had no intention to spend any energy on that, (other than here now expressing that I wonder if “EV adjustment” or something in that direction might have been somewhat better if one want to emphasize the capture related log2 aspect without linking it more than necessary to something outside of the image processor and its data …)
As repeatedly stated my focus is the terms related in general to amending the global light level of an image, (which I mainly do with exposure slider and therefore linked to use of that). You say:
which is in line with the dt manual’s “Increase or decrease the overall brightness of an image”.
Brightness is however a perceptual term – with similar meaning as lightness, I believe - for the subjective impression of some objective value/aspect. At the stage of the exposure module I believe we are in linear RGB space/domain (or something around that) prior to any rendering of an image we can perceive, and the objective change that happens from the computation of the exposure module should then be, from what I’m now able to see, luminance values.
This switch between perceptual terms and objective terms can sometime be quite bewildering when one try to get to grips with image processing. Like the Color Balance RBG using “luminance” labels for some sliders, whereas Color Calibration has its “brightness” tab, both modules being positioned in the scene-referred domain prior to the transform in filmic/sigmoid. This confuses me – although any objective change can only visually be observed from changes in brightness levels on the screen (unless we take a dump of the data) …
1 Like
The other thing to keep in mind is that many people use multiple instances of exposure often with masks to dodge and burn and move it around the pipeline so locking in on it as a single step in the processing from the sensor and trying to define that action might not be the way to look at it. In the end all the modules get an input from the previous one and create an output to pass along. In the case of exposure it will take that input it gets wherever it lands in the processing and apply a gain to it and spit that out… what that means to the image might be different depending on how it is implemented or blended with a blend mode so to me its a gain operation and I don’t really think too much about it other than to tweak it to get the look I am after
I have sought to be clear (lately) that it’s mainly global changes that I have in mind, but the theoretical aspect applies in principal also to partial amendments - and whether it happens in several steps should in my mind not also have any principal impact.
Furthermore, ref also previous post responding to @ggbutcher, when we next put the data in for local adjustments, we employ the Tone Equalizer, which happens to have a selection for “luminance estimator” …
So do I, but I still would like to know how to talk about it with the most technically correct terms …
1 Like
Dang, I thought someone actually read my joke… 
What I’ve found returning to photography after a 30-year abstinence is an almost fanatic attempt to abstract the mechanics to facilitate understanding, sometimes to the detriment of that. Concepts like the exposure triangle… Don’t get me going about ISO sensitivity, a misnomer actually propagated by at least one camera manufacturer…
What keeps me sane is the digging into the actual mechanisms to understand how they work. Call it what one will, I wrote some ‘exposure’ code so I know what it does, Adobe’s highlight rolloff not withstanding…