In the Exposure module we find this main slider, by many regarded as the pillar of image processing, labelled “exposure”. A fairly straight forward mechanism that we all understand.
However, “exposure” is what happened previously in the camera, (the density of light hitting the camera’s sensor, based on the illumination level, shutter speed and aperture value used).
So I wonder what is really the correct term/description of what this slider adjusts?
The exposure module is in the scene-referred part of the pixelpipe, so maybe its luminance (which I believe is described by the Y channel of the CIE XYZ 1931 space) – or value or tone or grayscale value or ?? (We can watch its effect on the screen, but then I guess it has become brightness – or lightness = perceived brightness, dependent on ambient light …)
Can somebody clarify - or perhaps more fittingly, elucidate - this?
(I think there was a discussion on this or something similar a couple of years ago, but can regrettably not retrieve it or remember the conclusion.)
In relation to this I also wonder if there is any difference in principle between what the exposure slider does, and the nature of what the global offsett luminance slider in the four-ways tab of CB RGB does?
There may of course anyhow be a difference in scaling, and since CB functions partly in linear RGB space and perceptual space there may be something related to that, I guess.
[EDIT: How do one compose a post with the category being just “darktable”?]
I guess “exposure correction” would be the precise term. The main function of the module is multiplying the recorded values by a constant (expressed on the \log_2 scale).
“Gain” is an accurate term. We might assume all software operations are digital, so that word is superfluous.
“Exposure” occurs in the camera, not in the post-exposure editing. “Exposure correction” is even worse, as it implies the camera exposure was wrong and can be corrected.
I think accurate terminology helps understanding. But I took part in that skirmish a couple of years ago, and lost.
Yeah, I like one-word names for my tools in rawproc, used to be just “exposure”, but a previous discussion on all that led me to add “compensation” to the term.
Me, I never use it. I take the camera exposure I so carefully crafted and tone-curve it to taste…
I‘m not aware of a whole bunch of issue reports indicating there’s a misunderstanding of the module ‚exposure‘
It’s a quite common abbreviation for the expected behaviour.
Renaming that will demand several changes in documentation and translations - causing effort without a benefit.
But for those who like it, there’s a way to add a pedantic.po translation file …
One, as I always did, could understand it as :
Now that the file is opened and the channel values are not bounded anymore at the defined bound of the original file, it is possible to linear increase or decrease the original value and have the display mapping displaying the result thus simulating chat could have append if the exposure time would have been longer or shorter ? (with the limitation of the SNR/quality of the acquired signal).
I’d like to be corrected if that understanding is flawed though, I’m here to learn
On another hand, exposure compensation is sometime linked in camera manual, if I’m not mistaken, to EV compensation variable…
Exposure is a film development term. You expose your negative film in the camera, and later, in printing, you expose your photographic paper. The exposure slider in editing programs mimics the latter: changing the amount of light in the printed picture, from an already exposed negative (in our case: file).
I am happy with the name. People understand it even if technically it is not the initial exposure. I shoot bracketed exposures all the time and by default DT corrects the three shots to have the same “exposure” within the limitations of being able to correct individual image files by post shooting adjustment.
I did not propose a change neither in module name nor for the slider label.
I am just steadily in doubt as to what is the precise term for the change I make with the slider for amending the tones of an image towards white or black.
The exposure slider isn’t the same as in-camera exposure, but it simulates a change in shutter time quite nicely (w/o the risk of movement blur for larger exposures as an extra)
Exposure is a film development term. You expose your negative film in the camera, and later, in printing, you expose your photographic paper. The exposure slider in editing programs mimics the latter: changing the amount of light in the printed picture, from an already exposed negative (in our case: file).
That “justification” is false. When we print from negative film, what happens when we increase the enlarger exposure? The print will be darker, not lighter.
It would be more accurate to compare with the process of printing from slide film, eg with the Cibachrome process.
There are 3 parameters that determine the image recorded on the sensor for each exposure: aperture, shutter speed, and sensor gain (“ISO”).
It is implicitly understood that the first two cannot be changed in post-processing, so “exposure” in Darktable just affects the last one.
Imagine a new user who is trying to understand a raw processor, and encounters an “exposure” slider. That user is supposed to understand: “Obviously this doesn’t really effect exposure. The aperture and shutter speed have already been determined. The third in-camera exposure parameter is gain, so that must be what this slider affects.” Hmmm.
Even worse: the raw processor might have a tethered facility, which does have a facility to adjust the camera exposure.
I won’t win this skirmish, of course. The term is too deeply embedded. But I empathise with those who find it confusing.
Please, Gentlemen!
I have, once again, in the OP not invited to a “skirmish” about the naming of the exposure module or its exposure slider.
As a non-native English user and one who try to understand the meaning of terms and their relation, I just want to know:
In the sentence “When we increase the value of the exposure slider, we increase the image’s [or maybe each pixel’s] __________ .”, which of the following words should be inserted?:
luminosity
light intensity
lightness
light level
luminous intensity
luminance
brightness
value
tone
grayscale value
or ?
When we use the exposure slider, the image data has been imported into an image processor and already been processed by several of dt’s other modules. Aren’t we then in some kind of “____space” with a dimension that describes light level? (I have therefore consciously not included the suggested “gain” since we are no longer in camera space and I believe gain is something that relates to processes prior to any raw data are formed.)
Should be: “When we increase the value of the exposure slider, we multiply all 3 RGB channels values by a constant”, that’s the maths behind the scene; but probably not the answer you’re waiting for
to a new user it’s highly recommended to read the manual. Especially if they don’t have an basic understanding what a raw processor can be used for …
There there‘ll be no misunderstanding of the purpose of that module.
darktable is focused on those who knows what they are doing. All those green rectangle kiddies aren’t in scope of darktable
and tone-curve it to taste…
