True. This is easy on exporting because you have time to wait for the full-size export to complete and then to downsize after, and if that takes 20 seconds, we can probably put up with that. However, the darkroom processes the image on-the-fly and in order to do that in a fairly fast interactive way it downscales the image first and then only processes the visible (downscaled) pixels. This doesn’t really impact overall exposure/contrast but it will impact any module that processes a pixel based on the properties of its neighbours (like sharpening and local contrast operations).
Not sure what you’re trying to say here. The halos are better (as in less bad)? Or you can see them more clearly? Maybe we’re talking cross-purposes
@gpagnon
I created a PR on github to add a pop-up window to have a convenient way to change the border width and ratio config variables, see screenshot:
Here is another update on my PR after some doubts by other contributors:
The mentioned standard (ISO12464) didn’t even exist, it was a typo and probably ISO12646 was meant. The author of the “colour assessment” feature wrote that he did not even have a copy of the referenced standard but got his main inspiration by a slidedeck of a presentation.
Another contributor - who has access to relevant ISO standards - checked them for references to this specific colour assessment schema, but couldn’t find any, so the actual measurements on how thick the grey and white frame should be are based on personal judgement and eye-balling.
In light of those findings the other contributors agreed on continuing with the effort on making the colour assessment conditions configurable with this pop-up window and getting rid of the ISO reference all-together to not cause more confusion about it.
The question was raised whether we want to still stick to the schema of having two values:
total border width (including grey and white part) as measured in cm
white border ratio
In my proposal I kept those two parameter, however I made the total border width relative to the screen size, which should scale well with different monitor sizes and physical resolutions (high DPI screens). I kept the 2 parameters because it is very convenient and fast to increase the white border width to quickly judge whether pictures are missing peak highlights (slight underexposured), this is way faster than switching to a bright / white theme. If the white border is too thin this might work well for judging overall colour and brightness balance, but not the highlights, IMHO.
Here is how the pop-up window looks like currently:
It was for prints not on screen but the prints were to be the same size and with a 9mm border so take any of that for what it might be worth…
I guess is ISO12646 which is D50 to match the pipeline and then I guess our display profile corrects to native or D65 however your monitor is set??
That iso is also the 80cd/m2 but some monitors are calibrated at 120 or even 160 in some cases… I have no idea if any of that impact the usefulness, probably not a lot I guess?
@da-phil I like all of your suggestions here. I presume once a user sets the width and ratios and it remains for future sessions. I personally currently use the frame module to put a white border around the image for assessment so your idea appeals to me. I doubt many if any users know what ISO126646 is anyway so it probably adds to confusion.
I presume once a user sets the width and ratios and it remains for future sessions.
Of course, these are config settings which are part of the darktablerc file, which was already discussed further up in the thread.
I doubt many if any users know what ISO126646 is anyway so it probably adds to confusion.
Yes, and for those who work in the media and print business who know those standards get confused as well, because they might have different expectations, except for only having a grey and white frame around the image.
FYI, my PR with the switch to relative scaling for the total border width and the pop-up window for parameterization was merged and should be usable in the next release (5.2).
I wonder, when we shall set the exposure value, which should anchor around middle grey, what is then the effect of having a white part of the assessment frame?
I would presume that it influences us to assess grey tones to be darker than what we would do without it.
Maybe that doesn’t matter since what we shall do in this assessment is not to judge the absolute level of mid-tones in the image, but to compare them with the grey part of the frame - the perception of the latter likely also impacted by the white frame in the same manner as the perception of the image, so that relative judgements come out the same.
We can with the new design eliminate the white frame under exposure assessment by setting the white border ration to 0 %. But we need the white frame for assessment of contrast and of colors, so it could be a lot back and forth with this slider.
I still wonder if it could be possible to have a menu under the button that first let us decide if we want a grey border only, or a combined one?
I know how it’s supposed to work and I know it’s there but really I rarely if ever use this feature in DT …do people really check all their edits with this…or just occasionally…
Actually it helps me a lot to balance out the overall brightness and color distribution of the image. The white frame also allows me to easily compare my highlights with pure white and makes it easier to judge whether they are “right”.
For me, the histogram or the RGB parade is only a tool for a technical assessment whether tones are over- or underexposed and fit into the gamut. But those tools are not suitable for aesthetic considerations or to judge whether an image will work for presentation as a print or on my website.
Thanks @da-phil your update has been incorporated into the latest windows weekly build of DT 5.1 and perfectly suits my needs to help judge exposure and contrast while editing. Another small but good improvement to DT.
I have this on per default when editing as it provides a perfect reference when it comes to correctly balancing the whole image with respect to exposure and color. So a nice feature for me
I will join the chorus of those who use it a lot. I sometimes just have it on by default because I much prefer to edit with it on. It’s a constant reference for my exposure and colour grading.
I would go as far as saying I can tell who uses it and who doesn’t from the Play Raw category…
Am I correct in thinking that it’s giving you 100 percent display white for your monitor…so it likely good overall between users and should help also to be consistent between your own photos but if your monitor or mine is setup darker or lighter won’t there still be that discrepancy between users even if they use it…
Shouldn’t that really be the effect of using the color assessment function, that we have something to relate to as a reference for what is white?
Then what has been edited as white, will be white (=the current maximum on the screen) however white that “white” actually is.
