In darktable documentation, when citing academic papers, please don’t just give a link, eg https://doi.org/10.1002/col.20128 . Links might break next week or next year. Please also give the paper name, author(s) and year, eg “Talking about color … brilliance, Rodney L. Heckaman and Mark D. Fairchild, 2005”.
5 Likes
DOI links won’t break, that’s the point.
1 Like
True. Then again, I was always taught that the full citation was the way to go, with urls as optional extra.
For printed goods, sure. But on the web, you are one click away…
Dimension is as technical as “width, height and depth”. That’s the kind of dimensions we are talking about, although you could look at them from an n-dimensional point of view, there is no need to. And these are quantitative properties as well.
Scene-linear is already used everywhere in the manual and defined in the color management section. The physiological approach is not older (the last revision of LMS space dates from 2006), it’s not old vs. new, they have different use for different purposes.
And yet it’s the whole point for color adaptation models.
Somebody else will have to do that 3D graph (expect 8 hours of computation).
Thanks Aurelien,
Give me a couple of days and I think I can propose a better draft for this section of the manual -ie more clarity and hopefully still technically accurate.
I think a lot of problems stem from the use of common English words for technical concepts. Using synonyms instead of the proper technical term, leaves the reader to guess whether something new has been introduced. This is why I questioned the use of “scene-linear” I can’t find that exact word anywhere else in the manual, so I guess it is the same as scene-referred. Better to use exactly the same term, or make a definition on first use of a new term.
Most people will recognise dimension in terms of width, height and depth, but it is a different concept to talk about colours or vision having dimensions. This is really about the colour spaces requiring 3 independent coordinates and how properties of colour can be represented by these coordinates. I agree that your final paragraph in the definitions section is the whole point, but for that reason it needs to be up front, not at the tail end.
More later.
I wonder if scene-linear is really just a case of scene-referred linear being truncated at some point.
When I came to try to understand the world of display vs scene referred/linear and the terms related around linear workflow I think the two sometimes got intermingled.
Quoting some discussion on the ACES website provide a description of the nuance
Scene Referred simply means the image data is maintained in a format that as closely as possible represents the original scene, without effective restriction on colour or dynamic range. This is not necessarily the same as the ‘raw’ image data as exported from the camera (after any necessary debayering, etc), but attempts to ‘correct’ the image to better match the scene the camera was originally pointing at, which may include white point correction, gamut correction, etc. Theses processes are often referred to as ‘Scene Reconstruction’ processes.
Scene Referred is also often in Linear Light, which while suitable for computer graphic rendering, is not suitable for grading workflows.
The process used to get images into Scene Referred space is to effectively ‘undo’ the ‘Capture Referred’, or ‘Camera Referred’ image, and reverse engineer it back into Scene Referred space. The theory being any camera pointed at the same scene would generate the ‘same’ image in ‘Scene Referred’ space, within the limitations of the capturing camera’s imaging capabilities.
So to me scene-referred is really working towards the scene-referred “reconstruction” process and in doing so if by using linear light/methods then it is scene linear…
Maybe that is not correct either but it makes sense
4 Likes
Indeed, scene-linear means proportional to the light emission but doesn’t say anything about bounds. Scene-referred is unbounded scene-linear. In the context of this doc page, I use “scene-linear” because we are indeed bounded to display white.
2 Likes
Yay! Contrast and his fulcrum is back again! 
I only wish that the slider is not limited to 50%. I am pretty contrast hungry. I know you can type in the higher values, but I’m lazy. 
9 Likes
I’ve raised a more general issue about exceeding soft limits on sliders here. The soft limits are generally about having a reasonable default UI but I agree it should be easier to exceed them.
1 Like
Could you save an autopreset after adding a manual bump. In most cases even when you remove the value the new max is adopted …so if you can set 100% and then go back to 0…it will leave the max at 100 or whatever you can raise it to…I will have to build and try it as my build is a couple of weeks old…
I know. That works too, but it’s still a detour. Loading preset or typing number is the same effort.
Actually, the whole thing doesn’t bother me that much. It’s just a habit I have with the old color balance. I have used contrast function extensively.
Color balance was - and now even more so - a replacement for a lot of other modules for me. I find it to be a very powerful tool with which I can do 80% of the work.
That why I suggested autoapplied preset….this would essentially override the defaults with a new set of defaults so no extra work……
The auto preset works fine in my limited testing…I went through all the sliders and set them higher and then set the values back to zero for the effect. I takes a min or so to go through all of them…Then save this as an auto preset and you have new defaults for any and or all sliders in the module…
4 Likes
This is really good news I just wish they were located together and not on separate tabs…
Yes, I would have put it together too and put both in “4 ways” tab. There it makes more sense, because there is luminance there anyway.
The module has really grown in length so masking is a long way down and when open masking it really makes the module a bit less friendly as it gets even longer. I wonder if its not hard about collapsing the high, mid, shadows leaving only global option initially exposed and then if you need to use the tonal ranges uncollapse them as needed. I like access to all those adjustments its just a bit of a beast now and contrast and the fulcrum for contrast would be nice if they were together as before. Just my 2 cents
We found some ways to reduce the size of the original color balance using a tabbed interface. Unfortunately color balance rgb is already using a tabbed interface and I’m not sure whether tabs-within-tabs is a good way to go. Individually collapsible sections might work I suppose, if the module remembered the last layout.
Ya I was just thinking out loud. Its nice that the new mask indicator also shows the mask and it might be nice to further leverage that in some way in the future, ie if you ctrl click it or something that you are taken to the mask setting dialogue or some type of integration that way. ON1 photo also access and present their mask settings from the top when activated so that you don’t have to scroll all the way down. I think all the DT tools are great for sure but I think organizing access to them could still be improved a bit. Its hard for sure when you offer access to almost every aspect of the edit there is a lot to manage
Not making any criticism just for reference this this is an ON1 module just to show what I was talking about…so you get a mini mask thumb and you can toggle the options quickly on and off without scrolling the module. I’m not sure to what extent the DT UI could even accomodate anything like that but fuel for thought
Clicking the mask icon to the left of the name toggles the mask options at the top of the module off and on to reveal this
And blending options are toggled by the gear
