I don’t think switching around the mask & bluring order makes this the same module. I personally don’t like the feel of the nodes in the advanced tab, they don’t pull on each other enough to give a smooth transition. I also don’t care at all about auto scaling the histogram, as I target specific parts of the tonal range of my image and often push the histogram off to one side to ensure that I’m not effecting specific parts of the tonal range.
If this was just moving the masking sliders to another tab, fine, I can live with that, but it is a lot more than that.
The reluctance is because we are speaking of a module very similar as it is almost the same code, with the same goal and a pretty close UI. So as a new module is just no to me.
Fair enough, and I’m fine with that decision in principle. It just seems a shame that if this new version ever makes it to master, it will still be a polarizing module. I know you can’t please everyone, but this module has always elicited strong opinions, and it looks like it will continue to do so.
My two cents just following the conversation for a while…
Is the current tone equalizer module actively maintained? I don’t know if there’s any routine changes that have to take place for new versions of DT or not but it seems to be otherwise in a mostly static state.
Many people seem to love the module as-is and that’s an important thing to keep in mind.
Many people seem to dislike the module as-is and find it confusing and difficult to use.
I get the sense that if maintaining legacy edits wasn’t a concern, there’s more changes that could happen and the “new” module could be even more different.
Understandably adding a new module is a big consideration but it seems like this is a case where it could possibly be worth doing. I don’t have any stake in this particular module. I use it sparingly for specific corrections and even though I don’t like the UI, I can deal with it.
Maintainability and compatibility are certainly important, but using that as the primary goal can stifle creativity in new ways of doing things. How different would a new module need to be for it to be considered more appropriate? Personally I think workflow, UI, and ease-of-use concerns are just as valid and important as technical concerns. We can see that expressed in the last several posts where on one hand these changes don’t fundamentally change the technical side of the module, but do fundamentally change how someone like Mica would use the module.
I’m not even sure if it’s “many people” who love it as is. But those who do love it seem to use it a lot, and it’s a fundamental part of their workflow. So I can sympathize with them if the module changes to something they like less.
I’m in the camp of those who don’t use it very often, so I don’t feel I can push my opinions too hard. On the other hand, when I think about why I don’t use it that much, it’s because I don’t like using it. On paper, it’s a module I would use all the time, but in practice I find it fiddly, imprecise and generally a bit frustrating. So I’m very intrigued to see if I prefer the redesign and whether I will use it more often.
I have been following this discussion here and Re: [darktable-org/darktable] Tone equalizer 2025-04-06 preview version (PR #18656) on github.
The only problems I have with the TE module as it is at the moment is that the magic wands don’t automatically adjust the histogram correctly and I have to jump between tabs to see the histogram when manually moving the sliders.
It’s current position is under the ‘advanced’ section, not the ‘simple’, because it acts on the nodes in the graph editor. So in your new layout, it makes sense to place it in the ‘main’ tab, since that is where most of our work will be done working in the graph editor. It would be highly annoying to have it only in the ‘curve’ tab (though you could have it in both tabs).
The option to manually scale the histogram is still available, but I guess for those who use the ‘simple’ layout it’s now 3 tabs they might switch between instead of 2, which does make things worse for them.
Perhaps the best solution is to have 2 tabs instead of 3? One tab with the ‘simple’ curve controls, and the other tab with the ‘advanced’ - or everything else. In vertical space, that would make it slightly larger than DoS, and slightly smaller than Agx (when ‘look’ and ‘advanced’ sections are both collapsed, and ‘show curve’ expanded).
I think I’ve mentioned this before, but I addressed that problem by assigning the up/down and left/right arrow keys to adjust the TE mask, so I can observe the histogram as I adjust. It also helps to enable to mask display to get an idea of how TE interprets the light and dark zones of the image
Don’t know if I’m to late for feedback, but here we go…
The tone equalizer is one of the most useful, but at the same time frustrating, modules in darktable.
One of my pet peeves is the extremely limited latitude . +2 EV to -2 EV is far from enough to tame a high DR scene. So almost every time I use it for this purpose I have to stack multiple instances of the module. Even with my extreme take on @s7habo’s linear shadow brightening preset shown below i have to stack two modules.
It seems this could be easily fixed without breaking older edits.
The limited latitude is unlike any module in darktable; most modules make make large shifts with very small adjustments .
The fact that even the slightest adjustment with high curve smoothing values totally flips the curve seems less than optimal too.
I think the mask exposure compensation and mask contrast compensation should be easily accessible from the advanced and simple tabs as it is usually a must adjust these. For me it means there’s a constant flicking between the “advanced” and “masking” tab to make adjustment to make the histogram fit within the boundaries of the “histogram space”.
I almost always use the tone equalizer before the exposure module as I find it extremely frustrating that there has to be adjustments done to my stack of tone equalizer modules every time I adjust the exposure. For my use case the default position would be before the exposure module and the “mask exposure compensation” and “mask contrast compensation” would be automatically adjusted to fit the histogram. I guess the automatic adjustment would be hard to achieve as it hasn’t been done.
Sorry if my pet peeves have already been mentioned in this thread!
And in addition, if you have one tonal range where you need to apply such a large adjustment, you can also use the exposure module with a mask.
Simply put, if you use a tool for things it’s not designed for (or in a way it’s not designed for, as in adjusting one node to the extreme of the range), expect things to break. I’d like to see the resulting image if you could limit the change to one node… very often, I get artifacts if I push one section too far using the scroll wheel on the image (affecting at least 3 nodes in the advanced view).
Do you mean you place the exposure module on top of the tone equaliser modules, or do you do the adjustments in the order first tone equaliser, then exposure?
If the latter, you’d still have to adjust the tone equalisers on any exposure adjustment…I find that I need very little adjustment to the exposure module later in the process, if I pay attention to it in the beginning. And if I need small changes , and have the tone equaliser active, those can be done in the tone equalliser.
If you don’t use the guided filter (preserve details = no), such as when you want to add contrast, you should in theory be able to create quite a steep curve without introducing artifacts. But the underlying curve math means you tend to get annoying oscillations with steep gradients, so it’s rare that you can just move 1 or 2 nodes and get anything more than a subtle change.
But I think your first sentence gets to the crux of the frustration for many users: it’s not designed to work in the way users hope/expect. The “equalizer” style of the module seems to promise a lot of control over 8 zones, but in reality it’s designed for adjusting about 3 zones max (because every adjustment involves at least 3 nodes), and adjustments need to be fairly gentle.
The old Zone System, which TE ostensibly replaced, did allow 8 zones to be adjusted independently. So, I think there’s a mismatch between what users expect from the module vs how it was designed.
Yes, of course, but humans don’t always act rationally. In an ideal world, everyone reads the manual and all associated content and then perfectly understands the module… problem solved. But we know that this doesn’t always happen.
And even if everyone does read the manual and understands how TE is designed, it won’t stop some people from not liking the way it’s designed. I stand by my point that many would like this module to work differently and/or have another module available that offers more/different control. I think we can all agree that this is a polarizing module, and all the explanatory materials in the world aren’t solving that problem.
That’s the goal of all the discussion here.
