Yes, exactly. They are mostly technical, not creative. Not even necessarily about tricky exposures, just a necessary remapping of raw image latitude to the fixed dynamic range of JPEG files and screens.
2 Likes
Lightroom is a black box and trying to make comparisons of DT and LR is not going to help you. LR gives you a single slider to brighten shadows and a single slider to darken highlights, but how this is achieved is not really controllable by the user. However. DT’s shadow and highlights module allows the user to control this process if they wish.
Do yourself a favour and download the image and work through the well explained lesson in this link
Also, you are tone mapping, but you are choosing tools not intended to be the primary tone mappers and making them do the job rather than selecting one of the tools designed for tone mapping. The tone mapper is just the start of the editing process and is not intended to give the final look alone.
My view on tone mappers is base curve is old school, it works but there are better options. Filmic is very challenging to learn and master and I no longer teach it to my photography students. Sigmoid gives nice colors out of the box, often without the user having to tweak its controls. AgX is for me magic. I mainly use it by activating it and using the auto tune levels to set the whites and the blacks. I also use the picker for pivot target output. Then sometimes I return at the end of my edit to the pivot target output slider and tweak it a little to ensure the final desired brightness for the image.
Most of your questions can really only be answered by doing some playraw posting with images. Posting your edit here is not informative enough to explain why a tone mapper will benefit your processing.
Good luck with your DT journey and its nice to see you asking questions on the forum.
3 Likes
Thanks, although I’m hardly a beginner with darktable, I’ve been using it for more than 6 years and have almost 10,000 fully edited files in my library.
In that time, like a lot of people who transitioned from Lightroom, I’ve experimented with lots of different workflows. More by luck than judgement (as there are so many modules in DT that do more or less the same kind of thing), I stumbled on a simple workflow that seems to get me where I want to be more or less quickly and without fuss.
It’s just when new versions come out and seem to be trying to change things around (like the new AGX module) that I start to get anxious and question the way I do things. Unfortunately, I find I don’t really understand what these tone mappers do, so I end up randomly pushing and pulling sliders and just hoping for the best. In the end, I usually waste a month messing with them, before giving up and turning them off once again. I expect AGX will follow the same path. Not sure why I keep listening to experts who tell me how I should be working when I already have a workflow that actually works. Possibly a case of the grass is always greener or perhaps a lack of confidence. Who knows. Unlike Lightroom which is so simple you can’t really do it wrong, DT has a lot of complexity and a lot of opportunities for making a Royal FU of things 
I should keep away from youtube and I’d be fine as I am
From the photos you’ve posted and the workflow you’ve described, it sounds like a lot of your images have less dynamic range than the output medium, which means you generally just want to stretch your histogram a bit. If that’s the case, there are lots of ways to stretch the histogram without a tone mapper.
In your case, I’d recommend just activating Sigmoid, and not change anything else about your workflow. In most cases, it won’t do anything, since your workflow already takes care of things. But in the rare case where some highlights escape, Sigmoid will reign them in.
1 Like
I’m sorry for my bluntness, but I’m not sure what you want to get out of this topic. You have a process that works for you; it seems to me you don’t want to learn about tone mappers (you could, if you wanted to, there’s documentation and dozens of good videos to learn them). Yes, pushing sliders at random won’t work, AgX has too many of them. Every time someone responds with an explanation, or asks you to upload an image as PlayRaw, you say you just don’t get it, and your process works for you. Which is great, of course, but if you know you don’t want to change a working process, then why ask? Keep using what works.
10 Likes
I find it weird that you start a topic with a question, and when you get nice, helpful, and detailed answers, you respond in the above manner, claiming that people are telling you how you should be doing things.
4 Likes
They are just tone curves. The only special thing about them, besides what I’ be already said about colours, is that the input is not restricted to 0…1.
Essentially, they are like the film in an analogue workflow: they take the extreme brightness range of the real world (as captured by the sensor, of course), and map it to something manageable (on the paper print / display). AgX is named after the silver-halide chemistry of the film era.
With a traditional (display-referred) curve-based workflow, you carefully manage exposure etc to make sure the input does not exceed 1, because that would mean clipping. With the scene-referred tone mapping curves, you tell the curve what the darkest and brightest input tones are that you care about, and they take care of the compression. This allows you to work with very few restrictions up to the point where the tone mapping occurs, keeping mid-tones as mid-tones, and the data in linear, which works better with many algorithms.
1 Like
A tonemapper isn’t a tool to optimize an image. It’s just fitting unlimited tonal range into a limited one.
E.g. if your image has a tonal range of [1…1000] and your target can just handle [1…255] the you need a tonemapper to define how to handle the 745 tonal values that doesn’t fit into the target.
The more users think about a tonemapper just like a tonecurve the more they get trapped expecting stuff a tonemapper isn’t made for…
I’ve never heard of playview, what is it?
I don’t understand this comment. I wasn’t referring to any posts in this thread, which have been useful and I appreciate the time people have given.
I was referring to influencers and commenters on the web always pushing the latest thing and creating anxiety and FOMO.
1 Like
don’t trust influencers without checking if they have fundamental knowledge of the stuff, they are talking about.
Most of those comparing darktable with Lightroom, C1, … don’t have any experience using darktable. And so they are simply using similar sounding tools in darktable and wonder, why they fail.
3 Likes
I thank you for the attempt, but this has just confused me again.
My primary output is intended to be paper prints (and I’m experimenting with my new Reflection Frame which has a mere 1:30 contrast ratio). I prefer matte paper as well, which has a reflectance range of between 1:50 and 1:100.
As you say, in most cases my file is going to have a higher dynamic range, so I am always compressing the image dynamic range into a reduced DR output medium. Always, and this has been the case for the 50 years I’ve been doing photography. We have histograms and clipping warnings to help us manage this.
I’m now back to my original confusion: what are film curves, filmic, sigmoid and AGX actually for and how does this differ from simply adjusting curves? I think my problem is that if these tone mappers didn’t exist, I’d have no idea they were needed. Which tells me I don’t understand what they are for and I am simply forcing myself to use them because some influential people on the web have told me I absolutely have to use them because otherwise darktable is all broken and obsolete. However, I have no practical experience that suggests darktable has ever been broken, it seems to be a wonderful product to me.
So, now I am back to square one:
1 darktable seems perfect to me
2 Influencers tell me I’m mistaken and that is has been broken for years
3 But, thankfully, there are a bunch of new modules and workflows that save the day and I need to switch to them
4 Unfortunately, despite the documentation, Q&A threads and endless videos on the purpose and use of these tools, I still can’t understand a word that is written about what they do, and have no idea when and how to use them, so in desperation I stop using them.
- But the influencer’s words have taken root and I’m left worrying and anxious that I’m doing it wrong.
I sense some frustration in this thread, so, just to reset and explain what I want from this thread:
- List item
I have worked out a simple workflow that seems to me to be every bit as effective as any other raw conversion software I’ve used and deals with every image I’ve thrown at it. I can make nice prints - at least good enough to satisfy the RPS assessors and earn my LRPS.
- List item
I don’t use any of the tone mappers
- List item
Influencers and web gurus are always pushing the latest thing. This applies to darktable, too. AGX as just been released, so the clamour has started again
- List item
I’m pressuring myself from FOMO into thinking I must have been using darktable wrong all along and that I need to up my game and get to grips with tone mapper based workflow
- List item
However, despite much reading and YT watching, I still don’t understand what problems tone mappers solve that isn’t handled perfectly well by non-tone mapper workflows; and I don’t understand what tone mappers actually do, except it sounds like they include some kinds of magical technology that will somehow transform my image quality so I need to master them.
And that is why I started this thread: to get some help understand what tone mappers do, how they work, what is wrong with my current workflow and what problems that has, what image problems tone mappers solve, how to recognise when I need to employ them and how to actually use them to solve problems rather than create them.
I understand this is no small undertaking and I’m asking a lot, but I have an appreciation of the expertise available on this site and a deep respect for it.
ps
The source of my anxieties on this subject go back quite a while: to the videos of Bruce Williams discussing the theories of Aurelien Pierre, exacerbated by Aurelien’s incomprehensible (to me) writings on various technical matters. I’ve been suffering ever since. The arrival of AGX has brought all this back
pps
With respect to paperdigits comments about my images having less dynamic range than than the output medium, I’m bemused. What evidence is there for this? I routinely edit my images so the histogram and clipping indicators are just sliding into or slightly beyond clipped. There’s no room left to stretch my histogram a bit, it’s already stretched to the limits…
Hi Kofa
This post sounds almost like the kind of explanation I was hoping for. Except I don’t understand it. This may mean I have some blank spots in my education that mean almost any explanation is going to be beyond me unless I can fill those blank spots.
You keep repeating this, so I am curious who these “influencers” are.
Can you link some sources?
—continued
What does this section mean? It implies that:
- Digital workflows don’t map tones, only film does
- Without tone mappers, you can’t make prints or screen images
I’m confused. If I switch a tone mapper on and off, it doesn’t usually do anything particularly obvious to the tonality of the image other than sometimes curtailing a clipped highlight or sometimes creating extra contrast and introducing clipping. What is does seems image dependent and apparently random which just proves I don’t understand what it is doing
What does “1” mean in this context?
I don’t understand how this differs from bending a curve to push tones up to the clipping points but no further. What is the tone mapper doing that is different?
Once again, this sounds exactly what a simple tone curve does, I don’t understand what the tone mapper is doing different.
Looking back over my questions here, I feel that there is simply something fundamental about image editing that I have misunderstood and your points are crashing headlong into that misunderstanding. Maybe it would help if someone could probe what I think I understand to work out the root cause of my misapprehensions?
Sure.
Deep writings: Aurelien Pierre
Youtubers: Bruce Williams and Boris, to start. Excellent trainers in the use of darktable but people who encourage certain techniques I don’t understand. Their confidence in espousing the use of tone mappers, for example, very much creates anxiety in me that I need to be using them, that the old ways are broken. I trust these people. I just can’t figure out the why, and the how of using them despite their best efforts.
100%. If mid-grey is at 18%, then perfect diffuse reflection with a bit of glare is 100%. Specular reflections and light sources are even brighter.
You can reduce exposure so you have a tiny bit of detail in your print at parts you care about, but that usually means you make mid-tones dark, which you then raise using a tone curve, and/or you dodge and burn the image to manage highlights and shadows (you can still do the latter, even if you use a tone mapper).
With the scene-referred workflow, mid-grey remains at 18%, diffuse whites are close to 100%, light sources above that, like at 200 or 500%. The tone mapper then takes care of mapping that to at most 100%, also managing colours (in case of AgX, so bright reds turn somewhat orange, but not yellow, skies turn a bit cyan, but not too much).
Ok, trying to parse this…
So, “1” in this context means what I would call the white clipping point. Specular reflections, if small hotspots, would normally clip and be allowed to go to full white.
You are saying that the tone mapper allows you to recover texture and colour in what normally would be these small pure white areas without depressing values lower down the scale, and without using dodging techniques?
How does it do this, it sounds physically impossible…