Ah, so you’d map your subject to middle grey in any case and then move it where you want it with tools later in the pipeline? Does that mean you use the color picker to adjust exposure first, to get the subject to middle grey then filmic to define the top and bottom end of the histogram, and then a third one to move the subject where you actually want it (in the cases where that’s not middle grey)?
That’s also what I’d like to use it for. Yet, I find that many changes I make in other places (or even in filmic itself) mean that I have to reiterate over filmic’s settings. And I think this could work better if it was possible to input the parameters which control filmic in a different way.
Example: If I adjust filmic for middle grey, then reduce overall brightness of my subject, I may find that the shadows are now a little too crushed. To counter that, I can adjust the shadow range in filmic, but then there’s a chance that the mapping curve overhoots (which visually crushes the shadows even more than before, which is counter-intuitive), so I also have to change the extent of the linear range in filmic or the contrast, but that affects highlights as well as shadows, so a bunch of other adjustments I made will need revisiting.
Of course, there will always be modules which require readjusting other modules, and once you have an established routine for dealing with your images, you can probably deal with it. But if I’m experimenting, or new to filmic, that is going to be a lot harder. Being able to set input and output for “(not-)middle grey” (dangit, this needs a name), independently of the upper and lower bounds can reduce the need for that iteration, and also make the whole thing more intuitive, because you’d no longer have to know from some external source that exposure must be set in order for filmic to work correctly – you could fire up DT, enable filmic, map the bottom, middle and top of the range and have a fairly decent result already.
I’m not trying to persuade you that your way of using DT was wrong of inferior(*), but I think with a module that uses an approach as different from what came before as filmic, it’d be a very weird coincidence that the first parameterisation/control layout/user interface are the best possible versions. Power to everyone who got it the first time around (and of course to the person who came up with it!) but I think filmic can achieve a lot more than it already has if we actively look for ways to improve it. I don’t think it needs to be dumbed down to do that, either. There’s no need to break anyone’s workflow in order to permit different workflows or come up with visually more intuitive, more robust or just plain different ways of controlling the module.
(*)or that filmic was somehow bad – if I thought that I just wouldn’t use it and walk away. I’m mostly working with RawTherapee these days, but filmic and the colour calibration tool (and the things they can do, as demonstrated by Pierre) are the reasons I’m using DT a bit more again. If I found a way to make them work as fast as my current process in RT, I’d be switching over for most of my stuff, nevermind my muscle memory and established routines. So, really, I’m somehow trying to contribute some useful suggestions to filmic to try and help it become more useful for myself and others.