And thatās not what I was trying to say either : ) It probably is what most theme designers are going for, but the UI itself ālooking pleasingā isnāt the question weāre discussing hereā¦
So this is exactly what I meant too:
itās about helping the worker assess the colour theyāre seeing
And I donāt think weāre coming at it from very different angles, but I do think our starting assumptions are a bit different.
Colour grades are, somewhat by definition, fairly arbitrary things with a wide range of artistic discretion latitude.
If all we wanted was āperfect photorealismā then that job is easy, and professional colourists are entirely out of a job. You calibrate your camera, and you calibrate your display. And thatās it, all done aside from drinks at the wrap party.
Or if your camera isnāt perfect, you fix it mechanically with an algorithm. Again, no judgement needed - there is One True Answer and you just need to fudge the camera output to correct its entirely measurable error. This is just standard fare for any precision instrument, so itās a well understood, completely solved, problem that needs no human judgement at all.
But none of that is what Colour Grading is about. Itās about creating a look, and a feeling, that often isnāt about perfect reproduction, itās about creating a mood.
It may have started with trying to fix the imperfections in early generations of print and photographic processes - but with modern digital imaging it can often be as much about trying to reproduce those limitations for visual impact as it is fixing real flaws in the raw footage.
A good colour grade is one that ālooks pleasingā to the viewer and communicates the mood the artist is trying to convey. And it may not be (and very often deliberately isnāt) even remotely like the original real-life colours in the original subject material.
Nobody would be enticed to buy a fast food hamburger that looks the colour of the mangy thing you see when you unwrap what they actually sold you. They want your mind to see the carefully coiffed false-colour thing they put in the advertising picture.
And it works because the human mind is really good at seeing what it expects to see in place of what it is actually there.
So I think where we differ on this, is that you are starting with the somewhat axiomatic assumption that a working environment with no intrinsic colour (sat=0, brightness=??) will help a Colourist be most effective at practicing their art. While Iām really not so sure about that at all, and think there is evidence in a lot of phenomenon that deliberately biasing the mood of the Colourist, in the right way, would actually help them more easily create the mood in the image that they are aiming for.
And if that hypothesis is true (and I do think some flavour of it is more likely to be correct than not), then there is no One True background colour for all colouring jobs, and the best colour to bias the working environment with will entirely depend on the subject matter being coloured and the mood the artist wants to colour it with.
Why use mid grey then? Maybe because itās a middle ground optimized for no one, but it doesnāt seriously disturb anyone either.
So Iām not sure I believe this conclusion is true either. Iād agree with you that āpureā white or black are among the safest bets in the general case, if you want a sterile mood that is least likely to distort your judgement in unintended ways - but mid-grey is itself definitely a Mood that biases your vision. And a sterile mood might not always be the best way to āhelp the worker assess the best choice of colourā either.
Yes, you can draw a line between them and build a perfectly logical argument around that - but vision isnāt a trivially logical thing. That strong colour saturation biases our vision is well known to be true, but that doesnāt automatically mean the optimum background colour is one with absolutely no bias at all.
Itās a bit like āIf I have 2 apples and give you 3 apples, I have -1 applesā. The basic math is true, but it has no relation to the real facts of how many apples I can give you.
The main reason not to use pure white or black as a UI background is to avoid eye strain, by backing off from those extremes just a little we actually improve what people see and reduce the fatigue of staring at it for long periods. And since we see colour non-linearly, there are likely advantages to slightly biasing that background level tooā¦
So do I think some background colours are better than others for this? Absolutely.
Do I think the current tiny bias toward blue in the current dark theme colour is actively harmful for āaccurateā colour grading? Very unlikely, though I totally agree that a much stronger bias could be.
Do I think there might be better colours than the current theme which we could use? Thatās probably almost certain - but my suspicion is that this is a spectrum, not a single truth, and the best colour will be a complex function of use and user, and that it will not have saturation=0, the hue may be variable, and the brightness will lie near but not at one or the other of the extremes.
Which means I do support the idea of being able to change the ambient colour (even though most people for whom this really matters wonāt have any of that colour visible on the monitor they are viewing the footage on anyway). But I donāt think ājust pick some shade of grey, any of them will be better than what we currently haveā has very strong legs to stand on. There are probably more shades of grey that would be worse than that would be better, which isnāt good odds for choosing at random.
not having the perfect solution doesnāt mean that what we have is worthless
I donāt think we need a āperfectā answer here. As that swatch shows, being in the right ballpark gets you most of the way there - and if that colour is āwrongā, then itās more likely to be because itās ātoo darkā than because itās ātoo blueā.
The problem we have with this, is that we donāt actually know which direction More Perfect is from where we currently are. We could guess - but that would be a bit antithetical to claiming that we did this to improve things for users.
I would read the paper of anyone who does it, though.
Iād be a bit surprised if there isnāt at least some relevant prior art already published in some corner of extreme colour-wonk space. The optical illusion people are starting to do some really clever things they arenāt just stumbling on by accident anymore.
sorry it took me so long to answer
No sweat there : )
Weāve gone this long without anyone raising this as a major usability problem, and its definitely a valid thing to be thinking about, and easy to agree that itās something we should be mindful of - but that still leaves us a few steps short of having some concrete action we can implement that we know will make a useful difference to users.
Which I think means it is worth keeping our options open until we get some better explanations of what really is known about state of the art best practice for this. I know there are many people who know much more about this than I do - I just know enough to know that, like most things, Itās More Complicated Than Most People Think.
So my hunch is the Ultimate Answer will not be a āmake it more greyā switch - it will be more like a Choose The Mood I Want To Edit This With thing.
Iāll write back when I have more time!
Enough for my brain to know that thereās something wrong, but not enough to really solve the ambiguity