The priorities are mutually exclusive: delegating scene data conversion to in-camera engine grants you the most simplicity and ease of use at the expense of control; the territory of technical details grants you the most ability to make the photos looks the way you want at the expense of simplicity. You dial one up, you dial the other down.
For example, your choice of demosaicing method can make a tangible difference in finer details: some methods would make them less noisy (better for some styles), others would better preserve finer details (better for other styles). Abstracting it behind one “more detail—less detail” slider isn’t going to work because “detail” can mean a multitude of things, of which sometimes you want one and not the other, and inventing new sliders with user-friendly but inscrutable labels a la “brilliance”, “texture”, and so on, can only get you so far.
There are shades between simplicity vs. control, of course, and so I am curious to know the answer from the horse’s mouth so to speak: to what end they choose to compromise simplicity.
It’s a case of diminishing returns. Shooting raw is a huge and obvious improvement if you want to post-process in almost any way. Conditional on that the workflow should be as smooth and simple as possible. Abstract controls like "clarity" are fine if the result of adjusting them is tangible and almost always does what you want. Giving the user lots of knobs that hardly have a visible effect (let alone a desired effect) is not an improvement.
Almost no professional photographer will care about the intricacies of the demosaicing algorithm, or the choice between a dozen different denoising modules, and Lightroom is entirely correct in not giving you a zillion knobs to adjust things that have no effect on image quality except in the rarest of cases. In 99% of cases the controls that matter are:
* Basic exposure/shadows/contrast etc
* Curves/levels for more control if needed
* White balance
* Cropping, obviously
* Cloning/healing brush
* Simple knobs for sharpening and NR
* Level/perspective adjustment
* Lens aberration correction (most of the time no manual input needed if the lens is in the batabase)
> Shooting raw is a huhe and obvious improvement if you want to post-process in almost any way.
See, you are saying “want to post-process”, which to me says that there is a different priority present rather than just “simplicity and ease of use”.
If the priority is “making the photos look the way you want them to look”, then we are in a territory where it is not as simple as “this tool is easy to use and therefore a better choice than that tool”.
You can want post-processing, but also don't want to spend 50 hours to learn a tool. Sometimes you just want "make it look close to the in camera jpeg, but let me adjust to exposure"
It’s not binary is the point and the whole reason I asked the original commenter why put premium into simplicity and ease and then immediately violate that by shooting raw.
When is the priority not “making the photos look the way you want them to look” ?
When you say “simplicity and ease of use” I think that includes taking the photos, and if I can defer some decisions that might make the overall process simpler.
I'm not the person you're responding to, but in my hobby experience it's sometimes the difference between a photo being fine or even great, and being completely unusable.
If the white balance is set wrong in-camera, then the JPG just came out all blue. It's effectively a black and white photo (albeit in shades of blue), and there's nothing to be done about it. Shot in RAW, the photo can be made color again, extremely easily and quickly.
In fact it gets worse, not better, if you on the day try to adjust the white balance, as you go from outdoors to indoors. Not to mention if you change from flash and back. Auto is safer, but when it's wrong, the photo is unusable, and the moment is gone.
But my DSLR is now over a decade old. Maybe "auto" has gotten much better?
So yeah, for me the main thing is to be able to post facto adjust white balance, which JPG does not support. (if you've done it with both JPG and RAW, you know what I mean when I say "does not support")
Right, I suppose shooting raw is good because you need to think less about your settings at shooting time.
I will say that “auto” is pretty decent on the phones in most common lighting scenarios like sunlight/shade/outdoor/tungsten/fluorescent—white point is an entirely subjective thing that cannot be reliably determined automatically, so in my experience you rarely get the correct rendition of, say, bright pink clouds at sunset, or a book with pink pages (the phone would think it must be the some weird lighting that should be corrected for, because obviously a book can only have nearly white pages, right?), etc.—but due to physical limitations of sensor size and inferior optics the phone is worse than even a decade-old APS-C DSLR in most regards overall.
I do think auto white balance is a little better these days. My old DSLR often had a strong magenta tint magenta in scenes with a lot of green (like forests). My new mirrorless camera from the same manufacturer no longer does that.
> The priorities are mutually exclusive: delegating scene data conversion to in-camera engine grants you the most simplicity and ease of use at the expense of control; the territory of technical details grants you the most ability to make the photos looks the way you want at the expense of simplicity.
The camera makes all those decisions even when shooting raw -- and there are stored in the raw file. So, by default, processing a raw file witout doing any tweaks will get you the jpeg you would have gotten.
My camera (Nikon) -- and I assume the others -- will even store both the RAW and the JPEG, so you don't even have to go through the automatic conversion step if you don't want to.
For example, your choice of demosaicing method can make a tangible difference in finer details: some methods would make them less noisy (better for some styles), others would better preserve finer details (better for other styles). Abstracting it behind one “more detail—less detail” slider isn’t going to work because “detail” can mean a multitude of things, of which sometimes you want one and not the other, and inventing new sliders with user-friendly but inscrutable labels a la “brilliance”, “texture”, and so on, can only get you so far.
There are shades between simplicity vs. control, of course, and so I am curious to know the answer from the horse’s mouth so to speak: to what end they choose to compromise simplicity.