gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
|
 p.4 #7 · Color discrepancies from Lightroom to Flickr, Imgur, Windows Photos, firefox, and chrome/brave browsers. | |
Camperjim wrote:
Dan, your comments clearly summarize some of the issues especially when it comes to the overall goal of matching screen and print appearance.
Of course there is an even bigger issue. Unless we are into major enhancements or digital art, many of us would like the real world colors and visual experience to match what we display on screens and prints. Some photographers think using a color checker will help solve that issue. A color checker might be of use for commercial photography applications where corporate colors need to be accurate. Even then there are better ways. For landscapes and most photography, trying to match what we see and what a camera records is impossible and adjustments in post are frequently a necessity. Camera sensors record differently than our eyes. The even bigger factor is processing of colors. Our brain processing is way different. Two of the big factors have been called color constancy and color relativity. Color constancy means that our brains try to attribute similar colors to objects regardless of the actual wavelengths of light reflected from that object. Purple mountain majesty is a good example. The camera may record intense purples and blues in the shadows of distant mountains. Our brains are likely to try to "correct" those colors to what we believe. That intense purple from the snow in the shadow of the distant mountains is likely to seem more white. The color reflected from the rocks might be more like granite or in the southwest more reddish. Color relatively is also well known. We interpret colors based somewhat on the adjacent colors. There are a lot of optical illusions that demonstrate this.
Anyway back to the issue of gamuts. In an ideal world we would have what grandmas wants. There would be a single gamut that matches what our eyes see, what the camera records, and what can be displayed on monitors and prints. Clearly biology and physics will never allow that. We need gamuts and profiles to try to bring those closer together. That approach still leaves a lot of loose ends. The out of gamut colors are one big issue. In post, we can only see and make adjustments based on what we can see on our monitors and the gamut that applies. If we work in Prophoto and there are a lot of colors that are outside of the gamuts for our monitors and prints, we can end up with undesired results. That is why I have decided to work in Adobe RGB. It is pretty close to the gamut of what my printer produces at least using a coated paper instead of a muted matte paper. It is mostly bigger than what I can see on the monitor but reasonably close. It is rare that I get an image that looks off when I print or compress to sRGB for online display.
I do agree that using ProPhoto for processing rarely causes issues. When it comes to online display, our monitors have already crunched those out of gamut colors in PP so we are making adjustments based on what we can see and that will also be close to sRGB for online display. The use of PP can be an issue for printing. The colors we see on a monitor have been crunched automatically but those colors that are larger are still in the PP file and when we print either relative or perceptual rendering intents may give us final print colors that we not desired. If I am wrong, I would certainly like someone to explain why.
P3 more closely matches the gamut of a newer, quality monitor so it would be a better choice to replace sRGB. As you mentioned it may not be the best choice for printing. The P3 gamut does not include some of the green colors that most inkjets can print. I am not sure if there is any push to replace sRGB with P3. We will see.
This whole topic can become a nightmare with gamuts, rendering intents, and all sorts of other biology and physics considerations. Fortunately, as you pointed out, the impact of our choices often makes little or no visible differnces. Again, to me, how our brains process colors and visual perception is way more important that all of these technical considerations.
...Show more →
Something like 20+ (actually almost 30 now that I think about it!) years ago I had a relationship with Apple’s higher eduction division that put me in touch with a bunch of college/university folks, mostly from the US but also internationally. One of them was a Canadian whose work was specifically about how to deal with these really complicated color issues that we run into with digital color systems and how to represent color accurately on various kinds of output media. Before I spent some time listening to him I had no idea what a can of worms it is!
And, as you remind us in your first paragraph, even when we rationalize color on monitors and printers, the darned human brain is so flexible and malleable and the viewing conditions are so variable that just about everything is in flux. That’s why I’m more about consistency than about accuracy. (I mentioned that in my previous post when referring to profiling, where I Think that, for example, literally matching monitor and printer output is a lost cause but where at least establishing a consistent relationship between the two is extremely important.)
I’ve written stuff like your first paragraph many times — in fact, I often use the “blue mountain shadows” example to point out how subjective our concept of color is and how much it is affected by what our brain knows about (or thinks it knows about) what we are looking at. We “know” that snow is “white,” so when it is actually quite blue we see that blue as being white. If we see something that we “know” is gray in the soft light under a tree full of autumn leaves, we see that rather warm toned thing as being gray… when it isn’t.
Beyond that, as you point out, creating “accurate” color is perhaps the province of things like product photography, where colors must be consistent with a brand. But for things like portrait or landscape (and many other types of) photography, we necessarily strive for something that is subjectively pleasing and very much NOT objectively accurate. One genre I work in quite a bit is landscape photography, and I’m hard pressed to think of any highly regarded photographer in the genre who presents literally accurate color rendition that is faithful to the original capture. That’s not a complaint but the way, not even remotely — it is actually quite the opposite.
By the way, the issues are limited just to color. We also perceive luminosity in similarly subjective and “inaccurate” ways. The friend I mentioned in my previous post likes to illustrate this with a series of visual puzzles that induce us to believe that two patches of identical gray are actually very different — depending on what other luminosity levels surround them. (See some fun examples here and here.)
It is a can of worms, really. Though a pretty rewarding can of worms when it works.
|
