Re: Niagara Spray: Results
I agree that this is a difficult choice but at best I think it’s tied with how to lighten the face. By my quick survey (hope I’m not missing one) I rate that only ##403, 405, 406, 407, 410, 411, 412, 413, 418, 422, and 423 have acceptable faces. That’s not a particularly good ratio, and I it’s lower than the number of people with acceptable waterfalls.
Ken Harris is right when he says that this shot is not amenable to a global correction. The face must be lightened but this tends to do unacceptable damage to the falls unless they are masked out. Some people tried to blend the red channel into the RGB in Lighten mode, which would lighten the face without doing it to the water, but there is only so far this technique will go. All of the successful people tried something in addition, such as a false profile/multiplication routine.
Even so, the best faces here tended to be paired with poor waterfalls.
I also agree with Ken that this is the direct opposite of the Veiled Bride exercise (which is why we are doing them consecutively). In one the background is so boring that a near-majority of us tried to suppress it in one way or another; in this one the background is so important that the majority tried to enhance it. In the other there was no color to be had almost anywhere, but this one has many opportunities to brighten things up. Ken also makes a good point that I missed, that the bride’s mood is somewhat ambiguous. Given that there are so few perfect husbands in the world, and that I am already spoken for, she is probably asking herself what kind of mess she has gotten herself into. And who knows how she will feel about this shot in ten years? The Niagara image, OTOH, is about pure happiness, nothing ambiguous about it at all.
I disagree about the importance of the image; I suspect that given the context of one’s first six-month trip to America this is going to be one of the most memorable shots, because few things are as recognizably North American as Niagara Falls, and the pose is attractive.
I do agree that the file is full of traps. The slicker is difficult. In real life, we perceive it as bright yellow. The camera sees it as more transparent and picks up a lot of what’s behind it, or makes the whole thing too weak. The result can be a distressing green. And there are these colorful blotches in the background, typical of many iPhone captures in my experience.
Almost nobody solved all the problems posed by this image. The closest to that goal is #423. Everything else has some kind of weakness, even when there are other areas of strength, as Paco pointed out in the description of his #403. This explains why IMHO the par version is decisively superior to any of its parents, which I didn’t think was the case in the last two studies. Here, though, an averaging of the imperfect yields an excellent result. Paco’s waterfall is too blue, but it is only weighted 20% in the par version so we no longer perceive that the color is wrong.
When I first reviewed these entrants I thought that we did a poor job on the whole, because it was easy to eliminate all but half a dozen or so entrants. On further review I see a lot more near misses, and fewer disasters, than I first anticipated. True, there are a few lemons in this assortment, but not as many as there were in Veiled Bride—and this image is technically more difficult.