Re: Chapter 12 in the Chevreul Book

Dan Margulis

On Apr 27, 2020, at 12:37 PM, Gerald Bakker <gc.bakker@...> wrote:

Interesting question. According to the text, figure 12.13B is an asymmetrical move from 12.13A. It's edge enhancement (Hiraloam color indeed), the same type that made 12.10B from 12.7A. However, I looked closely at 12.13B and it's hard to find an area that clearly shows that an asymmetrical method was used. All blues seem to have been enhanced equally, the same for all reds, greens and pinks. It would be interesting to see the result of a symmetrical move, e.g. a color boost application on 12.13A and compare with the given 12.13B.

The result would not be attractive. A setting that would get flowers as bright as in Figure 12.B would make the face quite sunburned. It would make the dress more of a peach color, and the illustration in the book she is holding would become a fiery orange. 

Sharpening, like other forms of correction, isn’t supposed to call attention to itself. The viewer hopefully sees that the image is better, but does not know how it was accomplished. So we hope that is is difficult to detect, even by experts.

Looking at the before-and-after, I’d say it’s clear that the Color Boost action wasn’t used, because that would have boosted colors everywhere. Here, the flowers are obviously more saturated in the corrected version, but the dress seems if anything less so, and that’s possibly true of the face as well.

As for whether it’s symmetrical: the definition of that is, if a certain color changes in a certain way in a certain location, then if the same color appears anywhere else in the picture it would change in the same way. I agree that the cool colors don’t indicate otherwise, but looking at the warm ones does show difference by location, e.g.

1) In the original the lips are similarly colored to the red flowers off her right ear, but in the corrected version they are darker and richer.

2) The face starts out similar to the pinkish flowers at image center left, but the fleshtone moves in various directions depending on what it’s close to. The “mustache” area, sandwiched bertween the red lips and the dark nostrils, gets lighter and more neutral. The reddish eye shadow, influenced by the blue eyes, gets darker and redder. Nothing like this is happening in the flowers.

3) The auburn-orange hair starts out as the same color as orange flowers to the left and right of the head. But the hair is affected by other flowers that merge into it, and changes color fairly drastically in certain areas but  not others. Nothing like this happens in the flowers.

4) The reddish illustration in the book starts out similar to the arcs (?) surrounding the flowers at left. The correction makes it darker but the hue doesn’t change much if at all. The arcs get lighter and perhaps more magenta.

This particular image, though not a photograph, reinforces a good point about them. Look for objects that can be identified by their color alone, and if you find them, boost that color. Things like the face, the dress, the hair, and the book don’t qualify: you need to see their shape to udentify what they are. But some of these flowers are so small that they couldn’t be identified as such in a grayscale context. It’s their color that tells us what they are. A case like that is where we need hiraloam color.

Dan Margulis

Join to automatically receive all group messages.