Still working on peonies

[DanK]

I'm still working on peonies, trying to capture the amazing patterns that are obscured by the uniform color and the density of petals on many of the flowers.

Here's a photo I tried both in color and in B&W. I had B&W in mind because there is no color contrast to speak of, and the lines are really the core of the image. I haven't finished dealing with the imperfections in the petals that cause tiny white dots under artificial light. I've removed more from the B&W image because they are more apparent there.

C&C welcome.

[purplehaze]

I love the b&w, Dan, but I would like it even more if those large black areas were filled with petals.

[billtils]

Dan

I like both of these quite a bit more than the original. It's hard to overlook the gorgeous red of these flowers in general, but the B&W does indeed do exactly what you say re the bringing out the lines in the image.

Very nice indeed.

[Manfred M]

Great example of how a pure B&W wins when it comes to contrast and tonal range over an image that is almost monochromatic.

As others have said, which of the two works better for them is purely personal taste.

[xpatUSA]

I like both shots but ...

Technically speaking, the color image is largely oversaturated (talking HSV space) and that could contribute to obscuration of patterns. For example, in some mid to low tone areas there is only red with the green and blue channels forced down to zero. Thus patternation there is only indicated by shades of pure red. This can occur if one edits in a wide-gamut space then saves as sRGB for posting.

Also there is some clipping causing some purple highlight areas.

HTH

[DanK]

I like both shots but ...

Technically speaking, the color image is largely oversaturated (talking HSV space) and that could contribute to obscuration of patterns. For example, in some mid to low tone areas there is only red with the green and blue channels forced down to zero. Thus patternation there is only indicated by shades of pure red. This can occur if one edits in a wide-gamut space then saves as sRGB for posting.

Also there is some clipping causing some purple highlight areas.

HTH

Interesting. I think this is indeed the conversion to sRGB. looking at it on my screen (I am looking in LR, which is a variant of ProPhoto, and my working space in Photoshop, where I did almost all the edits, is ProPhoto), and I can't find any areas where either B or G is zero, although it does come very close in a few intense areas near the center. I also haven't found any clipping. I hadn't given the magenta areas much thought, but if I go to the petal at the bottom right, which has the strongest magenta cast, I find values like R=38, B=15, G=26. The petals did have a hint of a purple tone, so I am guessing that areas in the shadows showed that more.

[DanK]

thanks, all, for the comments. I think B&W is the way to go. Janis, re the black space: I take your point, but the flower was asymmetrical, which is why the black is there. I decided to leave it alone.

[Manfred M]

Interesting. I think this is indeed the conversion to sRGB. looking at it on my screen (I am looking in LR, which is a variant of ProPhoto, and my working space in Photoshop, where I did almost all the edits, is ProPhoto), and I can't find any areas where either B or G is zero, although it does come very close in a few intense areas near the center. I also haven't found any clipping. I hadn't given the magenta areas much thought, but if I go to the petal at the bottom right, which has the strongest magenta cast, I find values like R=38, B=15, G=26. The petals did have a hint of a purple tone, so I am guessing that areas in the shadows showed that more.

Dan - I've often wondered how much damage is done through what happens between the time we finish our image and upload it to the time that it is displayed. When I look at a saved version of the image, it shows a fairly significant issue in the green channel, as seen in ACR.

[xpatUSA]

Interesting. I think this is indeed the conversion to sRGB. looking at it on my screen (I am looking in LR, which is a variant of ProPhoto, and my working space in Photoshop, where I did almost all the edits, is ProPhoto), and I can't find any areas where either B or G is zero, although it does come very close in a few intense areas near the center. I also haven't found any clipping. I hadn't given the magenta areas much thought, but if I go to the petal at the bottom right, which has the strongest magenta cast, I find values like R=38, B=15, G=26. The petals did have a hint of a purple tone, so I am guessing that areas in the shadows showed that more.

Yes, the effect of color space transformation can be seen easily in RawTherapee because changing the output profile changes the colorpicker values. For example. at position x,y = 235,635 we get percentage RGB values of:

Prophoto: 41.2 16.9 15.7
IEC sRGB: 64.3 0.8 20.0

So the green channel really got pushed down at that position.

[DanK]

This makes sense. This is an image for which conversion to sRGB imposes a cost. In general, I have frequently had difficulties with intensely red flowers. However, I decided to go with B&W anyway and just spent ages removing most of the tiny white spots, before posting it on my website.

[xpatUSA]

Dan - I've often wondered how much damage is done through what happens between the time we finish our image and upload it to the time that it is displayed. When I look at a saved version of the image, it shows a fairly significant issue in the green channel, as seen in ACR.

I think that the "damage" was done during the conversion from ProPhoto to sRGB prior to up-loading.

Green channel extracted in the GIMP:



Not much green at all.

For posting flowers here, perhaps Dan should edit in the sRGB working space, not ProPhoto ...

[DanK]

For posting flowers here, perhaps Dan should edit in the sRGB working space, not ProPhoto

Would that help? I haven't tested doing the identical edits after conversion, as you're suggesting, rather than before.

[Wavelength]

I love the colour one for its beauty(some how it makes me remember a cock's head);i equally like the black and white with a more serious appeal.

[Manfred M]

I think that the "damage" was done during the conversion from ProPhoto to sRGB prior to up-loading.

That is highly unlikely as the software simply maps one RGB colour space to another. I have gone back and checked and did not detect any gamut issues in the past, but that was several versions of software ago. Things could have changed so I'll try that again, but it would surprise me a lot...

Not much green at all.

Obviously not; there is lots of magenta, which is the complement of green.

For posting flowers here, perhaps Dan should edit in the sRGB working space, not ProPhoto ...

That is generally viewed as a poor practice as the colour space is so small. Most recommend either AbobeRGB or ProPhoto RGB (so long as we don't get artifacts by pushing the colour space too hard).

[xpatUSA]

Dan - I've often wondered how much damage is done through what happens between the time we finish our image and upload it to the time that it is displayed.

Could you please explain what is meant? It makes no sense at all that, after an image has been up-loaded, it somehow gets "damaged" before someone looks at it.

[Manfred M]

Could you please explain what is meant? It makes no sense at all that, after an image has been up-loaded, it somehow gets "damaged" before someone looks at it.

It gets uploaded and no one knows what manipulations the hosting site does to the image. Likewise, when we download, I have no idea as to what mechanism my computer uses to capture the file, Is it a screen capture that is saved, is it processed by the operating system, etc.

I am relatively certain of what the data looks like when I save the file on my computer, but once it leaves, it is at the mercy of whatever algorithms it passes through.

[xpatUSA]

For posting flowers here, perhaps Dan should edit in the sRGB working space, not ProPhoto

Would that help? I haven't tested doing the identical edits after conversion, as you're suggesting, rather than before.

It helps me because what I see on my monitor is what anyone else will see on theirs given correct color management. However Manfred has spoken to the contrary, so I suggest that you try it for yourself.

[xpatUSA]

It gets uploaded and no one knows what manipulations the hosting site does to the image. Likewise, when we download, I have no idea as to what mechanism my computer uses to capture the file, Is it a screen capture that is saved, is it processed by the operating system, etc.

I am relatively certain of what the data looks like when I save the file on my computer, but once it leaves, it is at the mercy of whatever algorithms it passes through.

Nothing specific then.

[DanK]

It helps me because what I see on my monitor is what anyone else will see on theirs given correct color management.

That wasn't my point. My question was whether it matters when in the process one converts, that is, would the product have been substantially better if I had converted at the beginning and then applied similar edits? I suppose I could check this some day.

I don't worry too much about getting online displays exactly right. After all, they will appear on all sorts of monitors, calibrated and not calibrated, using browsers with different color management. Even if I get it so that it looks dandy on my NEC, it doesn't look all that good on the uncalibrated Dell sitting next to it. So I am willing to be a little loose when it comes to online posting. When I really fuss is when I print.

[xpatUSA]

Originally Posted by xpatUSA Not much green at all.

Obviously not; there is lots of magenta

Obviously not much green, but there is next to no magenta in the posted image. Between the hues of 290 and 310 deg. there are less than 0.1 percent counts! Even between 280 and 320 deg, still less than 1 percent.