Adirondacks: Results


Dan Margulis
 

I’ve posted the results of the Adirondacks study, the penultimate in a series of 11 studies beginning with one provided by Roberto Tartaglione and continuing with ten chosen by me.

Reviewing: This particular scene was photographed in fall 2018. You are given to understand that unusual climate in the summer months of the year created conditions for an unprecedented display of autumn color, and you are to assume that this will be noted in any caption to your work. It’s hard to get a bad-looking result from this original but it isn’t easy to get an excellent one. And there is a bit of room for artistic judgment.

We have 21 entrants. Three others, including two that came in the last 12 hours, were disqualified because they had been downsized, presumably by the person’s e-mail app. Most people also submitted a list of their steps, thanks very much. I haven’t read these, because I’d rather get a sense of who was successful and who wasn’t before investigating why. 

The files don’t have people’s names on them, and were random-generator numbered from #1101 to #1121. As with past studies, we also have a “par” version, #1122. To get it, I chose what I thought looked like the five best entrants, and averaged them, each one weighted 20%. This often creates a version that is superior to most if not all of its parents.

I’ll have some things to say about this assortment, but as usual I’d like to open it up to group discussion first. What do the successful versions have in common? Meanwhile, if you’d like to know how your own version stacked up, download the par version and compare the two directly. Do you think you got the same kind of quality? If not, I hope you’ll find further discussion useful.

The folder is in the group Photos section, named Case Study: Adirondacks

I also have zipped all 22 files and uploaded a 50 mb file to our Files section,
Search for Adirondacks_entries_072720.zip
If you are going to study these versions I strongly encourage you to download these files. Many of these entrants vary only in a minor way and it is hard to see the impact of a change without toggling back and forth between them.

I look forward to your comments.

Dan Margulis

P.S. Our final case study is announced today, look for a separate post.


Paco
 

Hi! Mine is 1115.

Locking up the "king of excessive color," I tried to keep the warm tones from being too oversaturated and too R. The tree between the Australian and USA flags has not yet completely turned R and there is a lot of Y still in there. I wanted to keep the separation in those tonalities. Also tried to keep the subtle difference in the G tonalities. 

On this image I assigned a false RGB 1.0 gamma to the background, duplicated it and in Multiply mode applied an RGB composite mask which was blurred. Worked on contrast and colors in LAB. Flattened that and in sRGB duplicated the BG twice. On one layer HIRALOAM sharpening was applied to the R channel and on the other layer the same sharpening was applied to the G channel. Opened the shadows with S/H. Color corrected using the Speed Limit and road as the grey targets. The sky was color corrected so that it would gradate from a grey on the left side to a sky blue on the right.

All the best!

Paco



Robert Wheeler
 

For many years, I wanted to see fall foliage color in far-away New England forests, but September and October always had heavy work demands that prevented travel. Moving to Vermont for five years solved the travel issue, but I found foliage photography challenging. Masses of tress from a distance would look like carpets without being impressive representations of the beauty. Views with distinct branches and leaves up close and groupings of color further away worked better. Bright sunshine would cause harsh contrast. Overcast days could make the colors look more saturated. Finding vantage points with emphasis on reds, yellows, and orange colors worked better than those that mixed in substantial areas with green leaves. If not too overcast, finding a place that included leaves glowing from backlight could be effective.

 

Our Adirondacks image is not the ideal starting point to highlight impressive fall color. The closer trees are bright with a lot of contrast (leaves at risk of going crunchy if slightly over-sharpened). We find more greenery than desirable. In the distance, the trees on the darker hill seem dull in the original – needing enhancement if we want to convey the message of a very colorful foliage year. Same for the low branches in the lower right.

 

When I saw the submitted entries, 1101 made me say out loud “I wish I had done that.” Enhancing the darkness of the clouds makes the distant hills look very shadowed from the sun. That makes my visual system conclude that those autumn trees out there must be VERY colorful to show up so well despite the darkness. The dramatic dark clouds also provide a pleasing contrast with the brightness of the closer trees and give the car an excuse to have headlights turned on. The balance of reds, yellows, and oranges came out perfectly. The green trees are properly subdued. The neutrals are neutral. Sharpening is subtle. This produces a definite Wow from me and is my favorite, even more than the PAR version. Others may have different opinions.

 

Mine is 1116. After using camera raw filter to make neutrals neutral, I used Image-adjust-equalize (after trying many other things), along with darken sky and H-K actions in preparation for color adjustment. Next came MMM+CB action (with color boost layer changed to soft light at 20% opacity and endpoint adjustment used on Auto). Back in sRGB, I lightened the darkest clouds (thinking this would provide an excuse for the color-adjusted distant hill trees to look appropriate),then boosted saturation of red, yellow, cyan, and blue a little bit, and used Topaz Sharpen AI at low setting.

 

When I blend 1101 into mine at 50% followed by a mild curves adjustment to enhance brightness, I get a result that pleases my subjective taste better than my submission, perhaps better than 1101 alone, and also better than the PAR version. Looks like other image blends could be interesting, but I haven’t tried them yet.

Robert Wheeler


Dan Margulis
 

On Jul 28, 2020, at 10:36 AM, Robert Wheeler <bwheeler350@comcast.net> wrote:


When I saw the submitted entries, 1101 made me say out loud “I wish I had done that.” Enhancing the darkness of the clouds makes the distant hills look very shadowed from the sun. That makes my visual system conclude that those autumn trees out there must be VERY colorful to show up so well despite the darkness. The dramatic dark clouds also provide a pleasing contrast with the brightness of the closer trees and give the car an excuse to have headlights turned on. The balance of reds, yellows, and oranges came out perfectly. The green trees are properly subdued. The neutrals are neutral. Sharpening is subtle. This produces a definite Wow from me and is my favorite, even more than the PAR version. Others may have different opinions.
The two bookends of our series of case studies, Carnival and Red Rose, unabashedly ask us to make red objects as vivid as possible, consistent with realism. As against that, we have Cinque Terre, Monument Valley, and this Adirondacks shot, in all of which more color is welcome even if somewhat unrealistic, but there is a limit as to what viewers will accept. Of these three, Adirondacks is the most extreme example because we are specifically told to emphasize the strong coloring.

Yet we have to avoid blowing the viewer away with brilliance. And although in the autumn certain trees can display fairly bright reds, they can’t compare in saturation to the reds found in a country’s flag, of which there are several convenient examples in this scene. So, the flags need to stand out from the background. Fortunately they have so little detail that it’s easy to paint them in by hand if need be.

Also, there are large areas of dull red in the left-side background. These are very convincing if we want to suggest that color is everywhere. They have to be red enough for the viewer to get that message but they can’t be very saturated without looking ridiculous.

Almost all of us accomplished these goals so I would have to rate this as the most successful group effort so far. Reviewing the entries in more detail after they were posted I think we have about a dozen really good ones. This has the interesting consequence that three or four individual versions are IMHO better than the par, which is not common in these case studies.

Assuming that the things mentioned above have been dealt with successfully, the very best versions can be distinguished by their variation in the trees, which supports the narrative of a dazzling rainbow-type display. Particularly, behind the nearest three flags are two prominent trees, one basically orange and the other basically red. In the original they are hard to distinguish. Best to somehow lighten the orange-yellows, as well as move them toward yellow.

This suggests that MMM or something like it is going to play a critical role. Also, it’s likely that H-K or something similar will be useful, so that the viewer will see patches of brilliance rather than a flood.

For these reasons I’d recommend being very conservative about color in this image up until the time you hit LAB. Some people tried to achieve bright color in Camera Raw first. They pre-empted themselves from getting a better result later, because they found that MMM gave them lurid color.

Most of us tried to get some credibility into the sky so it would be something more than a gray blur. Three people went farther, maybe taking something from the Seated in the Grass exercise, by trying to make the sky heavier and blue enough that it would affect our overall perception. Of these, the most spectacular is #1101. Like Robert Wheeler, I find this possibly the best overall version. I didn’t dare pick it for the par, however, because it would create a discord with whatever else was chosen.

I’ve therefore created and uploaded a second par, called #1123 strong sky par, averaging the three that took that approach. An interesting alternative.

Dan


John Furnes
 

I agree with Robert – 1101 is quite something.

Mine is 1105, and I tried to blend 1101 with 1105, which gave a better result than each of them.

 

To get to 1105, I first started out with Lab, H-K, MMM, and eventually it became very light, but colours were so-so.

I then started over and did

  1. Colour adjustment (not much)
  2. S/H
  3. Lesser Hammer
  4. MMM Fine with a selection of the ground behind the flag poles. This also gave some light to the darker background.
  5. High Pass (7.0 at 100%)

This result, which I thought was a bit too dark, I blended with the first attempt.

 

I think the colours came out good, both in the flags, the foreground and the background, and  even though the background became a bit lighter, I think there is enough difference front to back to make the foreground be the eye catcher.

 

 

John Furnes