High Andes: Results

Dan Margulis

I’ve posted the results of the High Andes exercise, our final scheduled case study for this year.

Reviewing: This excellent original, by Vincent Versace, represents a girl wearing the colorful traditional costume of the Quechua-speaking people of the High Andes. We have 25 entries; one was rejected for being at the wrong aspect. 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 #1001 to #1025. As with past studies, we also have a “par” version, #1026. To get it, I chose five that I thought were among the 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.

Normally I don't comment on results for two days after they're posted. 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's Photos section, 2021 Case Study: High Andes,

I also have zipped all entries and uploaded a file to our Files section,
Search for 040521_High-Andes_entries.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, and hope you have found this series worthwhile.

Dan Margulis

P.S. As advertised, there will be an eleventh case study, but it is unusual and it is unclear when it will appear, tentatively two weeks from today.

Kenneth Harris

This one is a bit unusual, I see a couple-three that stand on their own, then another six or seven that what I see as problems which fairly easily blend out but would make them unacceptable for the expected usage. I'm 1001, but I prefer par, although it's just a touch too sharp on the face to me. 1006 is very nice.

Ken Harris

Dan Margulis

On Apr 5, 2021, at 7:56 AM, Kenneth Harris <reg@...> wrote:

This one is a bit unusual, I see a couple-three that stand on their own, then another six or seven that what I see as problems which fairly easily blend out but would make them unacceptable for the expected usage.  I'm 1001, but I prefer par, although it's just a touch too sharp on the face to me.  1006 is very nice.

I step in to avoid confusion/chaos. It is no wonder that Ken likes #1006, since that’s his own entry. #1001, though similar, belongs to somebody else.


Kenneth Harris


Kenneth Harris

In my defense, I was showing three kids under ten how to do a fast edit (on a laptop) right before school. 1001 is very nice. They were arguing about 1001 vs 1006. 1016, 1003, and 1007 were also in their mix.



Mine is 1020.  One issue here is that there is no area that seems clearly and undebatably white.  So I was guided by the idea, emphasized in the Mantillas discussion, that for Hispanic faces, b should be “considerably higher” than a.  Maybe I gave too much weight to “considerably.”  But a number of entries seem to have given very little weight to that stricture.  There was obviously room for discussion how sure we could be that any or all of the women in the Mantillas picture were Hispanic.  But it seems to me that we can assume that the lad in this picture is Hispanic.  Or at least that there’s no basis to assume the opposite.


More debatably, I also assumed that the embroidery on his cloak should be close to white.  Here, we can’t be sure.  Even if it started out white, perhaps the colors ran when it was washed or out in the rain.  But white seems a fair assumption, and to me it makes for a more attractive picture. (Of course, it’s difficult to measure the colors in this embroidery with precision, but I aimed for close to white.)


My picture also has more brightness and saturation than most.  One consideration was that this picture is about a charming young lad in a colorful outfit.  So why not allow it to be really colorful?  Another consideration is that as I’ve gone through these competitions, and read other evaluations in Dan’s books and blog, my general takeaway has been that he prefers more boldness and saturation than I have in the past.  There’s no right and wrong there, within reason, and he certainly has vastly more experience in judging what his clients will prefer than I can claim.  So perhaps, in aiming for “what Dan likes,” I overshot.


With regard to saturation, there’s also the phenomenon that the world at large has drifted toward more saturation, as shown by the evolution of Adobe profiles and elsewhere.  So I tried to not fall too far behind in this exercise. 


Finally, my photo is deficient in  bringing out the texture of the boy’s cloak.  I ran MMM, which did help a bit, but my efforts in that direction were probably weakened by the brightness and saturation I embraced.  I assume the better images in this regard found blends that worked.


Bill Iverson

Christophe Potworowski

My image is #1010. I also had difficulty finding a neutral or white point. I picked the white of the eyes, but this was a mistake, I should have picked a reasonably clean spot on the white hat as per the par image.

Christophe Potworowski

Dan Margulis

On Apr 5, 2021, at 10:05 AM, bill_iverson_washington <bill@...> wrote:

So I was guided by the idea, emphasized in the Mantillas discussion, that for Hispanic faces, b should be “considerably higher” than a.  Maybe I gave too much weight to “considerably.”  But a number of entries seem to have given very little weight to that stricture.  There was obviously room for discussion how sure we could be that any or all of the women in the Mantillas picture were Hispanic.  But it seems to me that we can assume that the lad in this picture is Hispanic.  Or at least that there’s no basis to assume the opposite.

There is ample reason. First, I believe that this is a girl, not a boy, because my understanding is that it is the women who dress in colorful clothing, and also traditionally women of a marriageable age wear many layers of clothing as a sign of their affluence.

Chances are that this individual has not a drop of Spanish blood in her veins, being descended from people who were indigenous to the region long before the Spanish arrived. Their counterparts in North America have all manner of different fleshtones, ranging from the heavily magenta skin of my Chickasaw ancestors in Oklahoma to the extremely sallow Makah people in Washington state.

I believe that this is also the case generally in Latin America, and recall being eviscerated by people who live there after one edition of Professional Photoshop for not having adequately acknowledged the difference between pure Hispanics, Mestizos, and 100% native blood. Plus, according to a Brazilian reader,  "E depois há os argentinos, que não passam de um bando de italianos que falam espanhol e fingem ser ingleses.” I regret that in this age of political correctness I am unable to provide a translation.


Doug Schafer

Mine is #1008
I did a Google/Bing search of images of Peruvian children skin colors, and after measuring several samples, I settled on colors I used in my image; also accounting for the cheeks reflecting some red from the clothing.
Doug Schafer

Edward Bateman

Hello everyone-

I am #1009.

In comparing mine to the par, I'm quite similar in terms of color. But in terms of luminosity, I have much more contrast... probably from a High Pass/Soft Light layer (at around 70%). Although the contrast is global, it is most noticeable in the face. And in looking at some layers, it seems that some Man from Mars exaggerated the effect. (Turning off MfM and High Pass makes mine much closer to the par.)

Part of me kind of still likes it... I feel it puts emphasis on the face. Another part of me thinks made developing cataracts are starting to rear their ugly head :) But I concede that it is probably a bit too much...

Thanks everyone!

Edward (Ed) Bateman

David Remington

My version is 1004 and appears cool and dark in the context of most other entries. Looking at it layered with the par I can see blending in a little color, Just a touch though maybe 20%. I can go either way with a blending in a little luminosity. I find the par and many other version too orange in the child's skin tone and in the red stripe of the fabric. I was aiming to avoid making the skin tone too warm or too saturated. In my base export from camera RAW the clothing was very saturated and there was little separation between the orange and red stripes. I increased that separation and reduced the saturation quite a bit. I'll think on this one a bit more and am interesting in reading Dan's take and more thoughts from the group.


Robert S Baldassano

Mine is 1005, and it looks quite different than the others. I was trying to accomplish 2 things. First I want the color of the face to come close to matching  Inca  natives that I had photographed a few years ago in Peru where I had contact with children, and older people, all who could trace their Inca ancestry. My second goal was to capture the saturated colors of the clothing I saw the people wear while there. I started with this image in Nikon NX  Studio where I had full access to all the Nikon Camera settings. Once I was happy with that result, I moved the image to PS2021 and the PPW panel and then applied the 2015  face recipe that in the LAB Color Ver 2 book. In looking back my A& B numbers in this file are higher, though in the same ratios I found in the other photos I mentioned, as they were  numbers like 15 A, 17b, 16A, 17B, 12 A, 12B, 21A, 20B and in the 1005 file the numbers are much higher. I am happy with the clothing saturation which is certainly more so than PAR and many of the other images submitted. I think in that area I am true to the colorful costumes people wear in Peru. I think I need some time to compare what I have done to the others and see how my A&B numbers compare. Some of the faces seem to be too Caucasian to me, but did I push my face color too far?


Robert S Baldassano


Sent from Mail for Windows 10


James Gray

I think my version is competitive with par.  Mine is 1021.  Maybe I am seeing this as too easy as I would characterize 14 of the 25 as acceptable to be included in par.  I have not been very impressed with about half of my efforts on these case studies.  This exercise may be telling me that one interpretation I have of my skill is not accurate.  I had assumed I was pretty good at starting with a mediocre image and making it good.  I tend to think I am not so good at taking a good image and making it excellent.  It would appear that my entries in this set of exercises do not support that.  This image is good from where it started.  I am pleased with my results.  A couple of technical comments.  I did these channel blends

  1. Applied the L channel to red in darken mode at 40%.  Applied the black to the red in multiply mode 40%.

  2. Apply L to green channel in lighten mode 15%.  Apply black to green in multiply mode 25%.

I neglected to mention in my steps that I did these blends on a separate layer that I converted to luminosity mode.

After I ran the MMM+CB action I seemed to have detail in the red stripes of her blanket. When I converted the image to RGB there were parts of the 2nd broad red stripe that lost detail and turned into red blobs. I could not figure out a way to maintain the detail present in LAB color mode when I converted to RGB. So I created two new versions concentrating only on that red stripe. One version used the Fixel Detializer plug-in. On the other, I used the clone tool to build in some detail. I created a mask so only the 2nd stripe contributed to the final version. The latter was blended at 57% opacity. It appears there were 3 versions that had the problem of the highly saturated 2nd stripe losing detail. They are 1005, 1011, and 1016. To my eyes 1011 and 1016 are pretty good except for the lack of detail in the red stripes.

Jim Gray

John Gillespie

My version is 1014. I have more contrast in the garment and less in the face than the par - I suppose this is better than the other way around. I did not want to over-emphasize the dirt on the child's face, which I think the par does a little too much.

Interestingly toggling between them in colour mode in RGB causes a noticeable shift in luminosity in the red part of the garment which does not happen in Lab -  something Dan has explained before as being caused by "impossible" colours which Lab can accommodate but RGB cannot -  a highly saturated but dark red.

John Furnes

Mine is 1011,

I went for a darker colour in the face – browner. This could very well be the true colour, but seeing that the dark red band (around the neck) has also become brown, I realise it was too much.

The people of the Andes have been able to make indigo colours – not easy, and the blue band would possibly be that colour. Mine is lighter blue. They have also lately used modern colours based on chemistry, but I think I would prefer the indigo.

Anyway, I think my colours are not too bad – the orange has got some ‘blobbyness’ over it, and lost some of the details in the weft.


I do like the par version. I think it is very much how many would perceive the scene.


John Furnes



Kenneth Harris

Most of the pictures we've gotten are cast-offs, nostalgia disasters, or salvage jobs. This is not. So my advice is to consider what the photographer was seeing and doing in taking the photograph. This photo is about the expression plus situation, and the light. It starts of in a great place. It gets to 85% of where it needs to be just with a white point move and some contrast. Your job is here is help tell the story. It's not strictly an anthropological shot either, so there's a lot of freedom to change relative emphasis. I see several comments about researching skin tones, but if I had gotten this as a job, first thing I would have done would have been researching the photographer's work. SOP, before accepting any job I look up the photographer and see if we were a good fit. Also, note that the shot is lit by extremely blue north light. It's in the eyes. You have to decide whether to correct that or not. The more you respect the photographer, the less likely you are to "fix" their "mistakes." And all but one of the photographers I work with despise sharpening. They let their lens and aperture do the work. Proceed with caution.

So again, here we've got expression, situation, and light. For expression, you just need to keep the greatest attention on the face, meaning good contrast, and not over contrasting the rest of it. For situation, that's the clothing, Dan will have much to say about that. I think a defect in many is hitting the knit under hat too hard, which, being close to the face, pulls away from the eyes. OTOH, having detail farther away is less of problem in that the eye can circle around and come back to the center. For light, the image starts with a wonderful blueness to the highlight on the nose. That's a feature.

A last thing, you've got that out of focus grey space on camera left. What is that? Opportunity with out obligation. Careful use of that can affect your perception of the reds. Someone should write a book about that.

Overall, there's so much to work with here, it's easy to overplay, which, in my experience, will lead to an unhappy client. First rule is to follow the light.

Ken Harris

Dan Margulis

After reviewing the entrants more carefully, I think I made some poor choices in producing the par version. As a couple of you have noted, there is an embarrassment of riches here, about a dozen versions that could reasonably be chosen for par. So I chose five new ones for an alternate par, which is posted in the folder as #1027. I like it better than #1026; perhaps you will as well.



While I probably didn’t add enough contrast to the face, I find most to be very creepy. The face of a cherub has been morphed into an extra from “Walking Dead”. The smutz on her face has become gaping scabs and scars. Not a pleasant look. You do know that you don’t need to use *all* of Dan’s techniques on every image, right?
Also, along the lines of the “if everything’s colorful . . .” rule, I would submit the “if everything’s contrasty . . .” rule. For me it’s really tiring looking at excessive contrast. So I kept the face soft and delicate. It offers some relief, and separation, from colorfulness and contrast found in the clothing.
Anyway, mine is 1025, and I was probably too delicate, but most were way too harsh. At least for my taste. Hmmm . . .
Steve J


With regard to Dan’s message on the nuances of Peruvian facial coloring, I was using “Hispanic” in a loose fashion.  I was really guided by the facial coloring of indigenous Peruvians that I have known.  Not hundreds, but quite a few.  Almost without exception, they have facial coloring similar to my submission, men, women, boys and girls.  None looked like they were on a year abroad program from Minnesota.

Kenneth Harris

There are things about the face which I like better in the new par, but both the relationship btw skin contrast and the surrounding area, and the directionality+cast of the light is weaker. I'd go 50/50 btw new and old par. The basis for my assessment is essentially to ask "would this look right in National Geographic." Old par too punchy, new par just a bit too weak on the face and loosing the light. I think there's some degree of freedom on the yellow plate, given acceptable compensation (or not) for source temperature (open sky). I was short Y, and plus M, on 1006, which is obvious in the sclera. This group was the hardest to edit of the ones I've participated in.

Ken Harris