Topics

Mantillas: Results


Dan Margulis
 

I’ve posted the results of the Mantillas exercise, the second in a series of 11 case studies.

Reviewing: Stock photography from the 1990s, scanned from film via Kodak Photo CD. No raw file is available. The three women are wearing traditional Spanish garb. We must assume that the subject of the photograph is not just the women’s faces but what they are wearing.

We have 40 entries; a 41st came in about half an hour after deadline and could not be included. 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 #201 to #240. As with past studies, we also have a “par” version, #241. To get it, I chose what I thought might be 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 to open it up to group discussion first. What do you think should and should not be done with an image like this? 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's Photos section, Case Study 2021: Mantillas,
https://groups.io/g/colortheory/album?id=260159

I also have zipped all 41 entries and uploaded a 45 mb file to our Files section,
https://groups.io/g/colortheory/files/
Search for 020821_Mantillas_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.

Dan Margulis

P.S. Our next case study will be announced today, look for a separate post.


Kenneth Harris
 

The par looks very unnatural to my eye.  Too sharp, trying to be something that the photo isn't.  Not pleasing.  Not engaging.  Too heavy in the 3/4 skin, too saturated, too open and gritty in the dresses, an weird lack of variation in the black hues.  It looks like an exercise.  We've lost what was good about the photograph.  There's no story here, just a bunch of regions of concern.  I'm sure many will disagree.

Ken Harris


Kenneth Harris
 

I suggest trying 205+210+212+220+230 to see what I mean.  The only problem I have with that blend is doubtful women seems a touch dark in the face.

Ken Harris


Doug Schafer
 

My first reaction to all was to look at images overall and looking for black dresses, correct skin colors, and most looked pretty good.
But I felt the distinguishing images had good hair, teeth, lips, eyes ... and many were less than "correct".
Then, transparency in the head veils thru to background .... and most everyone had color issues....left me wondering which is correct?

Doug Schafer


Kent Sutorius
 

I did 202. Shocked to see the blue tint on the girl's dress on the right. Was black on my screen. Many good shots. I liked 203, 205, 220. Faces were a challenge. Seemed a number either too red/yellow tint or over saturated. I told Dan I had trouble visualizing what a corrected image should look like.

Kent Sutorius


On 2/8/2021 11:18 AM, Doug Schafer wrote:
My first reaction to all was to look at images overall and looking for black dresses, correct skin colors, and most looked pretty good.
But I felt the distinguishing images had good hair, teeth, lips, eyes ... and many were less than "correct".
Then, transparency in the head veils thru to background .... and most everyone had color issues....left me wondering which is correct?

Doug Schafer



Gerald Bakker
 

I assume that facial detail is the most important aspect of this exercise. I myself focused mainly on that. My version is 223, it looks remarkably similar to the par version, except for one thing: the redness of the lips. I emphasized eyes, teeth, jewelry, but forgot the mouths.
Now that I review my entry, I wonder if the hair of the three women should not be darker, closer to black. In my version as well as in the par, it is more brown than black.
 --
Gerald Bakker
https://geraldbakker.nl


Hector Davila
 

Mines is number 232.

Of course Colors is what is most important
to bring out first.
(since the colors appear to be washed out
on the three women)

So, I focused on the little boy and the
woman behind him. (they have the most colors)

Increasing colors makes his hair a full brown color,
and her hair appears to be reddish brown.

That increases the hair colors on the 3 women, Brown.

(i don't do 'color by numbers' cause I'm allergic to numbers)

And I try to increase all the colors I can find
(regardless if the colors are correct or not)

Then if you want
you can adjust the colors to your liking
(or the clients liking)
and add any details you feel it needs.

Hector Davila


bill bane
 

Realizing I am probably wrong, but.....

I immediately wondered if these dresses were all over and all nearly black/neutral? I googled Mantillas and many were not black, indeed the majority were not. They were and are worn for many events not associated with deaths. And, the vast majority I reviewed were not connected with black underdresses.

Further, I assumed that these ladies were economically well to do and showing off. If so, since blue/purple dyes were historically more expensive (I think), I assumed these were not black, but blue, indicating affluence.

Finally, and this is what really tickled my fancy, by using a blue dress color, I easily was able to separate the arm skin color under the fabric on her right arm. Again, affluence on display, again I thought.

Since none of the Par selections made this "blue" choice, all seemingly going with black, I guess that I am wrong and black is correct. My question is if there is color evidence (?) showing that the undergarment of the lady (audience right) is essentially black as depicted in all par versions?

When I take the original image, and check colors, and measuring the black ornamentation in the over garment over the lady's right breast, my lab readings are 1,0,(1); the readings just below that, on the undergarment, is 20, 0, (17), very blue and quite unlike the over garment's tone and colors.

The par version has this same breast area undergarment region at 32, 0, (4) in one spot, and almost pure grey 25, 0, 1 just below that.

These blue to grey corrections in the par seem to suggest there was a blue cast in the original, and that everyone removed it? I note that in the original the stone on the upper left all has a yellow bias (B roughly positive 6). The stone on the right is a more neutral but bouncing plus/minus 3 on the B Channel. Neither suggests a blue cast. The child, lower right, and the hidden man (?) white shirt, are both yellow (positive B channel) in the original. Why was the undergarment judged to have a blue cast?

Of course, if the undergarment is more blue, the other two ladies' dresses also follow.

I am hoping someone can tell me if there is technical information in the original -- that I missed or do not understand -- creating such certainty that these garments were as black/neutral as most moved them to become.

My image (which has lots of defects) is 207. I am new at this game, only book lernt, so do not beat me up too much. :-)

Thanks,
Bill



On Mon, Feb 8, 2021 at 10:18 AM Doug Schafer <k_d@...> wrote:
My first reaction to all was to look at images overall and looking for black dresses, correct skin colors, and most looked pretty good.
But I felt the distinguishing images had good hair, teeth, lips, eyes ... and many were less than "correct".
Then, transparency in the head veils thru to background .... and most everyone had color issues....left me wondering which is correct?

Doug Schafer


Jim Sanderson
 

I think you and I were on the same page.

Jim Sanderson



-----Original Message-----
From: bill bane <bill.bane@...>
To: colortheory@groups.io
Sent: Mon, Feb 8, 2021 2:42 pm
Subject: Re: [colortheory] Mantillas: Results

Realizing I am probably wrong, but.....

I immediately wondered if these dresses were all over and all nearly black/neutral? I googled Mantillas and many were not black, indeed the majority were not. They were and are worn for many events not associated with deaths. And, the vast majority I reviewed were not connected with black underdresses.

Further, I assumed that these ladies were economically well to do and showing off. If so, since blue/purple dyes were historically more expensive (I think), I assumed these were not black, but blue, indicating affluence.

Finally, and this is what really tickled my fancy, by using a blue dress color, I easily was able to separate the arm skin color under the fabric on her right arm. Again, affluence on display, again I thought.

Since none of the Par selections made this "blue" choice, all seemingly going with black, I guess that I am wrong and black is correct. My question is if there is color evidence (?) showing that the undergarment of the lady (audience right) is essentially black as depicted in all par versions?

When I take the original image, and check colors, and measuring the black ornamentation in the over garment over the lady's right breast, my lab readings are 1,0,(1); the readings just below that, on the undergarment, is 20, 0, (17), very blue and quite unlike the over garment's tone and colors.

The par version has this same breast area undergarment region at 32, 0, (4) in one spot, and almost pure grey 25, 0, 1 just below that.

These blue to grey corrections in the par seem to suggest there was a blue cast in the original, and that everyone removed it? I note that in the original the stone on the upper left all has a yellow bias (B roughly positive 6). The stone on the right is a more neutral but bouncing plus/minus 3 on the B Channel. Neither suggests a blue cast. The child, lower right, and the hidden man (?) white shirt, are both yellow (positive B channel) in the original. Why was the undergarment judged to have a blue cast?

Of course, if the undergarment is more blue, the other two ladies' dresses also follow.

I am hoping someone can tell me if there is technical information in the original -- that I missed or do not understand -- creating such certainty that these garments were as black/neutral as most moved them to become.

My image (which has lots of defects) is 207. I am new at this game, only book lernt, so do not beat me up too much. :-)

Thanks,
Bill



On Mon, Feb 8, 2021 at 10:18 AM Doug Schafer <k_d@...> wrote:
My first reaction to all was to look at images overall and looking for black dresses, correct skin colors, and most looked pretty good.
But I felt the distinguishing images had good hair, teeth, lips, eyes ... and many were less than "correct".
Then, transparency in the head veils thru to background .... and most everyone had color issues....left me wondering which is correct?

Doug Schafer


Doug Schafer
 

Not familiar with a mantilla, I looked it up on Google: its the veil:
 
noun: mantilla; plural noun: mantillas
  1. a lace or silk scarf worn by women over the hair and shoulders, especially in Spain.

    Doug Schafer


bill bane
 

Right, this is from wiki, " Spain, women still wear mantillas during Holy Week (the week leading to Easter), bullfights and weddings. Also a black mantilla
is traditionally worn when a woman has an audience with the Pope and a white mantilla is appropriate for a church wedding, but can be worn at other ceremony occasions as well"

The site includes some images, and a few are shown below. The first I just found, and shows the use of a blue underdress/garment.

Bill




image.png




image.png

image.png



On Mon, Feb 8, 2021 at 6:20 PM Doug Schafer <k_d@...> wrote:
Not familiar with a mantilla, I looked it up on Google: its the veil:
 
noun: mantilla; plural noun: mantillas
  1. a lace or silk scarf worn by women over the hair and shoulders, especially in Spain.

    Doug Schafer


Ronny Light
 

Interesting, Bill.

 

My submission is 201. I was surprised by all of the black mantillas in other entries, and only a few blue mantillas.

 

The girls on the left, in my entry, had light blue mantillas, with black and blue in the headgear of the girl in the center. The girl on the right had black lace over a blue undergarment and black headgear.

 

I didn’t assume that all dresses were black. I was more concerned with skin tone.

 

I point out that, in the par version, there are no blue costumes.

 

 

Ronny

www.RonnyLightPhoto.com

5010 B Wilkerson Dr., Nashville, TN 37211

 

 

 

From: colortheory@groups.io <colortheory@groups.io> On Behalf Of bill bane
Sent: Monday, 8 February, 2021 7:11 PM
To: colortheory@groups.io
Subject: Re: [colortheory] Mantillas: Results

 

Right, this is from wiki, " Spain, women still wear mantillas during Holy Week (the week leading to Easter), bullfights and weddings. Also a black mantilla

is traditionally worn when a woman has an audience with the Pope and a white mantilla is appropriate for a church wedding, but can be worn at other ceremony occasions as well"

 

The site includes some images, and a few are shown below. The first I just found, and shows the use of a blue underdress/garment.

 

Bill

 

 

 

 

 

 

 

 



 

On Mon, Feb 8, 2021 at 6:20 PM Doug Schafer <k_d@...> wrote:

Not familiar with a mantilla, I looked it up on Google: its the veil:

 

noun: mantilla; plural noun: mantillas

  1. a lace or silk scarf worn by women over the hair and shoulders, especially in Spain.

    Doug Schafer


Hector Davila
 

Well of course you can make the dress Blue or Black

(but I have no idea what Color Dan would pick.)


But, I see a lot of color photographs prints where people
with black hair it comes out blue.

Are all those cameras incorrect
when they output blue hair?

The blue hair is the correct color, just in the real world it's incorrect.

I sometimes like that blue hair look. It looks like a... color photo.

I like to add the color blue to white snow. To re-create a Kodak Color Moment.

Hector Davila


KENT SOUTHERS
 

I think my submission is 240(M).

I assumed that the lighting is outdoor.
I noted the shadows (harsh / soft) and deduced soft lighting, inferring cool skylight
I compared the RGB values of the dresses, mantillas and gloves.  I assumed they were color matched
Decision between blue vs. black was given credence to black, since the lighting was deemed to be cool. 
Deduction that if the dresses were actually blue (and in cool light), they would be VERY BLUE, and there wouldn't be any question about blue vs. black, it would be obviously blue.
Noted the textural difference of the right woman dress (i.e. velvet to absorb light more than the sheen of the other two women)
Raised saturation to assess areas of blue (i.e. deeper shadow) to confirm blue lighting (as assumed at onset)

Remainder, dial to taste to reveal differing hair color in women, skin tone to taste.


Matthew Croxton
 

My answer to the question that many have raised regarding why blue is not the new black: I interpret the majority of the blue color cast to be an artifact of the capture medium (or its processing). In his book, Post Exposure: Advanced Techniques for the Photographic Printer, 2nd ed., Ctein describes the origin of this blue bias in his chapter on "How the film sees." If the image was taken for stock, it was probably on slide film, which is described below.
"This is color crossover. In the shadows on the shoulder of the curve and the highlights on the toe of the curve, the blue-light density is too low, which means a shortage of yellow dye in the image. In the darker midrange densities the effect is reversed; the blue curve is higher than both the others (i.e., the film is absorbing more blue light than any other color), so the darker midtones will be too yellow at the same time that the highlights and shadows are too blue." Although this passage is in reference to printing, development and exposure variables may result in similar outcomes for film [when done incorrectly].

I did not have this theory in my mind when correcting. My first version came out like many of the blue ones here, until I saw a version by a student who corrected to black. I immediately scrapped the first version and worked on a second, which eventually became Image 226. I went dark in my processing because I am not a fan of film grain in the shadows. In retrospect, I think I took the density intensity too far. However, in eliminating the blue color cast on the dresses in the subjects, I note that both the woman with glasses in the background and the boy beside her still have dark blue garments.


Kent Sutorius
 

As a side note, Ctein is giving away his Post Exposure book away for free. You can download it in pdf form from http://ctein.com/booksmpl.htm

Kent Sutorius


On 2/9/2021 6:27 AM, Matthew Croxton wrote:
My answer to the question that many have raised regarding why blue is not the new black: I interpret the majority of the blue color cast to be an artifact of the capture medium (or its processing). In his book, Post Exposure: Advanced Techniques for the Photographic Printer, 2nd ed., Ctein describes the origin of this blue bias in his chapter on "How the film sees." If the image was taken for stock, it was probably on slide film, which is described below.
"This is color crossover. In the shadows on the shoulder of the curve and the highlights on the toe of the curve, the blue-light density is too low, which means a shortage of yellow dye in the image. In the darker midrange densities the effect is reversed; the blue curve is higher than both the others (i.e., the film is absorbing more blue light than any other color), so the darker midtones will be too yellow at the same time that the highlights and shadows are too blue." Although this passage is in reference to printing, development and exposure variables may result in similar outcomes for film [when done incorrectly].

I did not have this theory in my mind when correcting. My first version came out like many of the blue ones here, until I saw a version by a student who corrected to black. I immediately scrapped the first version and worked on a second, which eventually became Image 226. I went dark in my processing because I am not a fan of film grain in the shadows. In retrospect, I think I took the density intensity too far. However, in eliminating the blue color cast on the dresses in the subjects, I note that both the woman with glasses in the background and the boy beside her still have dark blue garments.



David Remington
 
Edited

My image is number 214.
 
I agree with Matthew's assessment. The photograph was shot on film and there was no auto white balance back in the day. Open shade/overcast indirect outdoor lighting is very cool. That is not to say the black clothing is perfectly neutral, but it is most likely close.
 
I prefer a "natural" look in my work so generally leave some cast from the natural lighting for ambiance. More or less depending on the conditions and the image. Outdoor images with all the neutrals completely desaturated look artificial to me. A strong cast usually does not match how the scene would appear to your eye either (sunsets etc. aside). Somewhere in the middle. With digital work I often start with a daylight balance as film would see it and proceed from there.
 
Looking at my image with fresh eyes, I can see some small adjustments are needed. I might go a little warmer overall and maybe open up the blacks a bit and darken the faces a touch. Fix the cyan cast in the woman's hair on the left and the green cast in the veil on her right arm. Less cyan in the gloves. No big moves.


KENT SOUTHERS
 

Yes, the color of the garments for the boy are a different color than the color of the garments for the women ... yet, still illuminated by the same (or nearly so) color of light.
Alternate assumptions / deductions are the building is not blue, rather the non-colored stone would be neutral (or reflective of incident light color) vs. slight blue bias, and the boy's collar, also.
Globally moving the dresses toward blue, reveals more strongly the blue highlights in the women's hair.  As the highlights are AI=AR reflective indications of the incident light color, I deduced that women do not have blue hair, thus it must be blue light.  As the blues rise in the women's hair may seem subtle, a (temporary) increase in saturation aids the eye to view / confirm what the numbers are suggesting.

As they are all (seemingly) illuminated by the same color of light (dresses, gloves, mantillas, hair, collar, stone), the indication of color bias / cast from the lighting is triangulated cool ... i.e. direction of correction is away from blue.


bill bane
 

David,

Maybe I have found what has confused me. When I go back to the two close-as-I-can-find "known" white objects: a) the white collar of the child and b) the white shirt behind the rightmost lady, I now find a more complex color picture than I originally thought.

I had placed color points on each of these "white" spots, and they were, and are, yellow, B positive, not negative, blue. The unobstructed (by the veil) b) shirt is B = 5, 6, 7, certainly showing a yellow cast, and if corrected (along with me seeing the stones being B positive), will make slightly blue things more blue. This is what I did.

However, I was looking at the most open, least noisy parts of the two white collar/shirt areas. When I move to the more noisy, less clear (whatever) parts of these two white collars/shirts, which might be the shadowy parts, the B channel moves negative, with it going as low as -11 in the lower part of the "V" on the b shirt down low. The child's collar does not have these extremes and stays mostly B positive but as it is crunched down by the blue sweater, it begins to go B negative.

To me, this seems to mean that the ambient "light" conditions did not have a cool B channel cast, since the centers of these two objects had, and have, a yellow cast. However, as this "white" went into shadows, the film (or scanner settings) must have added a blue cast. Since the undergarment of the dress is also in whatever is this shadow condition, the film/scanner again might have added a blue cast there too. (I should note that I know almost nothing about film emulsions and if my theory is even possible. I do know that today's scanners have automatic and chosen colorshifting.)

But, even if this is true, I find all parts of the contentious undergarment to be B channel negative, most over -15. The blue to the right of the pinkish corded necklace is -21 and it is in no shadows. There are numerous parts that are -24.

I am expecting Dan to clarify this but maybe the undergarment is not "quite blue", as I have made it, but "somewhat" blue (more so than the Par I guess). In which case, the "right" solution might be for it to end up bluish grey, or greyish blue, taste determined.

Does this make sense?

Bill


On Tue, Feb 9, 2021 at 8:55 AM David Remington <david_remington@...> wrote:

[Edited Message Follows]

My image is number 214.
 
I agree with Matthew's assessment. The photograph was shot on film and there was no auto white balance back in the day. Open shade/overcast indirect outdoor lighting is very cool. That is not to say the black clothing is perfectly neutral, but it is most likely close.
 
I prefer a "natural" look in my work so generally leave some cast from the natural lighting for ambiance. More or less depending on the conditions and the image. Outdoor images will all the neutrals completely desaturated look artificial to me. A strong cast usually does not match how the scene would appear to your eye either (sunsets etc. aside). Somewhere in the middle. With digital work I often start with a daylight balance as film would see it and proceed from there.
 
Looking at my image with fresh eyes, I can see some small adjustments are needed. I might go a little warmer overall and maybe open up the blacks a bit and darken the faces a touch. Fix the cyan cast in the woman's hair on the left and the green cast in the veil on her right arm. Less cyan in the gloves. No big moves.


Harvey Nagai
 

I expected a wide variation in how people would see the faces but I didn't expect such 
variance in how people see blacks.

Mine is easily identifiable, although not because it is an extreme outlier (233).

Kenneth Harris' comment that the par looks like an "exercise" applies to mine more so,
because that was how it came to be: several problems solved separately and not brought
together very well (not to mention the timid color).

I would ask for an explanation of "3/4 skin", I've never come across that term before.

#216 looks like a Dan-blender special à la last summer's "Veiled Bride" case study.

Regarding the color of the dresses, the explanations offered by Matthew Croxton and
David Remington are probably close to the truth.

In the image posted above, if the graying gent's hair and the cobblestones are blue,
then probably the dress is also blue.  But I think that image has a blue cast.  It does
read above the three-quartertone, so maybe the dresses in the case study image are
not as dark as I interpreted them to be.