Topics

Hotel Lobby: Results


Dan Margulis
 

 

I’ve posted the results of the Hotel Lobby exercise, the first in a series of 11 case studies.

 

Reviewing: This image was part of the MIT study. We are asked to assume that this is a promotional image for a certain hotel, showing how tasteful and elegant their interior design is.

 

We have 141 entrants, a record. 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 #101 to #141. As with past studies, we also have a “par” version, #142. 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’d like 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: Hotel Lobby, 

https://groups.io/g/colortheory/album?id=259709

 

I also have zipped all 42 files and uploaded a 61 mb file to our Files section,

https://groups.io/g/colortheory/files/

Search for 020121_Hotel-Lobby_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
 

Very interesting.  Your par and my par differ by one image.  Unsurprisingly, I prefer my par, even though it brings back more of the reflections that I didn't like.  Also unsurprising, my bad habit of letting the black float too high once again weakens mine.

Ken Harris


Harvey Nagai
 

Right off the bat I can see a dichotomy between lighter more orange corrections
and darker browner ones.

I would guess that one of the reasons is the choice of source file: the flat
version which is lighter (after de-flattening) and more orange or the default
version which is darker and browner.

I'll reserve the value judgement for Dan, for myself I chose to use dng conversions
based on both versions and made a coin flip decision to correct towards the default
image (which was not towards the par).


Kent Sutorius
 

Mine was darker than the par but I think the colors are similar. Surprised an inside shot would be that bright but agree my should be brighter and more vibrant.
Dan, do you reveal the five you used for the par?

Kent Sutorius


On 2/1/2021 10:43 AM, Harvey Nagai via groups.io wrote:
Right off the bat I can see a dichotomy between lighter more orange corrections
and darker browner ones.

I would guess that one of the reasons is the choice of source file: the flat
version which is lighter (after de-flattening) and more orange or the default
version which is darker and browner.

I'll reserve the value judgement for Dan, for myself I chose to use dng conversions
based on both versions and made a coin flip decision to correct towards the default
image (which was not towards the par).



Doug Schafer
 

Wow, such a good response of 141 images....there is a lot of interest.

My quick, summary:
Very few really wrong.
Seems like way too many lack contrast and look flat (and likely too dark or too light too)...and I know I still have trouble getting excellent contrast and avoiding over-saturation.
#142 par version is excellent; probably the best of all.
Will wait to hear from others and Dan's critiques.
Excellent exercise!!!

Doug Schafer


Gerald Bakker
 

Interestingly, mine is yellower than all other entries.

Well, there is a reason for that. I sampled the greens in the original and found out that they are not yellow enough. In LAB, the A value higher than the B (absolute values). So I countered that quite radically, and #111 is the result.

Did I go too far? Maybe. I have the feeling that many versions are still on the blue side, even the par. On the other hand, my light marble floor tiles looks muddy compared to most other submissions. The question is what it should be. Blue-grey as most have, or more beige?

I prefer a bright orange background over a darker, more brownish one. With so many submissions, it's hard to pick favorites, but I do it anyway: 133, 119 and 106, and the par as usual.
--
Gerald Bakker
http://geraldbakker.nl


Doug Schafer
 

My "test" for WB was the top HVAC duct and assumed it should be near white/grey...could be slightly blue from skylight/windows or slightly yellow from indoor lighting (and it looked too dirty for a product shot but should not be too bright to steal the scene from plants and marble/granite and brass).
And where greens were off color, I corrected the greens separately (masks); especially the top of palm which seemed too washed out and bluish.

Doug Schafer


David Remington
 

I prefer the look of 114 which is similar to my 106 though a bit richer. Those two averaged with 119, more acutance, and 138, more saturation, and 128, lean and clean with good separation between the foliage and the flowers, make a nice "par" version.

--David


Bill Theis
 

I had the same issues as Gerald Bakker with the greens.  apparently having A value higher than B (absolute values) doesn't bother folks or is this one of those plants (like evergreens where the absolute value of A is 2/3 of the B - Page 57 of Modern Color Workflow) where I should have known that it would violate the general recommendation of making B "1.5 to 2.5" times higher than absolute A (p45)? 

Getting this answered is very, very important to my work since it almost always includes greeny!

I confess missing touching out the bright highlight, muddying up the color of the red flowers and not getting enough color difference to the background, snap and a variety of other things.  A very helpful exercise, but I really need the answer to Gerald's question


Frederick Yocum
 

A surprisingly demanding photograph. The colour corrections difficulties for me were the high tones of the palm fronts as they picked up ceiling reflections and the outright reflections on the wall. 

Quickly going through the photographs, I arrived at 114 as my top choice and par as my second. I didn’t spend anytime figuring out which one was mine, though it was neither of my choices.

regards,

Frederick Yocum
frederick@...


Dan Margulis
 



On Feb 1, 2021, at 11:28 AM, Kent Sutorius <kasutorius@...> wrote:

Mine was darker than the par but I think the colors are similar. Surprised an inside shot would be that bright but agree my should be brighter and more vibrant.
Dan, do you reveal the five you used for the par?

Yes, in a couple of days, I think group members deserve that long to discuss things before the bull enters the china shop.

I will, however, toss out one tidbit. There’s no shame in having a version that doesn’t compare favorably to the par. Not comparing favorably to what the MIT retouchers did is another matter. They occasionally get good results (one of them has been complimented a couple of times already  in this thread) but by and large they shouldn’t be at our skill level.

Let me reveal, then, that #109 is the average of the five MIT folk. They weren’t allowed to sharpen, so anybody who wants to compare it to their own work should, in fairness, sharpen it first. (I should have done this myself; we have three other MIT sets in this series and I’ll remember to do it.)

So I would hope that people would be able to look at #109 sharpened and honestly say to themselves that their own version is better.

Dan




Robert S Baldassano
 

 

 

Sent from Mail for Windows 10

Mine was #118 and overlaying PAR I can see that there were a few areas I could have done better, primarily in making the white marble a bit more so and making the greens a bit brighter. I was having a bit of trouble in adjusting to the new version of Photoshop 2021 just moving over from CS6 extended that I used for years.  I am curious how others handled the bleached out founds, as I think I did it the hard way. Instead of looking for a proper mask within the channels, which I should have done, I just went into LAB and created a color blend layer and painted in a mask. I also should have picked a better green color to blend as I can see from the PAR, that mine could have used a bigger L boost. This is the first time I have participated here and it was lots of fun. I hope to do better on the other challenges. I am not a professional; just a hobbyist who enjoys post processing/color correction. Dan through his books and some hands-on help when I was trying to interest other Nikonians in PPW, has taught me a lot. I hope to learn much more from this group.

 

Robert S BAldassano


Dan Margulis
 



On Feb 1, 2021, at 9:07 AM, Kenneth Harris <reg@...> wrote:

Very interesting.  Your par and my par differ by one image.  Unsurprisingly, I prefer my par, even though it brings back more of the reflections that I didn't like. 

We need to better define those “reflections that we don’t like” because we have several varieties here. 

First, though, I noted that one contestant named his file HotelLobbyPosh. I agree. According to Oxford, posh means “elegant or stylishly luxurious.” And this hotel, wherever it is, surely is that. This is a very expensive room. The floors are marble. The door in the background, and the planter, are brass. And the wall, I am pretty sure, is granite.

Finished granite is polished and almost as highly reflective as the brass. If all reflections in the wall are removed completely the result looks cheap and lifeless, not in keeping the the luxurious nature of the stone.

There are, however, several kinds of reflections in this image. From top to bottom, and I may have missed some:

1) Reflection of lighting at the top of the wall, behind the a/c unit.
2) Reflection off the top right leaf, making it lighter and bluer.
3) Reflection of three yellow lights in the wall, right below this leaf.
4) Large reflection in the middle of the background behind the plant.
5) Large reflection in the brass door at the same height as #3.
6) Lesser reflection lower down in the door.
7) Reflections at the same height as the last two on right side of wall.
8) A series of reflections within the brass planter.
9) A series of reflections from the planter onto the marble floor.

Now, which ones of these don’t we like? And what should be done?

For me, the clear loser is #2. The plant is the only object here that the viewer doesn’t understand to be shiny. Why should we allow it to be reflective and compete with all these other things? The clear winners are #8-9. The reflections there define their materials and don’t detract in any way. Why should we try to retouch them out?

I am neutral about the yellow reflections in the background objects. I could see a slight reduction of #5 if it would assist in the treatment of #4, but it can’t be eliminated altogether without losing the sensation of a brass door. I see no reason to take out the reflection of the three yellow lights (#3) in the wall.

The big question, which I think blinds us to some of the others, is what to do about #4. Take it out completely, and we risk the wall’s realism. Ignore it, and it’s a distraction.

Then the smaller question: assuming something is done with #4, should it also be done with #7, the smaller reflections on the right side of the wall? They seem to be a continuation of the reflections in the brass door, so maybe they should be left alone in the name of realism. Or not.

My strategy, then, is:

1) Get rid of the reflection in the top leaf.
2) Reduce, but do not eliminate, the main reflection at center.
3) Then make a seat-of-the-pants decision about #7.
4) Leave everything else alone.

I disagree, therefore, with the downplaying of the three yellow lights (#3) in the par version. Having them at full intensity emphasizes the reflectivity of the wall and draws attention from the big reflection. That big reflection, reduced without being eliminated altogether, seems to me a reasonable compromise in the par, but I think I’d prefer downplaying it even further.

I regret to say that my own version didn’t even go that far in reducing the main reflection, because I went about it in a silly way. Getting rid of the leaf reflection is easy, by blending red into blue Darken mode to increase yellowness, and then by blending red into RGB Darken and Luminosity mode for shape. Fair enough, and no reason to avoid doing it right away.

In reducing the main reflection I started by recovering highlight during acquisition. Then I did several time-consuming things to build it up and baby it along the way. I thought I had a good treatment at the time, but now I think I should have gone farther.

Instead, I should have ignored the main reflection until otherwise finished with the job. Then, make a duplicate layer and take the minute or so needed to retouch out the big reflection completely. Finally (since the two layers are otherwise identical), decide how far to reduce the opacity of the retouched layer. Then turn attention to #7 if needed.

Dan


Dan Margulis
 



On Feb 1, 2021, at 8:10 AM, Dan Margulis via groups.io <dmargulis@...> wrote:

The files don’t have people’s names on them, and were random-generator numbered from #101 to #141. As with past studies, we also have a “par” version, #142. 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.

There is now a #143 in the folder. Harvey Nagai kindly made a new version, a mathematical average of all 41 of our entries, exclusive of the par. It is interesting to compare the two, to get an idea of what are the characteristics that typify a good version of this original.

Dan


jorgeparraphotography
 

The interesting part of this challenge comes from having to decide-mostly in a subjective manner- what works for each one involved. Averaging 140 plus images to see how it looks is just one more variant, which could appeal to some and not to others.

If an internal calibration like the Custom White Balance is done in camera on the premises before the shoot, we would have a reference point as to how green is green and how yellow is yellow, etc, in a “naturally or neutrally rendered” file, and I dare saying that, even in spite of having such reference, each shooter/retoucher will vouch for whatever additional subjective changes that appeal to him/her, so each final result will still be different.

We are affected by our culture, values, education, training,our subconscious mind, emotional intelligence, psyche, etc in quite unexpected ways.

Best Wishes

Jorge

@Jorge Parra Photography
www.JorgeParraPhotography.com
Ph: 7 8 6 . 2 2 2 . 9 4 0 5

--
Jorge Parra
www.jJorgeParraPhotography.com ( http://www.JOrgeParraPhotography.com )
Miami


john c.
 

I totally agree. In the absence of having something to match, a lot depends on our subjective preferences, what we think of the subject matter, and who the potential viewer or buyer might be. In the current case of the three women, after I was done with it I thought that since this is supposed to be an image for stock photography, then maybe it might be better to go for a 'look' instead of a neutral correction, so I've been trying various color lookup options and found that many are very appropriate. Are these women in morning, afternoon or evening light? Is it winter or summer? Might they be in the open reflected shade of a pink building and so on. There are many options that could be correct and pleasing for different reasons, and even displeasing options can be legitimate. What statement are we making to make this image stand out and saleable for stock? It's not just about correcting it.

john castronovo

-----Original Message-----
From: jorgeparraphotography
Sent: Thursday, February 04, 2021 8:31 AM
To: colortheory@groups.io
Subject: Re: [colortheory] Hotel Lobby: Results

The interesting part of this challenge comes from having to decide-mostly in a subjective manner- what works for each one involved. Averaging 140 plus images to see how it looks is just one more variant, which could appeal to some and not to others.

If an internal calibration like the Custom White Balance is done in camera on the premises before the shoot, we would have a reference point as to how green is green and how yellow is yellow, etc, in a “naturally or neutrally rendered” file, and I dare saying that, even in spite of having such reference, each shooter/retoucher will vouch for whatever additional subjective changes that appeal to him/her, so each final result will still be different.

We are affected by our culture, values, education, training,our subconscious mind, emotional intelligence, psyche, etc in quite unexpected ways.

Best Wishes

Jorge

@Jorge Parra Photography
www.JorgeParraPhotography.com
Ph: 7 8 6 . 2 2 2 . 9 4 0 5

--
Jorge Parra
www.jJorgeParraPhotography.com ( http://www.JOrgeParraPhotography.com )
Miami


Rex Butcher
 

With the hotel lobby, we don't know the colour of the marble; so I took a average colour temp reading off the a/c unit (side and bottom), the gray inlay and the gray floor tiles.  I don't recall the range but set and average to 3100k and therefore assumed the marble, plant and door would be more of less,  'correct.'

"What statement are we making to make this image stand out and saleable for stock?"  I agree that we don't know the quality of the light, summer or winter, shade or not, reflective colour of the building, etc.  But I would assume that their black dresses should be as neutral black as is possible.  Any variation is certainly up to the individual, but my belief is that unless this was an 'artistic' photo (and it is not!), any library would want the blacks to be as neutral as possible.

Toodle pip

Rex


--
What if the hokey cokey IS what it's all about??


Paco
 
Edited

Hi! Mine is 131 and it is not as washed out as it is here. The problem lies with how C1 outputs the JPEG file. When opened in PS it opens as a flat RAW file. Dan and I tried to find out why and how to remedy it. Nothing I did in C1 worked and Dan was able to figure out what the problem was but too late for doing anything about it. After searching in the Capture One site's forum I found out that it is a problem which they know exists but are not willing to do anything about it. The solution is to alter the PS RAW preferences which is something not very convenient.

Anyway, the way I approached it was to find a white point and contrary to what others have said, I easily found it in the left side of the marble floor. Once that was done, I proceded to fix the highlight in the palm leaves. The reflection on the wall was too distracting and served no purpose and as a matter of  fact, it detracted from the richness of the material. Getting rid of it entirely could not have been easier. Since most of the image is composed of warm colors, the red channel is an already made mask because the leaves are rendered as black. Curving the channel further cleaned out the whites and darkened the leaves even more and with a bit of extra touch up with black on some of the leaves, the mask was perfect. Then I used the clone stamp tool to fill in the highlight behind the palm's leaves and because the leaves are rendered as black in the mask it was very easy to do. The reflections from the lamps though, serve to highlight the materials reflectiveness and so I left them in.

The colors I pumped up to highlight the golds so as to enhance the feeling of "poshness." Then I retouched out the detail in the ceiling because it is something I automatically do when part of an architectural detail, when for advertising and not for the architect's work, detracts form the image. This then put me in the verbotem position of having messed with the image in an unacceptable manner as per Dan's rules.

So believe me when I tell you that it looks really good when properly rendered.

All the best!

Paco


john c.
 

I don’t agree that it’s not an ‘artistic’ photo. If it’s being sold as stock, it’s commercial art and can certainly have a bias even if it’s just being used as a background to a product in an ad. In fact, I did make the dresses neutral in my submission, but then I wondered if given todays trends, buyers wouldn’t prefer an orange/teal result as has been discussed here with other images. Warm flesh tones and bluish shadows is certainly a look that appeals to many buyers today. And that’s just one possibility; the oranges and pinks of Mediterranean sun is another. If the foreground product is a warm toned subject, say a flamenco guitar, then the art director could be looking for a subdued and cold look for this image. I think stock buyers are absolutely interested in a look.
 
john castronovo
 

Sent: Friday, February 05, 2021 9:38 AM
Subject: Re: [colortheory] Hotel Lobby: Results
 
With the hotel lobby, we don't know the colour of the marble; so I took a average colour temp reading off the a/c unit (side and bottom), the gray inlay and the gray floor tiles.  I don't recall the range but set and average to 3100k and therefore assumed the marble, plant and door would be more of less,  'correct.'

"What statement are we making to make this image stand out and saleable for stock?"  I agree that we don't know the quality of the light, summer or winter, shade or not, reflective colour of the building, etc.  But I would assume that their black dresses should be as neutral black as is possible. Any variation is certainly up to the individual, but my belief is that unless this was an 'artistic' photo (and it is not!), any library would want the blacks to be as neutral as possible.

Toodle pip

Rex


--
What if the hokey cokey IS what it's all about??


Hector Davila
 

Mines is # 115

Product and landscape is not my forte.

So, I just want to make it pop.

The client might exclaim, "What the hell happen! Why did you make the marble Blue?? Can you make the marble grey?"

I would say, "Sure, but I didn't make it Blue."

If you look at the left side of the pot on the floor you can see Yellow reflections.

I'm guessing the Blue must come from the blue light on the ceiling.

I like 'attention to detail'.

If you're doing a layout for a magazine ad, you got 'one second' to attract the consumer to your ad before they turn the page


I imagine when I look at that hotel lobby picture, me standing there in the lobby.

Try to take a one dimensional picture and make it 3 dimensional.

Now, there are Yellow reflections on the marble floor on the left side of the pot. It is there.

Should it not be seen?  If you are standing in front of the pot in the hotel lobby you can see it.

It means increasing the details of light and color.

It's just me, I like 'attention to detail'.

Hector Davila