Curious about saturation preferences: the poll

I was intrigued by Michael’s question last week (here) about saturation preferences, and it struck me that a poll might be a good way to figure that out. I prepared six versions of an image, each with modifications to the saturation or vibrance level, and each adjusted to be within the range that shows up on photonet (in other words, no version is “wrong,” particularly given the usual differences in calibration between monitors). I’ve also inserted copies of an ugly b&w version between the color images to cleanse the palette. As always, all discussion is welcome!

Which version do you prefer? (some are similar, so you can vote for two)


Version A



Version B



Version C



Version D



Version E



Version F

 

In making a selection, I asked myself which one I'd like to look at on my wall every day.Then I thought about Goldilocks.

 

Version A seems to me the most naturalistic while still having photographic power. Version A has enough drama without feeling like it's a Photoshop creation. I think most important to this is the blue of the sky, which stays in the most palatable realm to me. The less saturated versions both feel kind of washed out. It's hard to fine tune things when looking at photos against the awful white background PN supplies in forums, but I brought Version A into Photoshop because it looks a little bit like the mid section is coming across a bit "hazy." I prefer it with the tree line and the bit of road beneath it a little darker, while still suggesting the detail.

 

Version A for me as well.

Not sure if you're already aware of this, Leslie, but including the B&W version actually distorts the perception of saturation for each due to adaptation. And if it was possible posting in a grid arrangement would make it easier to compare. I had to keep scrolling up and down where I kept seeing 'E' & 'B' looking the same.

But I'm glad you posted this topic because I've been wondering what others are seeing on their displays. I even doubt myself and have several times had to go back and either reduce or increase saturation on images in forum threads and re-upload before edit option timed out.

 

I like version E.

From my monitor I think version A and D look very much alike.

 

Thanks, Tim. That’s actually exactly why I arranged it as I did. I know you already know this, Tim, but for others who are interested, this is an example of Chevreul’s concept of simultaneous contrast. Because the perception of a color is strongly influenced by the colors next to it, there needs to be a common surround if different versions are to be evaluated on their own merit. Otherwise, a low-saturation version surrounded by higher-saturation versions would look very drab indeed; and a high-saturation version surrounded by lower-saturation examples would look even more saturated than it actually is. This can be really awkward (like when I'm trying to figure out which image to use from the Lightroom line-up), but it can also be a very powerful tool—if I need to make a watercolor pigment look more saturated than is physically or chemically possible, I juxtapose that color against a drab version of its complement.

In this case, I decided that a neutral gray would make a pretty good common surround, so the mangling I did to the b&w was intended to turn it into a gray rectangle that was disguised as a photograph. I chose the size to make it more difficult for the viewer to see two color examples on the screen at the same time. My allusion to “cleansing the palette” was actually intended to be literal—it was the perceived color palette that I was hoping to reset using those blobs of gray.

I was looking for a good link to provide for more information on this topic, and I was delighted to see that the first one that showed up is one that has been a favorite of mine for years: here’s Bruce MacEvoy’s description of Chevreul’s work on the simultaneous contrast of colors. There's lots of other good color info on that website, too.

 

I am one of the 2 that prefers C so far.

Maybe it's why I enjoy shooting slide film but the costs is getting to me and I look at sunset forecast and chase the saturated skies. So it's not just a camera / software setting to me.

 

I don't ever want to tell other people what colors they "should" choose to use. One of the pleasures of looking at other people's pictures is being surprised by what colors they choose to use, especially if they are not the colors I would choose to use.

But what if this was my picture, you might ask? I have no idea which of the colorings I would prefer. It would depend on what my motivations were when making this particular picture. I didn't make the picture, so I have no idea which colors I would want to use.

If I were to take this or even one of my own pictures and edit it with no particular intent, just to see what "looks nice," to make it have "pretty colors" I would smack myself in the head and say "Stop that!" I think that letting oneself get into the habit of editing pictures to make them "look pretty" or have the colors "look nice" is the worst kind of laziness. What colors do I use? It always depends on my intent for a particular picture.

 

A&E (Arts and Entertainment, that's why I come here)

 

But that's not how we judge saturation during long subjectively influenced edits when we don't realize adaptation has kicked in from staring at one image much longer than judging objectively at an array of the same finished image each having various levels of saturation.

Editing is when the problem starts caused by adaptation. Quick comparisons remove the adaptive influence only leaving subjectivity. I was not influenced by the saturation of one image next to another in your OP because adaptation takes longer for this to happen at least for me it does. The grid arrangement without the B&W versions speeds this process up to reduce the adaptive effect. The interruptions going from one image to the next from having to scroll up & down impedes the subjective process.

 

From editing over one thousand Raw images captured in mixed and available lighting, IMO it's not that simple, Julie.

What I've outlined above about the adaptive effect in long image edits suggests that some folks aren't aware they've given their image an artificial look by over cranking the saturation making it difficult to derive their intent. Now maybe if we knew that they did do a reality check and walked away for a while and returned only to be shocked at the saturation levels and decided then and there to make a change, then we'ld know they should've smacked them self and said "Stop that"! Consistency in their body of work would tell us if they intend to make their images way too vibrant.

And of course there is the possibility that a curator chose the version from a contact sheet or supplied portfolio. I think this thread is important in this respect.

 

I agree, color saturation perception (and color balance) can be very subjective. I try to minimize that by turning my had around the room from time to time while staring at the monitor. Also, I review my edits several times between long breaks before deciding on the final version. One thing, in Photoshop, the vibrance slider gives better result with color saturation than the saturation slider.

I voted for version A, but I think it's shadow areas are a bit subdued for my taste (imagining how a scene like this would look like in person). I prefer the shadow areas of version D, but the sky tonality for me is better in version A. Since the topic is about colors, I think version A has it.

 

Probably why to me C, D, and F look virtually identical - I can't see them together and hence can't really determine their differences.

 

There you go ...

 

Good point. I'm not sure how this exercise gets to that defect ... but carry on. We shall find out as we go.

 

If you want to be Explored on Flicker, or Trending here, you need C or F.

 

Thank you. Still C, D, and F look almost identical. Earlier, I had A and D as my two choices; that choice is still valid after seeing them together. B might well be close to the real thing and E appears to be a slightly pushed version of it.

 

What I've come up with to control my perception of proper saturation levels is to take into account the character of lighting illuminating the subject. The saturation sells the light.

As an example the two images both have vibrant colors but completely different lighting both in luminance and white balance. These two images are next to each other within the same folder. I've spent so much time getting the saturation levels to look convincing due to my first working on the one on the left of the cups and then checking the saturation of the image on the right which always looks too desaturated especially in the tan skin tone of the girl due to the adaptive influence from the one on the left. I grappled with it just now before I posted it here. I still have my doubts on whether these two look right.

 

I figured that before the thread goes further into discussion, I'd recap results to date.

First, if you’re curious about the images, here’s what I did to get the sequence:

I selected a photo that I’d edited last February in Lightroom, made a virtual copy, and reset it without looking at it closely (Version B—the sharpened raw file). I took the reset version to Photoshop and reedited the colors in Lab mode to get it where I was happiest with it, then brought it back to Lightroom and adjusted the clarity and fine-tuned saturation and luminance in the yellows (Version A—my personal favorite). I made a virtual copy of that one, and reduced the vibrance to -25 (Version E; this turns out to be the closest to the final result I landed on last February). I increased vibrance on a second copy by +35 (Version D); and increased saturation on a copy of Version D by +21 (Version C). Finally, I made a copy of Version C and reduced saturation in the yellows (-40) and oranges (-18) (Version F). As I mentioned above, I was trying to keep all versions within the range that we often see on Photonet.

So here are the results so far, in order of increasing saturation:
B 2 E 6 A 11 D 3 F 0 C 3

The bottom line here seems to be that there’s a lot of latitude in what’s considered optimal saturation. At the same time, though, about two-thirds of the responses were distributed over one-third of the range of choices—there’s a general preference for images that are more saturated than a neutral raw capture, but not by a whole lot.

But note that there’s one very important piece of information that’s missing here: we have no idea what people are actually seeing when they look at the photos, because we don’t know which responses are based on calibrated monitors. Before I calibrated my laptop last year, I’d have thought that anything less saturated than D was abysmally drab. I would have needed to process images to the F level to arrive at something that looked like A on that particular screen.

Here's the sequence in order of increasing saturation, and I have another question for you---had you seen this sequence in this order, would your preferred image have been different?

 

Don't know. I never view images that way.

There used to be a color edit comparison module panel showing similar thumbnail size progressions but in a grid pattern in Photoshop that allowed clicking a radio button under progressed versions of the same image in hue, saturation and white balance. Can't remember the name of the module right off hand and don't know if it's still included in current CC versions.

I never liked using it because I didn't want to edit that way being visually bombarded assessing color with that many thumbnail preview in a grid formation.