Color Perception -- What Is Real?

[Norm Dulak]

After images are posted on BPN, we often see comments such as "there appears to be a color cast", or I see a problem "on my monitor", or "the color saturation seems too weak or too strong". But are such observations always correct?

Even if we all calibrate our monitors to international standards, there is no guarantee that what one individual sees will be viewed in the same way by another. That is of course obvious when one of the viewers is colorblind. But the fact is that far more subtle factors can cause problems.

There are likely to be perceptual differences of an object viewed by a 70 year old person and by a youngster. That's because as we age, there is gradual deterioration of both the physical components of the visual apparatus and of the neural system, that transmits and processes visual information. Then there is the matter of bias, both in our personal opinions of what is visually correct, and in the devices we use to capture images, whether they be film or digital capture systems. Even the coatings on the glasses some of us wear may have an effect. It very much becomes an issue of what is real and correct to different individuals using different image capture systems.

My question to the experts in this forum is what, if anything, can be done to make it more likely that what we perceive as a correct rendition of an image will be seen in the same way by others?

Norm Dulak

[Roger Clark]

Hi Norm,

I agree with everything you said. I'll add that as we age, the lenses in our eyes yellow, and then if one has cataract surgery, the new lens transmits more blue.

So then the cynical view. The vast majority of web viewers are using either fading CRT monitors or new cheap LCD that have too much contrast and saturation. Our images from calibrated monitors don't look good on either! So I just make it look good for me (calibrated monitor).

Roger

[Desmond Chan]

"there appears to be a color cast",

RGB values can help you out on that though. I check that myself.



or "the color saturation seems too weak or too strong". But are such observations always correct? That could be subjective and so is not about correct or not, I think.


But the fact is that far more subtle factors can cause problems.The environment can affect our perception of color. For example, the same color can be perceived differently depending on what the color it is appearing next to.



My question to the experts in this forum is what, if anything, can be done to make it more likely that what we perceive as a correct rendition of an image will be seen in the same way by others?
Is that possible?

The same photo doesn't look the same even to the same viewer when it is shown on a computer screen, printed out and hanged on the wall in a gallery/museum, printed and hanged on the wall of an office, etc. Is that "correctness" really worths pursuing?


Forgot to say: I am not an expert but I'm curious sometimes.

[Lance Peters]

Always two sides to every story - Isn't there :)
Agree on Both sides --- I think if you have for instances a white Egret and the feathers are blue and the RGB values indicate that - then it has a blue cast.

Regardless though in the end - IMHO - only one thing matters...

Did you have fun making the image and Do you like it!!

[WIlliam Maroldo]

I think that regardless of what appears on any given monitor, or any given eyes see for that matter, the color, saturation, contrast, etc. of an image can be quantitatively measured and we can do it easily in Photoshop or other image editors. Take the example of blown whites. Blown whites are always R256/B256/G256. We can download an image and analyse it, and this is not subjective, or an opinion. We can use curve graphs, color sampling, and other methods, not just in the blown whites example. In other words its not someone's opinion that the whites look blown on my monitor, or my eyesight ain't what it used to be. It is scientific fact. Of course ther are subjective aspects of images, most notibly composition, but many other aspects can be measured, and thus are truely objective. regards~Bill

[Norm Dulak]

I think that regardless of what appears on any given monitor, or any given eyes see for that matter, the color, saturation, contrast, etc. of an image can be quantitatively measured and we can do it easily in Photoshop or other image editors. Take the example of blown whites. Blown whites are always R256/B256/G256. We can download an image and analyse it, and this is not subjective, or an opinion. We can use curve graphs, color sampling, and other methods, not just in the blown whites example. In other words its not someone's opinion that the whites look blown on my monitor, or my eyesight ain't what it used to be. It is scientific fact. Of course ther are subjective aspects of images, most notibly composition, but many other aspects can be measured, and thus are truely objective. regards~Bill

Bill:

I suppose such quantitative analysis is one approach. That's something I should probably check more carefully. But it is complex and something I'm not presently very familiar or comfortable with, and I'm not sure that the more subtle differences that cause experienced photographers to proclaim one image to be a "wow" image while another seemingly similar image is not could ever be quantified. That's one of the things that make photography so interesting and challenging.

But, as in the case of careful monitor calibration, even if we adopted such a rigorous analysis, the perceptual capabilities and biases of others would remain a problem as they view our more perfect images.

Norm Dulak

[Andy Wai]

I think that regardless of what appears on any given monitor, or any given eyes see for that matter, the color, saturation, contrast, etc. of an image can be quantitatively measured and we can do it easily in Photoshop or other image editors. Take the example of blown whites. Blown whites are always R256/B256/G256. [...]
While 256 is certainly blow highlight, to me any large area of bright highlight that is complete flat is blown. So it can, and often do happen well before 256. To throw even more confusion into the mix, different monitors have different luminance sensitivity. Try looking at the pattern below on various monitors:

http://www.drycreekphoto.com/Learn/Calibration/monitor_sensitivity.html

And worst, blown highlights are not always bad thing. Spots of it here and there sometimes add sparkle to the image. It all depends on the image. So the objective analysis tells only one part of the story. For sure, you have to get the fundamentals right. But what it doesn't tell can be just as important.


[...] But it is complex and something I'm not presently very familiar or comfortable with, and I'm not sure that the more subtle differences that cause experienced photographers to proclaim one image to be a "wow" image while another seemingly similar image is not could ever be quantified. That's one of the things that make photography so interesting and challenging.
I think the wow response is very subjective and individual. Go to Flickr and look at images with a string of awards, some (emphasis the some here) of them you scratch your head and wonder what the fuzz is about. People have different responses. So you can please some people all the time, you can please all people some of the time, but you can't please all people all the time :).

One little observation here: The wow images that I saw on one monitor often will be wow images on any monitor at all, no matter how crummy. There seems to be something that make them transcend the limitations of hardware. Of course, being able to identify a wow image is intuitive, being able to look at an image, see its potential, and knowing what to do to take it to the wow level is much more difficult. I believe extensive practice in looking at images and quickly identifying their faults and strengths would be a good starting point. But even if you can do it, that ideal wow level is ones own subjective wow. Not some universal wow thing. I guess that's what makes it impossible to quantify...

Andy

[Norm Dulak]

I think the wow response is very subjective and individual. Go to Flickr and look at images with a string of awards, some (emphasis the some here) of them you scratch your head and wonder what the fuzz is about. People have different responses. So you can please some people all the time, you can please all people some of the time, but you can't please all people all the time :).

One little observation here: The wow images that I saw on one monitor often will be wow images on any monitor at all, no matter how crummy. There seems to be something that make them transcend the limitations of hardware. Of course, being able to identify a wow image is intuitive, being able to look at an image, see its potential, and knowing what to do to take it to the wow level is much more difficult. I believe extensive practice in looking at images and quickly identifying their faults and strengths would be a good starting point. But even if you can do it, that ideal wow level is ones own subjective wow. Not some universal wow thing. I guess that's what makes it impossible to quantify...

Andy

Andy:

You may or may not be right that a "wow" image at one place is a wow image everywhere. But I mentioned this, because following a posting I recently made, a BPN member sent a repost making some color adjustments that was acclaimed by others as a "killer repost." I liked the repost and the advice that accompanied it, but didn't feel that it was really such a dramatic improvement over my original posting. Was my perception of the two images in some way defective?

Norm Dulak

[Andy Wai]

Andy:

You may or may not be right that a "wow" image at one place is a wow image everywhere. But I mentioned this, because following a posting I recently made, a BPN member sent a repost making some color adjustments that was acclaimed by others as a "killer repost." I liked the repost and the advice that accompanied it, but didn't feel that it was really such a dramatic improvement over my original posting. Was my perception of the two images in some way defective?
For me, color balance is a rather relative thing. It's quite easy to look at two different color balances side by side and tell which one is better. But unless it's way out, it isn't always easy to tell by looking at just one instance. Also, it might be more of a cognitive rather than perceptual thing. So with training, one should be able to get better and better at pick out the color cast and how to counter it.

The other thing I find is, since color balance is a little relative, there is usually some room to play with. I have 5 different displays at home, including CRT and cheap LCDs. They are all calibrated by the same sensor and 4 of them uses the same software. No two of them look exact the same. Some of them has gross distortion in certain color regions. After working with them for a while, I know how each one would react in general. But the difference isn't just in color, it's also in the way the brightness scale is handled. The one I use as reference is the NEC 2180WG. But it's very easy to get into a situation where things look good on the NEC but look odd on the others. And often, it is possible to use the feedback from the others to tweak things on the NEC. The NEC always has the final say, but it's usually possible to have things look good everywhere. So that's my "trick".

Andy

[Robert Amoruso]

Norm,

Good points above. The short answer is No.

Many times I have received feedback on a critique that they don't see the color cast I am talking about. When I repost, saturation has a tendancy to get amp'ed up, etc. When I make a colro cast correction, I can readily see the difference as I toggle the changes on and off but they are hard to see viewing the repost and the original in the BPN thread.

The thing to remember is that critiques are subjective; meaning that personal preferences, monitor color calibration, eyesight, plus a whole lot of other factors govern how someone sees and looks at your images. If you are unsure of a comment, ask that person to clarify.

[Jackie Schuknecht]

Excellent link Andy by Dry Creek. My mediocre monitor did not too badly. It is a good fast test to check out one's monitor.

[Norm Dulak]

Many thanks to everyone! I didn't know what kinds of responses my thread would generate, but the results have been helpful to me and I hope useful to others as well.

Norm Dulak

[Jackie Schuknecht]

It is like having an encyclopedia at your fingertips Norm:) with all the knowledge we have here.