Update on the vanishing blue.

[Manfred M]

Manfred,

A very clear explanation. thanks for posting it.

This thread began with the mystery of Brian's B=0 areas, and this thread has clearly documented that merely converting to a smaller color space can create this sort of thing. However, I think it is worth keeping track of what I showed in the middle of my tests of the sunflower image, which is that simple edits, in particular, contrast adjustments, can cause saturation problems. I first discovered this years ago when processing images of intensely red flowers. The final products had red areas in which detail had vanished, creating a plastic look. Part of the problem was unintended saturation increases. It was the first time I started using luminosity blend for contrast adjustments, in order to avoid inadvertent increases in saturation. For folks who are unfamiliar with this issue, here is a web page I found some years ago that explains it.

Dan

I absolutely agree Dan. There are three "root causes" to the issue:

1. Decisions made at capture that underexposed key parts of material that is naturally quite saturated. This caused loss of shadow detail in the captured material and moved the data into a range where it is close to the hull of the colour spaces, especially the Adobe RGB and sRGB ones.

2. Converting to a narrow (or medium) colour space without making mitigating edits (increase lightness and decrease saturation) ended up pushing more material out of gamut. In a workflow where part of the work is done in a parametric editor / raw converter and part is done in a pixel based editor, this can be considered a separate step. The separation is less clear in a tool like Capture One where both types of edits can be made at the same time.

3. As you pointed out additional edits that increase saturation and decrease brightness move even more material into where channel clipping occurs.


When Ted asked in #19 if I agreed that this was a saturation issue, I hesitated to agree because that really identifies the symptom rather than the problem. I would have been just as "correct" if I had said that is a rendering intent issue or an out of gamut issue; both are equally correct answers but not terribly useful to the reader.

Ignoring the exposure issue, it is a saturation issue because of the way that the relative colorimetric rendering intent works. This rendering intent brings all out of gamut values to where they just come into gamut, so one of the channel values will be set to 0 for all affected colours; giving a saturated value. The perceptual rendering intent redistributes all of the colours that are out of gamut, which effectively reduces the colour values sitting at the hull of the colour space. For the reasons Ted points out, this is not an option for on-screen display, but is restricted to preparing images for print.

[xpatUSA]

...

Another thing to note is that there are no areas of highlight clipping showing, which means you have some headroom on that side of the exposure. A raw file is often called a "digital negative". Ideally if you get a capture with no clipping of highlight or shadow detail, you are generally open to more flexibility in your workflow.

I'd like us all to be clear about the amount of clipped shadow detail in Brian's raw image. There is very little (0.2% and that only in the blue - none in the greens or the red). In RawDigger, I set an under-exposure warning level of 16. Looks like the raw has a bit depth of 4096 (12-bit) - that makes my warning level about 0.4%.

I'd say that there is next to no shadow detail clipping at the sensor in Brian's raw image file, even at 16 out of 4096 levels. That means that all the clipping shown by Manfred is either due to the conversion or to his selected warning levels or both, upon which we seem to agree.

Proof for the technically-minded:



The hard-to-see dark blue bits are the warnings.

[Manfred M]

Ted - unfortunately the RawDigger analysis is only going to be valid if you are using that software as your raw converter. There is no set standard on how the camera data should be interpreted and turned into image data. Different software uses different solutions, which is why a started with the same tool that Brian uses, Capture One. If you look at the areas that Adobe Camera Raw and Capture One show as clipped, the two pieces of software are not really aligned all that well. The only reason I chose to show the examples I posted using Camera Raw is that it has a feature that makes it easier to show.

I know a number of excellent photographers, both amateur and commercial that use Capture One because of the "look" of the files it produces. This is not a group that does not use Photoshop; they do. They just prefer a different raw processor. I use Capture One as well, but mostly because of the tethered shooting functionality. I find it to be a lot more stable than Lightroom.

[Cantab]

Having read all the posts, I'm going to provide my simplistic take away points for dummies (me).

The lack or near lack of blue is the result of two primary factors working both in combination as well as independently:

1. The underexposed original image did several things. But most importantly, it set the stage for the future loss of data in the pp process.

2. Editing was done in the sRGB colour space rather than in a larger colour space. Because of gamut and other issues, this resulted in the loss of digital data.

Is this a correct summary from a practical point of view?

[Manfred M]

Having read all the posts, I'm going to provide my simplistic take away points for dummies (me).

The lack or near lack of blue is the result of two primary factors working both in combination as well as independently:

1. The underexposed original image did several things. But most importantly, it set the stage for the future loss of data in the pp process.

2. Editing was done in the sRGB colour space rather than in a larger colour space. Because of gamut and other issues, this resulted in the loss of digital data.

Is this a correct summary from a practical point of view?

Good concise summary Bruce.

The only thing I would add is that conversion to sRGB before posting online is a necessary step.

[xpatUSA]

Ted - unfortunately the RawDigger analysis is only going to be valid if you are using that software as your raw converter.

I see. It is indeed unfortunate that examining the sensor exposure is not valid. I really must stop culling raw files based on that.

As you should know, RawDigger is not a converter and nobody uses it as such, so that puts me dead in the water since I have neither Adobe Camera Raw nor Capture One.

Perhaps I should look at the raw in RawTherapee and post screen shots from it?

[xpatUSA]

Perhaps I should look at the raw in RawTherapee (RT) and post screen shots from it?

Here we are ...

RT's raw histogram, just for the record, showing no clipping:



RT set to ProPhoto output:



RT set to sRGB output:



What we see here is similar to your examples with ProPhoto still showing no clipping and sRGB showing a lot.

I kept the clipping warnings turned off in these screenshots (after all, they are set by the user), so it is the histograms that tell the tale.

[DanK]

Having read all the posts, I'm going to provide my simplistic take away points for dummies (me).

The lack or near lack of blue is the result of two primary factors working both in combination as well as independently:

1. The underexposed original image did several things. But most importantly, it set the stage for the future loss of data in the pp process.

2. Editing was done in the sRGB colour space rather than in a larger colour space. Because of gamut and other issues, this resulted in the loss of digital data.

Is this a correct summary from a practical point of view?

Not quite, IMHO.

First, the underexposure, while an issue for other reasons, was not the cause of the problem that is in the OP: yellows that had blue=0. The yellow areas were not the areas that were underexposed.

Second, and more important, it is the conversion to the sRGB color space at any point that was the main problem. Even if Brian had edited in ProPhoto, saving/exporting the sRGB image at the end would have caused the blue=0 problem. That was what my example showed. I edited only in LR, hence only in ProPhoto, but as soon as I exported a JPEG, the blue in the yellows areas went to zero, just like in Brian's.

And third, with highly saturated colors, as one often has with flowers, even the process of editing itself--regardless of the color space--can cause these saturation-related problems.

So, I would summarize this way:

1. Don't underexpose. It can cause a variety of problems.
2. Independent of that, be cautious in editing areas with highly saturated colors. One way to do this is to use a luminance blend if you use layers for increasing contrast.
3. Moving to a smaller color space at any point can place some colors out of gamut. If you are ok with the appearance of the result, you can just leave it. I never noticed that at least one of my flower images online has blue=0 in the yellows. However, if you are going to print, it is helpful to softproof, which will show you--among other things--what is going to be out of gamut for the printer/paper combination, and it will let you experiment with ways to respond.

[george013]

There's no blue=0 problem.

George

[xpatUSA]

There's no blue=0 problem.

True!

255, 255, 0 = yellow
255, 255, 127 = yellow
255, 255, 254 = yellow

All are yellow: therefore, no problem.

[Manfred M]

I see. It is indeed unfortunate that examining the sensor exposure is not valid. I really must stop culling raw files based on that.

As you should know, RawDigger is not a converter and nobody uses it as such, so that puts me dead in the water since I have neither Adobe Camera Raw nor Capture One.

Perhaps I should look at the raw in RawTherapee and post screen shots from it?

Understanding the limitations of a tool is the important thing. It looks like something that is quite interesting for exploring raw data. The only issue is that the various raw converters effectively create an intermediate, proprietary image that they build from that data. It is only visible and workable within the software itself.

[Manfred M]

True!

255, 255, 0 = yellow
255, 255, 127 = yellow
255, 255, 254 = yellow

All are yellow: therefore, no problem.

The first one (which has no blue content) is something that is unlikely to occur in "real world" photography (but then neither are the other two examples). Monochromatic colours and light sources will be man made.

[JBW]

Fascinating, and much thanks to everyone. Here's what this old troglodyte has taken away from this exploration.

(1) photography and philosophy have at least one point of commonality. There are multiple ways to approach any problem and as many fine tune solutions as their are photographers or philosophers.

(2) I am going to have to live with the B = 0 reality in some of my shots. The day before i posted a macro of this flower and though exposed at -0 (I have both - and + 0) I still got B=0 in the yellow.

(3) non techies such as myself really shoukld limit the number of processing programs we use because of all the porential conflicts.

(4) the closer I shoot to neutral and the more I build my 'attitude' in pp the better my chances for a better shot.

(5) and perhaps most importantly for someone like myself whose likes and dislikes are a touch eccentric... a technically perfect shot is not always going to be the shot that produces what I see and feel.

Thanks again for all the help and please don't consider this a shutting down of the thread. I and I am sure many others are learning a lot.
Brian

[xpatUSA]

Fascinating, and much thanks to everyone. Here's what this old troglodyte has taken away from this exploration.

I am going to have to live with the B = 0 reality in some of my shots.

A small caveat for the day you shoot a really colorful red flower and discover to your amazement that not only B=0 but also G=0. Same old reason ... and unlikely to be that whopping -0.3EV "underexposure" that you favor - as you have already proven below.

The day before i posted a macro of this flower and though exposed at -0 (I have both - and + 0) I still got B=0 in the yellow.

Yep, comes as no surprise to me in spite of certain gloomy forebodings posted here ...

[DanK]

A small caveat for the day you shoot a really colorful red flower and discover to your amazement that not only B=0 but also G=0. Same old reason ... and unlikely to be that whopping 0.3EV "underexposure" that you favor - as you have already proven below.



Yep, comes as no surprise to me in spite of certain gloomy forebodings posted here ...

Right. As I pointed out above. the B=0 problem was not fundamentally an exposure issue. The yellows were not badly underexposed, and the underexposed areas weren't yellow. It's first and foremost a gamut issue. Or am I wrong?

[xpatUSA]

Right. As I pointed out above. the B=0 problem was not fundamentally an exposure issue. The yellows were not badly underexposed, and the underexposed areas weren't yet.

"It's first and foremost a gamut issue"

Agreed.

"Or am I wrong?"

Not from my POV, Dan

[Manfred M]

Let me respond to each of your thoughts in a way that give something else to think about

(1) photography and philosophy have at least one point of commonality. There are multiple ways to approach any problem and as many fine tune solutions as their are photographers or philosophers.

It's not just philosophy and photography, but rather just about anything we do in life. Some of the approaches might end up with "better" outcomes than others. The level of expertise and experience is likely to have a positive effect on the outcome.

(2) I am going to have to live with the B = 0 reality in some of my shots. The day before i posted a macro of this flower and though exposed at -0 (I have both - and + 0) I still got B=0 in the yellow.

I would refer that back to the previous point. The outcome will also be the result of expertise and experience. The one thing that does not change is that unless every step of the way is built on a solid foundation, the subsequent steps will be weaker than they should be.

(3) non techies such as myself really shoukld limit the number of processing programs we use because of all the porential conflicts.

Here I would have to disagree as the important part is to use the appropriate tool(s). Limiting oneself a set of tools that cannot deliver the outcome one is looking for is not going to work. It's hard to use a nail when the only tool one has is a screwdriver.

(4) the closer I shoot to neutral and the more I build my 'attitude' in pp the better my chances for a better shot.

I'm not quite sure what you mean here. When shooting, the most important technical aspect is to capture high quality raw data that gives you maximum flexibility in processing.

(5) and perhaps most importantly for someone like myself whose likes and dislikes are a touch eccentric... a technically perfect shot is not always going to be the shot that produces what I see and feel.

A technically perfect shot can be a very boring image. In addition technical strength, one also looks for an image that is well organized and one that has a strong emotional impact on the viewer. In general, a great image should have all three elements working together.

[Manfred M]

It's first and foremost a gamut issue. Or am I wrong?

I would suggest a "gamut issue" is often an outcome, rather than an indicator of an issue. It simply reflects ones inability to capture and process an image in a way that respects the colour space's limits while satisfying the photographer's intent.

There are conditions where the colours (or more specifically a colour range) in an image cannot be satisfactorily be reproduced in a low gamut colour space.

[DanK]

I would suggest a "gamut issue" is often an outcome, rather than an indicator of an issue. It simply reflects ones inability to capture and process an image in a way that respects the colour space's limits while satisfying the photographer's intent.

There are conditions where the colours (or more specifically a colour range) in an image cannot be satisfactorily be reproduced in a low gamut colour space.

Yes, I agree. However, my point was just to clear up a confusion that seemed to be cropping up in the thread, which was confusing the exposure issue affecting the darkest shadow areas of Brian's image with the B=0 phenomenon in the bright areas, which is what he started the thread with. Had Brian exposed half a stop more, the former would have not been an issue, but the latter would have been.

[JBW]

Yes, I agree. However, my point was just to clear up a confusion that seemed to be cropping up in the thread, which was confusing the exposure issue affecting the darkest shadow areas of Brian's image with the B=0 phenomenon in the bright areas, which is what he started the thread with. Had Brian exposed half a stop more, the former would have not been an issue, but the latter would have been.

Not that it matters ut I co0mpensate in steps of 1/3