Indicating First Stage OOG Colours

According to this text, PhotoLab has two stages of colour conversion:

  • First stage conversion from sensor to working colour space, governed by the Color Rendering tool and its “Protect Saturated Colours” (PCS) slider
  • Second stage conversion between working and output colour spaces, governed by the Soft Proofing tool and its “Preserve Colour Detail” slider

While the second stage soft proofing tool enables the oog indicators, the first stage rendering tool does not, which means that we have to rely on DPL’s estimate of how PCS should be set through its magic wand function, or else go with trial and error, pushing and pulling the PCS slider while watching the preview.

If the position of the PCS slider is indeed meant to be the definitive indicator of oog colours in the first stage, I propose that this should be (and possibly is) communicated. Nevertheless, I often find that I have to push PCS to its right extreme to make an effect visible.

Proposal: In order to improve the situation described above, a feature that shows OOG colours in the first stage should be added.

What is the difference between the two, I don’t quite understand it.

From the documentation is says:

Protect saturated colors: The Protect saturated colors correction prevents some specific saturated colors from being clipped, which may lead to unnatural colors and loss of texture when a particular color channel is close to the minimum or to the maximum luminance intensity (0 or 255). This process is performed automatically; you can fine-tune or modify the result with the Intensity slider. Clicking on the magic wand restores the image to the original automatic setting.

Preserve color detail: this slider applies only to matrix based ICC profiles, so only display profiles – not printer profiles. Set at 50 by default, the slider lets you alter saturation and details in saturated portions. By reducing the value (to the left), the slider maintains saturation to the detriment of the details and, To the right, the details are preserved to the detriment of the vivacity of the colors. Note that the Export menu has a Preserve color detail checkbox which gives you the option to apply the slider’s effect to the exported images.

Is this not the same?

They are very similar indeed, except that clipping in the first stage translates to colours lost, no matter what you do afterwards…except that you could go back to colour rendering and shift the PSC slider to see if some lost colours and/or details might re-emerge.

Having a first stage oog indication could save us from having to re-iterate the PSC customising step, again, except in case of absolute perfection of PSC’s magic wand setting…which I find to be less than perfect. Anyway, one point is t know, what we sacrifice whan we set PSC to this or that position…

The “Protect saturated colors” intensity slider in the Color Rendering palette is an adjustment to the selected color rendering (defined by Category, Rendering, Intensity). Some colors can be rendered from RAW too saturated - sometimes to the point that there is clipping that causes detail to be lost in the monitor output or final export. The monitor out-of-gamut warning may be used to see parts of the image that might be clipped. The detail isn’t irretrievably lost if the camera captured it and you have a large enough working color space - though you probably want to change the rendering options if the colors are really bad. Raising the “Protect saturated colors” Intensity slider increases the adjustment to saturated colors so that (a) they are more within the visible gamut - less saturated - and (b) there is more microcontrast in those saturated areas so that tonal gradations are more visible.

The “Preserve color detail” slider is for soft proofing. It performs the same function on the final image, adding to whatever was done in Color Rendering.

Here is an example of all of this:

Note that the Soft Proofing destination out-of-gamut warning indicator is like the monitor out-of-gamut warning indicator in that it only shows what is being clipped in the selected color rendering. Adjustments made in Soft Proofing aren’t reflected in either warning indicator.

So I don’t understand this feature request. I believe its initial premise is incorrect.

Check out the white paper and look into the section How DxO’s reimagined color processing fixes the problem

The first stage (Protect saturated colors in the Color Rendering palette) has been reworked and improved compared to PhotoLab 5, the second stage (Protect color details’ in the Soft Proofing palette) is entirely new.

Note that the wording does not match for the second stage.

I still do not understand why there are two stages? Would one not be enough? And in that picture we also see only a final image, not two steps, so the description and the text do not really match?

Besides, so I understand it correctly that always by default that Protect Color Details is applied with value 50 when I export?

Personally, I’d rather have everything set to 0 and keep the original saturated colors, even if I lose a bit of detail. Well, it depends on the image, so having that option is nice. But if I don’t activate soft proofing, the Protect Color Details algorithm will be applied with value 50 and I will only know about it at export, is that correct?

In that example, it seems that you are applying both algorithms, and it has an even stronger effect. Basically you are going higher than 100, up to 200. But it still does not explain what is the difference between the two. And why two steps are necessary.

Because the output gamut might be different from the visible gamut you are working in (your monitor). The algorithm in Soft Proofing is to adjust the conversion during export. Let’s say you’re working on the image in the Adobe RGB or Display P3 color space. You have a wide gamut monitor that lets you do this. The working color space is larger, though, so it’s possible that some of the original colors need adjustment in Color Rendering. So you adjust Color Rendering to your liking. Now you want to export to sRGB so that the image is suited for display on the web. That’s a smaller gamut, so another conversion has to take place. You can manage that conversion with the Soft Proofing palette.

1 Like

Thanks for the explanation, that makes sense. But if I activate soft proofing for the srgb profile, and see that some colors are out of gamut. Why should I not make the adjustments with the Protect Saturated Colors slider in the camera rendering menu? What is the difference? And why do they have two different names?

Also in this workflow, what if my output colorspace is bigger than my monitor color space, e.g. with a good printer? Why should I apply the protect saturated slider first, to bring the colors down, and then apply the second slider afterwards in soft proofing to bring them back again? I could just do both in one step, or not?

Let us assume, for a moment, that you want all information that was captured by the sensor transfer to an intermediate image that allows you to decide, which parts to let go and which parts to take along.

  • Assume that the sensor has captured information that is out of (working colour space) gamut. This information would be lost completely if it were simply clamped.
  • Instead of clamping, the info from the sensor is scaled in order to make this info fit into the working colour space. Imagine letting some air out of a balloon in order to stuff it into a box (=gamut). Some air (=colours) will be lost in the process, but there’s still enough air in the ballon for us to not notice the difference. “Letting out some air” is performed in the colour rendering tool’s “Protect Saturated Colours” slider.

Okay, now we have stuffed the balloon into the box, the size of which corresponds to the working colour space. Now, we need to do something to stuff the balloon into an box that might even be smaller (e.g. sRGB) - and we can do this in a similar way as described above. Only now, it is initiated in the soft proofing tool - and the “Preserve color details” slider lets us decide, whether we want to trade intense colour for less detail or less intense colour for more detail.

As you write, this could all be done in one step or one of the steps could be hidden from the user, but from a resource and processing time point of view, it makes sense to split a long process into smaller parts, specially if the smaller parts don’t have to be recalculated every time. In DPL, the smaller parts have been chosen to look like this:

Usually, we don’t change denoising, WB and optical corrections a lot, which means that DPL can render images more quickly from an intermediate image that results from the conversion based on DxO Calibration and DCP profiling, which corresponds to the big box in the middle of the screenshot I took from here.

Note how customizing steps precede the conversion to working or output colour spaces. This helps to reduce processing losses in a process that starts with an extra large box (sensor captured information) and cascades down to smaller boxes towards the end.

…but the whole thread is evolving away from the original request: Make OOG visible in stage 1.


But that first step of squeezing colors into gamut should not be necessary anymore, I thought that was the whole point of introducing the wide gamut working space. That space is big enough to include all colors. I am sorry to stray away from the original question, but I feel the question cannot be answered if we do not understand the problem.

Read the whit paper I linked to above. Look for “clamping” to start with.

I read it. I also read the corresponding chapters in the PL5 and PL6 user guides. And there is no differentiation between both algorithms. If I have missed it, please show me the part where it says when to use the one and when the other algorithm? I have a feeling that the ‘Protect Saturated Colors’ algorithm is a left-over from PL5 and not required anymore when you use the white gamut working space. The Protect Color Details replaces it.

I don’t know what else to say that hasn’t already been said. One is for color rendering from the camera color space to the working color space to your monitor; the other is from the working color space (with all adjustments applied) to your output destination. One affects all rendering, the other is ONLY FOR SOFT PROOFING AND EXPORT.

You asked about printing. “Preserve color details” doesn’t even come into play when converting to a CMYK ICC profile from an RGB color space. Select the ? icon next to the “Soft Proofing” palette heading to read more about that.

But “Protect saturated colors” in the Color Rendering palette is always available and will have an effect if there are very saturated colors in the working color space that need to be displayed to your monitor and exported to an RGB image file or printer.

That is not so. It is true that you may use “Preserve color details” instead if you want. You can do that for RGB output (JPEG, TIFF, DNG). It won’t help your prints, though, and won’t affect the out-of-gamut warning indicators, whereas “Protect saturated colors” will because it is applied earlier in the pipeline.

1 Like

Thanks for all the explanations, I guess we can leave it at that, I think I can understand now how it is supposed to work, but I don’t find it a very elegant solution, it feels a bit like the software Darktable where so many things are to consider and one module is incompatible with the other. I prefer an approach where the software does the thinking for you behind the scenes.

Then turn Protect Saturated Colors all the way up (and keep Preserve color details at 50 or higher). This way you’re telling PhotoLab to do the thinking for you behind the scenes. The algorithms apply the adjustments only when PhotoLab thinks they’re needed.

1 Like

I rather set Protect Saturated Colors to 0, as in 90% of the cases I prefer the more saturated colors ;-).

I’m having difficulty in understanding your concern too, Mr. P

Changes to Color Rendering’s PSC slider (first stage) are reflected via Monitor OoG warnings
Or, am I misunderstanding what you mean ?

For an image where changes to PSC and PCD are quite easy to see (visibly), try this one.
To make changes very obvious, set PL’s zoom to 100% over one of the red leaves.

Yes - that last point is curious behaviour …

  • With Monitor (1st stage) OoG warnings activated, changes to Color Rendering’s PSC slider are reflected in OoG warnings (blue) - and in the Histogram (which, if SP is activated, reflects SP’s ICC Profile) … which is what I would expect.

  • Also, changes to Color Rendering’s PSC slider are reflected in OoG warnings with Destination (2nd stage) OoG warnings activated … again, as per expectations.

  • However, with Destination (2nd stage) OoG warnings activated, changes to Soft proofing’s PCD slider are NOT reflected in OoG warnings (red) - Tho, they are visible in the Histogram and in the image itself (with OoG warnings switched off).

I don’t understand why that behaviour is different/inconsistent (?)

Your balloon-in-a-box is an excellent analogy :+1: … However, just for clarification; “deflating the balloon” is not necessary for DxO’s Wide Gamut (“Huge box”) WCS … which is claimed to encompass;

… every possible color that a photographer might encounter in nature … [It] uses spectral colors as its primaries. It is big enough to contain all real-world surface colors, and it achieves this without imaginary colors — i.e., every combination of R, G, and B in this color space represents an actual color. {Quoting DxO’s Whitepaper}

And apologies if I’m perpetuating that drift - but I’m doing so in attempt to understand your request :thinking:


Not always …You have two opportunities to prevent PCD being applied;

  1. If, in export dialogue settings you specify “Same as Soft Proofing” AND Soft Proofing is activated, but its PCD slider is set to 0/zero


  1. If, in export dialogue settings you DO NOT specify “Same as Soft Proofing” AND the PCD check-box is not ticked/checked.

With the qualification that PCD is applied ONLY for display / matrix-based ICC Profiles (not for printer profiles). For printer profiles the Intent option can be used - for similar purposes.

In double-checking my expectations in this response, I discovered something curious;

  • I was expecting the result for “Same as Soft Proofing” AND Soft Proofing deactivated to be the same as for above (PCD not applied)… but that’s not the case (with my saturated-reds test image).

Edit: The reason for this unexpected outcome is explained here … In brief, it’s due to the image having been captured by camera with Color Space = AdobeRGB.


but SP deactivated nothing happens colourwise (thank god, that trap is gone).

For “Same as Soft Proofing” AND Soft Proofing deactivated … a “little bit” of PCD is applied !!

I can’t see this (checked with several saturated files). All I noticed that the histogram moved a little
when comparing to the exported as jpg.