Hello David,
interesting question, which I agree with.
I can only guess:
“Protect Saturated Colors” is just about saturation, “Reserve Colour Detail” is about ‘details’ in the colors.
Thanks!
Manfred
You’re right, it’s in the manual.
Interesting that the ‘Preserve color detail’ slider is not shown anywhere.
Hi Manfred,
while I’m on Windows it should be the same for you.
.
As you know for web use (and often for printing services) it is recommended to export as sRGB – if you care about others.
When exporting, DO NOT select “As shot” but explicitly select “sRGB IEC61966-2.1” to avoid a profile mismatch (you may have your camera set up other than sRGB or want to keep your work in a wider color space).
Your calibrated P3 display will show the exported file correctly.
.
Now, exporting with a ICC profile → … you get the option “Preserve color details”,
which when checked applies the default setting [ 50 ] from Softproof.
.
While when exporting with “ICC profile → Same as Soft Proofing”,
the option is checked / grayed out → to then apply the “Preserve color details”
.
from the Soft Proofing → “Advanced settings”,
where the “Intensity” is set to [ 50 ] by default, but here can be adjusted.
.
The above should clear up your first question …
.
while it has ‘nothing’ to do with printing & paper profiles. *)
*) paper profiles are NOT matrix-based ( → the Intensity slider / value is grayed out )
.
That is, this option
is there for convenience (e.g. straight export w/o Soft Proofing),
.
but use Soft Proofing when*) to have a correct Preview
- with Color Space deviations / changes
- to check Color Saturation vs Color Details/Texture **)
*) depends on several factors (pic, color space, monitor, settings, eye sight … )
**) talking about colored textures, not ‘textures’ from local contrast / sharpening
( and please see explanation / statements by @John-M )
.
To your 2nd question …
.
The “Protect Saturated Colors” slider ( as part of the “Color Rendering” ) → controls & affects the Color Saturation / Texture balance … and then also the subsequent export (be it with or without soft proof).
It may be helpful to consider the Export → Preserve Color Details setting as an additional correction when exporting to a smaller color space or to account for limited color and contrast reproduction when printing.
.
Lastly …
.
In the Nik Collection, there is no Softproof and Color changes can affect Color Space and Color Saturation / Texture balance. … Why not to check the processed pics in PL to then finally export (or print).
have fun, Wolfgang
(and sorry for this longish answer)
addendum
for more details about Nik please see → here …
1 Like
Hi Wolfgang,
Thank you very much for your detailed answer! No need to apologize, I really appreciate your illustrations and explanations. I’ll have to read through them carefully tomorrow, my head is already full today.
Thank you and best regards
Manfred
Continued, from above …
Addressing more of your points, Manfred;
In this case, you’re exporting for Paper-&-Ink (not to a display device).
So, the best approach would be to ask your print lab for an ICC-Profile that best describes the printer+paper that they’ll be using … and set-up Soft Proofing for that profile … so that, once again, it will be a case of WYSIWYG.
No change with PLv7 … DxO chose not to contrive to provide a WYSIWYG experience by default … One needs to have SP activated to achieve that.
- I’m fine with this - It just means (for simplicity) that SP should be activated for one’s intended output target (which, typically, is sRGB).
Regarding your question about the “Preserve Color Detail” check-box in Export to Disk;
@wolfgang addresses this question in his response
I’ll elaborate for some emphasis …
Preserve Color Detail refers to a DxO-specific algorithm (PCD) that determines how colors and details are “squished” to fit into a smaller color space (such as sRGB) … In the SP interface, the slider associated with this setting determines the strength or degree by which this algorithm is applied.
- I recommend using the default=50 in SP settings - because;
i) my experience is that this setting works well (I’ve never found a reason to change it)
ii) using this default avoids potential clashes with related settings for Export to Disk
-
In the case of 1) - the PCD algorithm is applied at its default strength (=50)
-
The the case of 2) - the PCD algorithm is applied at the same degree/strength as is defined in Soft Proofing … which may be other than the default.
-
If the “Preserve Color Detail” check-box is left unchecked then the PCD algorithm is not applied at all … but I cannot imagine why one would want that (?!)
See this thread, posted by @KeithRJ, for a process-flow explanation of how, and when, the PCD algorithm is applied.
Note: There’s a glitch/fault in the Soft Proofing UI that probably causes confusion for some users (esp. those coming to PL from “other” software);
-
When an ICC Profile is selected that targets a display device (such as sRGB) then, wrongly/redundantly, the drop-down control for Intent is left activated on the UI.
-
“Intent” is a legacy setting that predates the introduction of the PCD-algorithm - which was described to beta-testers, during development of PLv6, as being a proprietary algorith that’s a combination of Perceptual & Relative (the Intent options), providing a “smart” implementation that does not require the user to make a choice between one OR the other.
-
(The Intent dropdown IS relative ONLY when the selected ICC Profile is for a non-display/printer target … in which case, the PCD slider IS correctly de-activated).
-
DxO is aware of this issue - and, hopefully, we’ll see this fixed in an upcoming release.
NOTE:
The “Preserve Color Details” setting for Soft Proofing (which relates to the PCD algorithm described above) is NOT the same as the similarly named “Protect Saturated Colors” setting in the Color/B&W Rendering tool - - they’re each doing “different things”.
Addressing some specific points;
Yes - that’s probably what I would do in his case - - thereby enjoying the benefit of his P3 screen in “general” circumstances (say, in games or movies).
However, when using PL, he would therefore need to have SP activated (with sRGB target - - as per his export intention for consumption via a website) … so that what he will see within PL will then match with what he expects to see reflected from a website.
- Otherwise, he may be surprised and disappointed when a 5-star image (as it appeared on his P3 screen) looks “flat” when displayed from a website.
No - they’re (albeit, subtlety) different concepts … See above.
Manfred puts it well like this;
A good related point by Wolfgang too;
Nice !
Application of this slider (which determines the strength/degree by which the PCD-algorithm is applied) is subtle - it’s a “smart” algorithm, that’s applied according to the needs of each specific image … it’s not a linear result (unlike, say, exposure comp).
For simplicity, the PCD slider (in the SP UI) can be (reliably) left at its default setting = 50
Yep … Exactly !
I don’t know which working color space is used by the Nik tools - but I’m confident that, at very least, it would be AdobeRGB (the same as PL’s legacy WCS) … So,
-
When exporting an image to any of the Nik tools, I specify ICC Profile = AdobeRGB (with Preserve Color Details checked/ticked)
-
I re-absorb the result from Nik back into PL (with the AdobeRGB color space retained, as I understand it) - and then I Export to Disk via an appropriate ICC Profile (typically, sRGB).
Yes, I definitely agree.
Having the Export to Disk option for ICC Profile = As Shot will be fine, provided your camera’s Color-Space setting is exactly the same as your intended target’s color-gamut … but having it set that way has potential to bring you undone - with, most probably, mystifying results !
On the point of having Soft Proofing permanently activated (as I do, for sRGB output) … If it irks you that PL then puts a bright-white border around PL’s working display, note that you can change this via Preferences;
Yes, forgot to point out … when exporting from PL to Nik
.
in the PlugIn Selector → Export Settings
.
you get the same export options to choose from
.
including the (Export) PCD option
( = no effect when exporting with the same or wider gamut than the source )
.
It depends very much on what you are doing, e.g. …
-
when you invoke Nik to modify something
you may not want to restrict to a smaller color space that early in your workflow. -
when you export to Nik and choose a smaller color space than the source,
the smaller one is kept … with the PCD option to do ‘its magic’.
1 Like
First of all, I would like to thank everyone for the detailed and great answers! I now understand some things, or rather some things better.
This tells me that as long as I don’t adjust the intensity slider in soft proofing, it doesn’t matter whether I set ‘sRGB IEC61966-2.1’ or ‘Same as Soft Proofing’ in the ‘ICC profile’ export option. Of course, assuming ‘Preserve color details’ is checked.
Got it. 
This is great information! Let me think:
I edit an image in PLv7 in WGCS. When I export it to NikCollection, it stays in WGCS (as long as I set this up in the plugin selector - and the ‘PCD option’ has no effect in this case).
When the processing in NikCollection is finished and the image comes back to PLv7, the TIFF (which I use as the format for processing in NikC) is still in WGCS and I still have the advantage of not being “surprised”, as soft proofing is still set in PLv7. 
Correct?
Yes John, thank you for pointing that out. It’s important not to forget about that.
But this can only be a workaround. What does a user who is not aware of the situation actually do? Rhetorical question…
Thanks for your more detailed explanation and experience on this point - and on this topic:
I now have a better understanding of the diagram.
I think you referred to this circumstance in an earlier thread.
That was exactly my thought. Fortunately, I found this setting.
Wow, what a lot of information! I really appreciate the help and discussion, thank you all so much!
(And please, keep the ideas coming. )
Yes - true that PCD will have no impact … because, you’re not changing the color gamut (so, there’s no need to “squish” colors into a smaller working color space (WCS)
BUT, I’m not sure that the Nik tools can handle the WGCS (which is unique to pure-DxO tools - such as PhotoLab). I am confident, tho, that Nik tools can handle AdobeRGB - which is why I exchange files with Nik tools via that ICC Profile.
A rhetorical question that I’ll respond to any way !
Yes, that’s the point I was attempting to make in my suggestion that PL default to a SP setting that matches the user’s monitor’s capabilities (which, typically, would be sRGB) … but, that proposal was not taken up.
I guess the answer is that “Average Joe” who is not aware of Soft Proofing requirements also probably doesn’t notice differences between what is seen within PL and what he gets in his exported image (!)
Great questions, Manfred … Hopefully, this thread will be useful to others too.
John
That is a justified thought. Transferring an image to the NikCollection in AdobeRGB should not cause any “damage”, as we only had a maximum of AdobeRGB available up to PLv5.
Yes, indeed, you addressed this weakness of PLv6 (and still in PLv7) very clearly in your thread I referred to and vehemently demanded an improvement. So far, unfortunately, without success.
One thing, however, that I only became aware of through the discussion here:
Suppose I work with the WGCS and have soft proofing permanently activated.
I set sRGB IEC61966-2.1 as the profile for soft proofing.
This allows me to see on the screen what should also be visible on the Internet, as in my case.
Under these circumstances, what is the advantage of working in WGCS?
I mean, even if the WGCS can display so many and very saturated colors, I can’t see them due to the permanently switched on soft proofing (with sRGB IEC61966-2.1, for example), because this limits the colors. So I could set the legacy color space right away.
Where is my fallacy?
That would be fine.
something to consider …
DxO’s WGCS is their own ‘spawn’ (for additional input …), but not a widely used color space. Returning to PL with such an intermediate Tiff file doesn’t (shouldn’t) cause any problem as long you don’t use it “outside” PL ( → add a descriptive suffix as a reminder).
Otherwise you might want to use Rec. 2020 or ProPhoto instead, while AdobeRGB does not fully cover P3, but is suitable for printing.
[ note – those paper profiles (2nd graph) reflect the color range of my paper & printer ]
Having Soft Proofing with sRGB IEC61966-2.1 permanently enabled can be a “solution” for sRGB users (screen, output…) and make their lives “easier”. Also, they could use the legacy color space, as long DxO will support it, and don’t mind a couple of disadvantages (has been discussed somewhere – just don’t remember in detail).
But the sRGB/Legacy color space isn’t for everyone. You may want to take advantage of the larger color space (like your P3 capable screen) or print (which is what I do) and see in advance …
The real problem for sRGB users is that they can’t “see better” and therefore have difficulty following what’s happening when using DxO WGCS as their internal color space.
Monitor & destination gamut warnings are helpful tools, but they cannot replace the impression you get when you are on a wider gamut screen and switching between the (VC) versions.
The problem these days is that it will be impossible to please everybody, when it comes to the colorspace they are using when watching images on the webb. All people using Apple devices (and there is a lot out there) whether they are using stationary computers/workstations, laptops, I-pads och i-phone phones will be using Display P3 and even Samsung is using a wider color space than sRGB both in their phones or their TV-monitors today.
So sRGB is not at all as totally dominant as it was 10 years ago. The world is moving on a broad front to wider color spaces than sRGB. I even use my standard JPEG-form made for 4K with Display P3 ICC when watching them om my Samsung TV and it works very well I think even if the TV-colorspace might be even wider than Display P3. That compromise is nothing that disturbs me at all.
Infact many Mac-users and even others that do not use any Apple products might prefer P3-workflows of other reasons as well. In my case I have seen the benefits sticking to P3 all through my workflows regardless if my pictures are going to be displayed on screen or be printed. It is a big advantage not needing to prepare and export several different files with different ICC:s. It is much more efficient and easier to handle and also eliminates the risk of by mistake (like many do) get problems when these files get mixed up by poor order in peoples picture silos.
I also happen to prefer prints from pictures prepared with Display P3 than Adobe RGB. Display P3 is about as wide as Adobe RGB but is leaning more towards yellow and red instead of the heavy green and blue bias often in Adobe RGB, which I often too much to that end for my taste.
The design goal with Display P3 is said to have been to get a color space closer to what the human eye really sees generally and maybe that is why I prefer it too but I can understand that nature photographers prefer Adobe RGB.
1 Like
The main reason is as I pointed out in an earlier response;
- That is, before the process of “conversion down” to a smaller color space, internal (working color-space) details are maintained at a higher level of mathematical accuracy … which gives this process, and the PCD-algorithm, etc, much more detail to work with.
- For a simple example; 8-bit JPG files are very susceptible to color-banding (when manipulated by an image editor) simply because they have limited detail to start with.
Regarding using WGCS for interchanges with Nik tools;
My concern here is that I don’t know whether or not the Nik tools can “fully consume” a file (say a TIF) generated via ICC Profile = DxO Wide Gamut color space.
-
If Nik is not WGCS-aware (which it’s probably not) - then how does it handle this file ? … At worst, it may assume it’s an sRGB file, and “trim” details accordingly !! !!
-
I’ve tried to check this, the only way that I can (visually !!) - but I’ve not been able to come to any confident conclusion.
-
So, I reckon the safest way to interchange images between PL & Nik tools is via AdobeRGB - which I’m completely confident Nik tools are able to properly consume.
Wolfgang: Do you know whether Nik tools can handle these wider color gamuts ??
(That would be really useful to understand … … Perhaps a question for DxO Support).
Some interesting points here - which I’ll respond to from my own perspective:
I haven’t yet progressed to printing my own images - I have used a photo lab a few times (with varying degrees of satisfaction !) … Mainly, I consume images on my own monitor, or share them with others (such as via my wife’s Android tablet) - - and I submit photos into my camera club’s monthly competitions (with some occasional success).
-
I have a very good quality BenQ monitor … it can reproduce 99% of the AdobeRGB color space (and, therefore, 100% sRGB).
-
So, in theory, I could use PL without soft-proofing (knowing I’m seeing the AdobeRGB color gamut within PL) … then I could export to disk with ICC Profile = AdobeRGB … and happily consume the result correctly rendered on my AdobeRGB-capable monitor.
-
BUT …
i) If I were to share those images with others then they’d be unlikely to see my images as I see them on my BenQ monitor … more likely, they’d be seeing a squished-to-fit-sRGB version created by a process that I had no control over
ii) My camera club requires that images be submitted in the sRGB color space … So, I would need to separately soft proof any such images (with ICC Profile = sRGB) and export accordingly. -
I reckon it’s just much easier and simpler, with a lot less hassle, to have PL set (via SP = sRGB) to the color space via which most of my images will be consumed (and judged) … and I suspect this is true for most others too.
For preparation for printing, would one not be soft proofing with an ICC Profile that described the ink & paper … rather than with any of the matrix-based profiles ?
(Or perhaps that’s exactly what you’re saying - but I’m misunderstanding ?)
I don’t follow your logic on this point, Wolfgang … I certainly DO advocate use of DxO WGCS as our working color space (even tho I’m rendering and exporting to sRGB) - specifically because of the much higher degree of internal accuracy that this enables (see above).
Here’s another analogy (supporting the argument for using the DxO WGCS);
- If we were to carry out a large number of separate calculations while maintaining, say, 5 decimal-places for each step - before rounding down to 2 decimal-places as the fine step … the result will be much more accurate (and almost certainly different) than if we used only 2 decimal-places for each separate calculation.
I doubt that.
- Working in a higher color space with the same bitdepth is less accurate. Expressed in wavelength the steps are getting bigger.
- This example with 8-bit only counts for sRGB. AdobeRGB is 16 bit for that reason.
George
Hi George,
My simplistic analogies may not stand up to technical scrutiny - but, my basic assertion (working in the widest gamut possible is always better, even when we intend to export to a smaller color space) is undisputable.
John
PS. Let’s not have this thread drift off into tangents.
Hi Wolfgang, for my part I do everything with DxO products. So I don’t leave the “ecosystem” and the “language” in it is (I hope) consistent.
Printing myself is not my issue. I only have an office laser printer here.
I completely understand, Wolfgang. The thing is, if I work with soft proofing enabled to always get WYSWYG while editing in PL(v6 and v7) (as John does and recommends), I’m forcing myself to limit the capacity of my display. That’s the point.
Of course I would like to use the WGCS in PL, but what good is it if I realize at the end of editing (because only then do I switch on softproofing) that the image does not fit my output profile (sRGB for Internet) and I have to make changes.
If I am editing images for printing, that is of course a different matter.
What John says in my next quote, however, makes sense to me.
Thanks for reminding me, John, you’re right.
There are a lot of things going on in my mind right now.
That seems to be the best option.
Very interesting views and opinions, I can only say thank you for dealing with my topic so intensively.
@John-m:
If you work with soft proofing enabled for output in sRGB (internet), do you create a Virtual Copy of each image as recommended by PL?
No, I don’t - - I don’t see any need or point of doing so.
(As you know) I have SP activated for all images (via my default Presets), and I have the background for Soft Proofing set to be the same as for “normal” images (via Preferences) … So, all my images have the same look; no differentiation.
I use VCs for their natural/original purpose; to create different instances of the same image - for comparison purposes, or whatever.
Thank you for your answer.
Me neither. So I wonder why DxO recommends this. Perhaps purely for security reasons.