Choice of Legacy or Wide Gamut and export sRGB or Display P3? and softproofing or not?

Did anybody watch the third linked video Tutorials zu Affinity Photo für Desktop-Computer from beginning to end :innocent:

He tells exact the same story.

No use to edit in a bigger colorspace then the supposed export size.
My test shows this bleeched image he tells about.
1 i edit colors inside WG. (so dxo renders all colors of the rawfile in WG or shows out of gamut warning with scpt, sun and moon and the histogram.)
2 when i just export in sRGB, dxo’s rendering intent convert this WG-image as they find suitable to jpeg sRGB. (same as v5 but then from WG.)
3 when i want control of this action i use softproofing. (turn it on and i see by red later which parts of the image will be touched by this rendering intent. Do nothing and just hit export same as 2 happens.

The problem start when you alter your image accoording to that red warning layer.!!
Desaturate in working space means your shifting the colorvalue distribution of the rawfile to work image. That’s why you images turn up pale.
By overruling the relative and perceptual compression and change the hole range of the color distribution your image has all color values inside but it’s not the captured color distribution any more.(that was based on the latent image of the rawfiles sensorcolorspace.)

So to put this is perspective:
Legacy: Sun moon, histogram and SCPT react on all colors out of gamut based on AdobeRGB. YOU are more in control of this out of gamut color section. Leave it outside Adobe or squeeze it inside by using the visiualisation of sun moon, histogram and SCPT.
Closer to your screen sRGB so much more what you see is what you get. Use monitor tab to see difference of AdobeRGB and your screen srgb gamut.

Wide gamut?
Sun and Moon and such are quite useless and blunt as gauges because export rendering intent is dealing with a much larger piece of the imagecolors when export is sRGB.

What’s the point in living in a house, only to have to end in a coffin?

The bigger space allows manipulations that could lead to lost colours in a smaller space.


This is the main question.
Do we?
Lets assume we have one conversion (demosiacing) from raw file to working file.
Yes then your correct.
But as we think we know dxo does realtime adjusting from the start file.(sensorcolorspace)
So changing highlight, shadow, tonality and WB does “work” with the complete colordata.
(see the SCPT reacting when you play with contrast and saruration.)
Turning down saturation shows that SCPT value gets less, so more colors outside the working space are usable inside.

That’s the first statement I believe…I think in my coffin I will have a lot of lost colours, maybe I will plan LED light inside :rofl: :rofl: :rofl:



( just fell of my chair – and sorry to be disrespectful )

1 Like

Dear Peter,

I’ve watched the video 4 time and I unterstand it this way

it’s better to work in RGB 16 bit Format and with the colour profile ROMMRGB to take advantage of all the colours
Display has to be capable to show this colours
then develop your Photo
because not all devices are colour managed or colour space aware>

  • exporting in ROMMRGB the color information could not be handled and the colours will look desaturated
  • so the best is to export in sRGB

@OXiDant and @Joanna
Thanks for the empirics and opening this discussion Oxidant and thanks for sharing your printing experiencs and reflections on them and the editing and soft proofing Joanna.

I have just bought a new Epson P900 A2 printer and started to use it and also tested the softproofing in PL 6 a little and I have to say that I will never use it IRL.

I fully agree with you that the important is that the colors, tonality and brightness I can see on my calibrated monitor, when editing an image, corresponds to what I get from the printer.

This match is extremely important, because if what you see on the monitor doesn’t sync very closely with your prints, then you don’t really know what you are doing.

I have a good sync between my monitor and my prints and am pretty pleased with what I get out of my P900 but what I see in the soft proof is mostly something duller and more desaturated. I could not understand why because it did not sync with my reality.

I had a plan starting a tread like this since the results from the soft proofing kind of disturbed me. So, thanks all of you for your contributions that help me eliminate that mental un sync. :slight_smile:

1 Like

right to be clear, RGB is the pixelized rawfile data and 16bit is bitdepth (2^16 possibilities) regulair jpeg is 8bit format.
and ROMMRGB is an other name for the same colorpsace as Prophoto.
So he said use as big as possible first conversion towards a Working (edit) space.
translation: 16Bit Wide Gamut.

No screen/monitor is capable of reproducing Prophoto only some printers.

That aside:
i stil swirle around about does DxO cut off colors or does it “float” the working colorspace inside the bigger box that is called Camera/sensorcolorspace? What i mean if i drag a certain color saturation back does it drag “out of gamut of that color” inside my working space?

here you see the explaination of the horsshoe.

And watch this for how typical RGBscreens create color.(it’s rather simplified but it shows about 8-10(-16) bitdepth difference

Maybe this makes it more visible what i try to say: (i extracted this from several posts and answers of dxostaff.)
The old before wide gamut colorspace engine of dxopl

rather “simple” flow chart.
one size fits all like system.

this became this:

(i assume they ignore WG rendering and go straight to Legacy (adobeRGB))
The hatched blocks are the öut of gamut colors which are handled by Protect saturated colors tool in colorrendering.

and the bigger one:

sun and moon and Histogram change when changing working colorspace. (That’s the main problem for some to understand.)
Step 5 is more or less a 16bit Tiff file in Wide Gamut color space. denoised and WB and CA and sharpenend.(microcontrast applied)
And yes microcontrast isn’t applied on your screendump same as sharpening and prime denoising.
when zoomed in above 70% then the “jpeg version” HQ-denoising, sharpening and microcontrast AND CA is applied. changing WB means all this is done again from scratch (step 1 to 2.)

i hope it’s more clear why i think it doesn’t matter if you use Legacy or WideGamut on Rawfiles when you export a jpeg as end result. Even a 16bit tiff adobeRGB doesn’t matter.
Aslong as you don’t export as 16bit tiff or DNG to a second editor.
For those people WG is a big step forward.

Sorry for my not so exact statement
is “…ROMMRGB to take the closest advantage of all the colours…” better?

but this are the terms James Ritson uses in his video und sets in the dialog

so who should i believe

When working in color spaces with such a large gamut, it is **recommended to work in 16-bit color depth to avoid posterization** effects. This will occur more frequently in 8-bit modes as the gradient steps are much larger.
I didn’t say he was wrong :slight_smile:
i did say it’s also called (ProPhoto.) and bitdepth is about “smoothness” in the color hue.


But it’s good to read also in other forums that a lot of people are asking the same questions, jumping from article to article and remain perplexed

Thanks for your patience

1 Like

No problem, i think everyone is confused in this area.
It’s that complicated that every different viewpoint.
(In dutch aanvliegroute) causes a WTF i thought it’s…

I still don’t know exact how dxo does the conversion between rawfiles data and the working colorspace. The under the hood system. (and they never will tell us because that’s holding the primedenoise en optical module secret.
I gues continues demosiacing process where WB and thus CA reversed change the settings of the proces and along other parts. So that’s why i think that it doesn’t matter if you use Legacy or Wide Gamut as working place.
Legacy is better in normal export as jpegs for general use. (the visualisation gauges for interpreting which way you should go in edit are better connected(scaled) to the end colorspace. (So no need to softproof then.)

One article of whitewall, where i also order prints sometimes Softproof und Druckvorbereitung - tolle Expertentipps | WhiteWall
Am Ende speichern Sie die Fotos im richtigen Format. Während es wichtig ist, während der Bildbearbeitung mit 16Bit und einem möglichst großen Farbraum zu arbeiten, halten Sie sich beim Speichern der Fotos an die Vorgaben von Whitewall. Deshalb konvertieren Sie den Farbraum nach sRGB und stellen die Farbtiefe auf 8Bit, bevor Sie das Foto als TIFF oder JPEG speichern(Sorry no deepl at the moment)

The last one for today
Good night

checked for your paper profil and put it in the software, that I still use for those kind of stuff
(hardware = no more supported → outdated driver = another way to force customers to ‘upgrade’)

both papers exeeding AdobeRGB colour space

while ProPhoto covers everything

the green line = Canson Platine Fibre Rag w/ custom profile

1 Like

Agree. Another thing is, that it’ s almost impossible to actually see how much/far a certain colour is out of gamut (and thus, wether it matters much at all). Could be almost in gamut (and hardly noticeable later), or completely off (and very much noticeable later).

The new “out of gamut” buttons on the Histogram (bottom row, RHS) are intended to provide that info;

Note: The proffered shortcuts do not work, tho.

John M

Thank you John. In your example it looks like the red and blue are out of gamut. Histogram also indicates that. But…is it possible to see how much these colours are out of gamut? How far is the blue out? How far is the red out? Just slightly, or maybe very far? Lighter or darker reds/blues in that “gamut checker”, possibly indicated by numbers, would be of great help. But maybe that’ s the case with the new “out of gamut” buttons?