Soft Proofing

Thanks for the recollection, Luca.

Just one correction (as I understand it); the Protect Saturated Colors Algorithm (PSCA) that’s applied when Soft Proofing is activated, and when Exporting-to-disk, is not associated with the slider provided with Color Rendering (although, unfortunately, they have a similar name).

There is no slider associated with the PSCA that’s applied when Soft Proofing / Exporting … it is completely automatic.

John M

1 Like

I’m not an expert on color management, I’m just digging in it like most of us. I understood color management is part of the os. One can view and change the settings. Win 10.

It’sin Dutch, sorry :confused:

George

1 Like

yes, true – and a bit more in detail

  • The perceptual rendering intent compresses / squeezes all colours (all out-of-gamut + in-gamut colours) until they fit in the target’s color space. – This RI shifts all colours, but keeps the visual relation between them. Therefore it is commonly recommended and less critical for the average user.

  • The relative colorimetric rendering intent compresses / squeezes (only) the out-of-gamut colours until they ‘meet’ the border of the target’s colour space. – This RI leaves the ‘reproducible’ colours intact, but crushes those ‘outside’.


DxO uses it’s own ‘automatic’ method with export – see description by @Lucabeer in post #20
– and is not user configurable.

1 Like

Hi Wolfgang,

That is a very interesting post and much to try. I have certainly learned something From that (and the rest off this thread). Many thanks for making the effort - I often think DxO is as much about its great community as it is the software itself.

Colin

2 Likes

There’s no contradiction between my post and of @Lucabeer .
I just wonder that if soft-proofing means to see how an image looks at another device, how that interacts with an own rendering intent. If everybody uses it’s own rendering intent where is then that uniformity that color management tries to gain.
If I download a printer profile from my printservice for soft proofing then I expect that pl is using that one as it is and not changing it.
On export that isn’t important, the image is the end product. But here we’re talking about soft proofing.

George

I did do some further research. From The Manual of Photography.


The second diagram and its explanation is describing exactly whta happens.
The conversion between color spaces goes via a ‘profile connection space’.
I can’t get it rotated to portrait. :worried: :worried:

George

Hi George,
all good – and I was referring to @Lucabeer’s summarize, how DxO handles rendering with export.

Softproof is nothing else than an on-screen simulation, what to expect with export to a certain target / colour space. It is useful when colour spaces differ … and usually the last step after editing.

As explained, it makes sense to use a virtual copy for softproofing, to then adjust the proof version to taste (if you think it’s necessary) and export from there (e.g. with the sRGB IEC… profile) – without changing the master file.
While a sRGB screen limits to see the full range of a wide gamut pic, softproof gives the possibility to check (and adjust) for the export output.

Now, with DxO using an ‘automated’ not user configurable rendering for export (in case of sRGB / AdobeRGB / DisplayP3 profiles), one might think to get a visual discrepancy in softproof … from choosing the fixed rendering intents perceptual or relative colorimetric.
– But that is not the case. The same ‘automated’ rendering is applied to the softproof so that it looks identical to the equivalent export.
To check yourself, you can export from the softproof copy (“same as softproof”) and compare the softproof version to its output. – It’s the user interface, that is so far unclear. :frowning:

I just wonder that if soft-proofing means to see how an image looks at another device, how that interacts with an own rendering intent. If everybody uses it’s own rendering intent where is then that uniformity that color management tries to gain.

Well, it’s up to us,

  1. what hard- and software we use
    and
  2. how we interpret & edit the file before we export the pic – as recommended with the colour profile included, so that the colour managent can pick it up.

With an sRGB monitor you can check (and adjust) for sRGB output,

  • especially when working with wide gamut pics
  • before sending pics to your printing provider

If I download a printer profile from my printservice for soft proofing then I expect that pl is using that one as it is and not changing it.

PL is not going to change your printer profile, but you can adjust your file (–> the VC / softproof version), so that it looks good to you, before you export that one with the sRGB IEC … profile – that’s the usual way.


And to wrap it up – personally, I don’t count on other user’s devices, as everybody uses something different, from high quality calibrated screens right to smartphones, and not to forget, people have their habits (reasons) how to set up things – if they do.

Wolfgang

1 Like

I think you misunderstood me. Softproofing is meant to see how an image is reproduced on another device, not connected to my pc. That’s only possible when the used rendering intent is the same. There’s no room for pl to make it’s own rendering intent.

George

Yes. And also it’s worthwhile pointing out the obvious; that we can actually see the result of Soft Proofing for a different color space only if our monitor is capable of rendering that color space. (For example, we cannot see the result of Soft Proofing for, say, export to the wider P3 color space if our monitor is only capable of rendering the sRGB color space).

In this case, Soft Proofing is still a useful workflow step, as we can use the “Out of Gamut” buttons on the histogram to check how/whether the image will “fit” within the target color space.
image

Yes - that applies particularly (as Wolfgang then goes on to explain), when the intended target is a printer, or via an ICC Profile to a different Color Space.

For the most common case, however (where we are working with PLv6 on an sRGB monitor - - and exporting-to-disk via the sRGB profile - and consuming the result on that same sRGB monitor - and/or sharing with other common/typical users with sRGB monitors - and/or posting to the interwebs, etc ; where the expected color space is sRGB) - - - my personal approach is to activate Soft Proofing right from the start, and simply leave it on all thru my editing process. (In fact, I have it activated via my automatic preset) … It’s all just simpler that way, I reckon.

John M

Spot on :white_check_mark:

Actually, the implication of the behaviour that Wolfgang describes above results in that not being so;

  • PLv6 is applying an additional, automated “Protect Saturated Colors” algorithm (PSCA) when we Export-to-Disk … Note: This is NOT related to the same-named slider with Color Rendering.

  • Therefore, even if we’re doing everything on the same sRGB monitor, including Exporting-to-Disk via ICC Profile = sRGB, it is always advisable to Soft Proof if you’re expecting WYS-is-always-WYG.

See more details on this on my post just above this one.

Also see here for attempt by @KeithRJ & me (with help from others too) to map PLv6’s color management pipeline. Note, tho, that it’s a Work-in-Progress.

Regards, John M