Strange behaviour of DCP profiles (monohrome) for Ricoh GR3x

[pnajbar]

Hi,
I am struggling with the strange behaviour of monochrome adobe dcp profiles for the Ricoh GR3x. This affects any of the profiles, whether hard or soft monotones.
When I select one of the monochrome profile, strange artefacts appear on the image, especially in places with more highlights:

This is all the more strange as I have no problem with any colour profile. I also check this in Lightroom and Luminar neo - in those other programmes there is no such artefacts.

Any hints or guess how to solve this problem?

I am using Windows version of Photo Lab. this problem is present in both versions 6 and 7.

[Wolfgang ([ Win10, PL+VP+FP, Nik6 (+3) ])]

It took me a while to find this dcp-profile …

Why don’t you use one for another camera, that works? There are plenty in Adobe’s DNG Converter.

[pnajbar]

I want to use for the raw file to best match to the rocoh’s profiles, especially monotone and soft monotone.

[Wolfgang ([ Win10, PL+VP+FP, Nik6 (+3) ])]

… to best match …

is relative.

Post an example (raw-file) that you want a monochrome version of. Then I could check in PL6 or 7 (I’m on windows). – There are different ways to do so.

Do you also have a FP license (which one)?

[JoPoV]

@pnajbar

strange artefacts appear on the image, especially in places with more highlights:

Wouldn’t that be related to the raw highlight clipping and loss of datas demosaicer bug photolab has with some cameras ?

[pnajbar]

I am sending a RAW file and converted to jpg with only applied Adobe soft monotone profile (for comparison one jpg from DXO and another from Luminar).

RAW RX008365.DNG (29,6 MB)

Photo Lab

Luminar

[maderafunk]

I noticed the same problems when I was using monochrome Adobe profiles for the Nikon d750 with a Z6 camera. When I used the Adobe Z6 profile with the Z6 it worked. So somehow some profiles of different cameras are not compatible. But it should not be a problem normally, as it works in other raw converters.

[pnajbar]

Do you also have a FP license (which one)?

Yes, I have also Film Pack in the newest version.

[Wolfgang ([ Win10, PL+VP+FP, Nik6 (+3) ])]

@pnajbar

I checked your raw-file in PL7
please see VC 1 to 7 → RX008365.DNG.dop (72,8 KB)

---

First, the mentioned problem is NOT the culprit
( at least I couldn’t see that the existing overexposure was responsible for this ).

---

• VC1 with Adobe RICOH GR IIIx Soft Monotone.dcp
  
  → and it does NOT look good

• VC2 with Adobe Nikon Z6 II Camera Monochrome.dcp
  

• VC3 with Adobe Nikon Z6 II Camera Monochrome (Orange Filter).dcp
  

• VC4 with PL7 / Neutral B&W
  

• VC5 with PL7 / Neutral B&W + DxO FilmPack / Orange Filter @ 100%
  
  → play with the Density

• VC6 with PL7 / Neutral B&W + Channel Mixer
  
  → play with all sliders to adjust the colors’ brightness

• VC7 with PL7 / Black & white film or Digital film +/or Channel Mixer
  
  → play with the choices

---

A workflow like in VC6 is fast & simple … shouldn’t you want to mimic ‘real’ B&W film.

have fun & explore,
Wolfgang

[noname (Captain Obvious)]

resulting tone curve ( internal + dcp ) →

something like this… I will test on Ricoh raw, but it seems DxO code might not be compatible w/ Look table (that clamps everything to greyscale )? pure speculation and nothing more at this point

[…scratch that for now…]

[noname (Captain Obvious)]

I might have dcp fixed… somebody test this .DCP vs “RICOH GR IIIx Camera Monotone v2.dcp”

attached =
mono-fixed v1.7z (58.2 KB)

and let me know how it goes, as I am myself monochrome-blind, obviously

PS: I mean test in DxO PL…

[noname (Captain Obvious)]

showcase

original = https://postimg.cc/8J9hCSQp

[pnajbar]

@noname

and let me know how it goes, as I am myself monochrome-blind, obviously

PS: I mean test in DxO PL…

I tested on several images and yes, it seems that it works nice. I don’t know how you dit it but thank you very much!

[pnajbar]

@Wolfgang

A workflow like in VC6 is fast & simple … shouldn’t you want to mimic ‘real’ B&W film.

Thank you for your answer, I was afraid of using profiles from another cameras I thought that it might looks not good but actually as you advise I can see that it can be used like this.

However, this does not change the fact that this is a problem in DxO that is not present in other raw editors.

Thank you,
Pawel

[Wolfgang ([ Win10, PL+VP+FP, Nik6 (+3) ])]

As you might have found out by now, the last one (VC7)
was a classic B&W film simulation + Channel mixer

---

note

I even did not manipulate the overexposure,
which you can “cap” by setting → this value to 253.

Of course, it does not restore lost textures,
but black and white reproduction “loves” clear tonal separation.

To do this, use the color filter or, even more finely, the channel mixer.
What counts is the convincing result, not what any company idea offers.

[noname (Captain Obvious)]

@pnajbar

However, this does not change the fact that this is a problem in DxO that is not present in other raw editors.

Thank you,
Pawel

did you test the fixed DCP profile that I shared above ?

PS: I missed the post where you said that you did, sorry

[noname (Captain Obvious)]

@pnajbar

I tested on several images and yes, it seems that it works nice. I don’t know how you dit it but thank you very much!

tell me which Adobe’s monotone DCP profiles for Richoh you need to be fixed and I, obviously, will correct them and post here after working hours

[pnajbar]

Captain Obvious, you are too kind
However, since you have suggested it and you have time for this I’d be grateful for to more:

• RICOH GR IIIx Camera Hard Monotone

• RICOH GR IIIx Camera Monotone

These generally covers the range that I use.

Thank you,

[noname (Captain Obvious)]

no big deal, I already wrote a Matlab script ( you can do the same manually in a text editor or in excel or in whatever - it is a very simple fix ) to deal with the matter in semi-automated manner… the issue it seems is between how Adobe creates profiles and puts numeric data in HSV “look” lut table for hues math-rounding / bounds wise and how DxO deals w/ that data in there ( resulting in see-saw tone curve )…

I quickly tested a fixed DCP profile vs original DCP profile in ACR yesterday and visually they are identical ( and surely no artefacts in ACR for original DCP as in DxO PL ) - I will check output from ACR bit by bit after work to see that it is not only my eyes ( because there might be something that is visually too small to notice ) and post fixed profiles…

[David_McA (Scotland (Win 10, PL7, VP4, FP5e, Olympus))]

I wonder did you see my query to you regarding highlight clipping with DCP profiles here: LumaRiver Profile Designer v2 is out ... after a long while
I would appreciate your viewpoint, particularly if there is something I can do to alter the offending profiles so that they work in Photolab.
Thanks in advance.