Big differences of RAW rendering between Lightroom Classic CC and PL5

[Captain_Bumper (Fred M.)]

Hello to all,

I have an Olympus OM-D E-M1 Mk II and a Panasonic G9, I do more and more RAW with it (usually RAW+JPEG). I mainly use Lightroom Classic CC 11.5, and a bit of DxO PhotoLab 5.4 for denoising the most complex exposures. I used to use the “Edit in… Dxo PhotoLab 5” from LR’s menus which exported a TIFF that I corrected then in PL5. The problem is that the TIFF exported files are in 8 bits (thus loss of dynamics, the RAW being in 12 bits) or in 16 bits, and that creates very very voluminous files of which a part is useless (the TIFF cannot be exported in 12 bits).
So now, I use the External Module of PL which allows to transfer the picture from LR to PL5. Except that the RAWs are not displayed in the same way in the 2 softwares: in LR, they look good, but in PL they are very saturated, lack sharpness, with more chromatic aberrations and slightly less exposed… This is the case with the .ORF of the Olympus camera and the .RW2 of the Panasonic camera. PL5 is however announced perfectly compatible with both.

I have - it seems - no settings activated (all small switches are set to OFF) or automatic presets activated (Presets → None) in PL5, and yet it seems that it applies a modification of the images because for half a second, when opening the file in PL from LR, it seems that the image has the same aspect as in LR before becoming ugly (it’s hard to say, it goes so fast)…

I found on the internet someone complaining about a similar problem (but with just a sub-exposure problem) with a Canon R5, and it seems to be a setting on the Canon to avoid burning out the whites in case of highlights, a Canon setting that PL5 misinterprets… See here : https://www.chassimages.com/forum/index.php?topic=324727.0 (in french)

But in my case it seems that more settings are modified (saturation clearly, sharpness/peak reduced, over contrast/under exposure, chromatic aberration quite visible and exaggerated), and I don’t see what common setting between the G9 and the OM-D in their respective RAW could be misinterpreted by PL and give this result (if we assume that it is a similar problem with the Canon R5 RAW). Or maybe the problem is elsewhere…

I’m up for any advice/help.

Thank you.

My computer: 2019 iMac 27" Retina 5K, 8-core 3.6 GHz core i9 (9th generation), 40 GB RAM, AMD Vega 48 GPU with 8 GB VRAM (GPU acceleration is active in both programs) under Mac OS Monterey (version 12.5.1).

[platypus (Canon & Mac (Intel and M1))]

Welcome to the forum @Captain_Bumper

What you see is quite normal: Different RAW developers produce pictures that don’t look exactly the same. Comparing does not really help to discover which representation is correct and which isn’t…because “correct” does not exist without taking further precautions.

If you work an image in PhotoLab launched as an external module, DPL will not take into account any changes you’ve done in Lightroom. DPL will simply start from scratch, do its and your things and deliver an output file back to Lightroom, in which you can then continue your work on that image.

[rrblint (Mark)]

Hi and welcome,

Could you store an example RAW file each of an ORF and a RW2 file in a file sharing service(such as Dropbox, p-cloud, etc.) and post the link here, so that we can better assist you with this problem?

[IanS]

A couple of points for clarity if I have understood what you are trying to explain.

You cannot use the best de-noising techniques in DXO if you are exporting a Tif. Neither PRIME or DeepPrime operate on a Tif file. Both of these de-noising algorithms actually work on the raw data before demosaicing into rgb. Which is why they give superior results compared to programs like LR which operates on the demosaiced rgb data, but LR will, of course, work on a tif, jpg etc. The high processing power required to work on the raw data means that you don’t see the result of DeepPrime on the whole screen, only in the preview window in the Noise tool.

All raw converters apply default settings to a raw file. For example LR will show the noise reduction slider set to a default value, it’s a long time since I used LR can’t remember if it is zero or an actual number. This doesn’t mean that no noise reduction is being used but that a default number is used “under the hood”. Try opening a base ISI file and a high ISO file, both will show the same “number” on the noise reduction slider but will be using different “under the hood” adjustments.

If you have turned off Photolabs Default preset then you are crippling the program as it’s designed to use the Default preset when opening the file.You will have turned off chromatic aberration correction for example.

I suggest that you try opening your files in Photolab stand-alone, to avoid any impact from transferring from LR. As others have said any adjustments in LR are not present in DXO-PL when you open the raw file in DXO.

If you don’t like the default preset it is quite simple to adjust to taste eg increase contrast, saturation etc and make this your default every time DXO-Pl opens an Olympus file or a Panasonic with its own default.

Although I normally use C1, DXO-PL does an excellent job with Panasonic and Olympus files regarding noise and optical corrections.

If I haven’t understood correctly please provide extra clarity and I will try harder

[Captain_Bumper (Fred M.)]

@IanS : it’s not about TIFF or JPG files, but RAW files, without any correction done : initial aspect is slightly better in LR than PL (more blurry, more saturated and moreover with a framing/composition wider in PL than in LR or on Preview/Finder’s Desktop).

[IanS]

Yes, I appreciate it’s about raw files but in your OP you said you were sending Tif files from LR.

Is Photolab set up in preferences to use the DXO default preset to new files? If DXO has seen the files before and you had "no preset in Preferences then you will need to manually apply DXO Standard from the preset manager.

Adobe uses the lens correction data built into the raw file. This is the same lens correction that is used when you take a jpg in camera. Due to camera cpu power restraints and time the jpg correction has to rather basic and by designing the lens with residual barrel distortion, which is relatively easy to correct, coupled with a crop, is often used.

DXO calibrate lenses in their optical labs and produce lens profiles that, with the power of a desktop computer and unlimited time, allow the true focal length of the lens to be used by avoiding much of the crop.

Here is a comparison between Adobe (jpg embedded raw file corrections) vs DXO for the Olympus 12-100mm F4 lens. Note the significant difference in fov at the wide end for this lens.Using DXO effectively upgrades your lens

[Tbolt (Darrell Thompson)]

Adobe seems to crop a lot of the image. Before I came to PL I used Photo Ninja and that cropped even less than PL, but Adobe seems excessive.

[jorismak]

@Captain_Bumper

it’s not about TIFF or JPG files, but RAW files, without any correction done

Raw files don’t have a ‘correct’ look. It’s just captured sensor data, that is up to interpretation by the RAW software. One of the reasons people tend to like one piece of software over the other: They will simply not look the same.

Lightroom is notorious (specially in review circles) for always applying a certain amount of sharpening and denoising, even if the sliders at set to 0. Like they try to have each file have the same ‘base’ presentation in sharpness / denoising.

And they open with quite a lot of contrast by default, at least for what I’m used to.

If you ‘send a TIF from Lightroom to PL’, Lightroom is rendering the raw file as a tif, and sending that to DXO. So you have a bit of the base Lightroom look in it, even if you don’t want it.

The same goes the other way around. If you open the raw file in DxO and export it as a tif to Lightroom, you always have some of the DxO base look in it.

Software that has a mode or a special reset or something else to start with the most linear data possible (ON1 linear mode, NOT Lightroom’s ‘linear curve’, Darktable / Rawtherapee) can be seen as sort of neutral… but it’s not something you are going to like if you are used to Lightroom .

Right, next thing is that DxO applies it’s - quite good - base sharpening only if a real supported raw file is opened. So a TIF file from Lightroom will probably disable any kind of sharpening out of the box.

Another pitfall, is that DxO’s sharpening is only visible from 75% and upwards, I think. So when just opening the file and resetting everything to default, it might look softer, unless you start zooming in and the sharpening-preview ‘kicks in’.

(BTW, DxO as the first step, using the denoising, and then saving as a DNG ‘with optical / denoising only’ is a very fine method to then use that data in Lightroom to edit the look there, if you want. Just be careful with PL6’s Wide Gamut option…).

[IanS]

Adobe is simply using the embedded jpg lens correction. In DXO there is a slider where you can reduce the amount of lens correction to trade off fov for correction. You can also remove aspect ratio constraints which can be useful in some circumstances.

The embedded lens correction crop means you end up at 12mm focal length in the 12-100mm F4 lens example.DXO gets you a 11mm (no idea, never bothered to calculate ) or so lens - hence the "upgrade your lens comment. Obviously with lenses, particularly older lenses which relied on optical corrections only you don’t see much difference.

[Captain_Bumper (Fred M.)]

@platypus

Hello

yes, I know all of that but :

1. the differences are beyond the simple variation observed between 2 derawtisers
2. I did not make any change in LR when I transfer de ORF/RW2 file to PhotoLab via the external plug in (except lens geometry correction, automatically applied from the incorporated profile, and anyway, PhotoLab do exactly the same thing.

On an another forum some people mentioned a problem with colorimetric spaces between AdobeRGB (the one I use on my cameras) and P3 of my iMac’s screen. And it’s true that by choosing “iMac” profile in Monitor settings instead of “P3 display”, I see differences of rendering that are more important in PL than in LR, to such an extent that rendering in PL is closer to LR’s rendering. As if PhotoLab react or use differently this system setting from LR… I don’t know why picture aspect is changing so much in PL and not in LR or in Finder (files explorer) when I change the monitor profile.
The only “big” difference at this time is sharpness (RAW in PL seems more blurry, especially in vegetation) and geometry (pictures taken with my 12-40mm PRO has a wider framing/composition in PL than in LR…).

I will continue my searches, and if you I advices, I’m interested.

[platypus (Canon & Mac (Intel and M1))]

Had a go at @Joanna’s lobsters and compared the looks in DPL5, DPL6 and Lightroom. Here’s what we get when we do the same thing in all apps:

Note the profile I used in Color Rendering and the softproofing with AdobeRGB (where applicable).