Settings in camera for (raw + jpeg) and processing raw in PL


I have recently acquired a camera which has wireless features which enable transfer of JPEGs to a smartphone. To try this out I will capture the images as raw + JPEGs. I am concerned if the in-camera settings needed for JPEGs may affect the post processing of the raw files in PL.

When shooting raw alone I turn off high ISO noise, Lens Aberration Correction and Auto Lighting Optimizer. I assume that that PL ignores the Picture Style which includes settings for contrast, sharpening, saturation etc. which cannot be turned off in-camera. I usually set white balance to auto.

For simultaneous capture of raw and JPEG I expect to turn on the in-camera settings (except for Auto Lighting Optimizer which I never use) and to choose an appropriate Picture Style.

It would obviously be convenient if I did not have to adjust the in-camera settings when changing between using raw alone and raw + JPEG. I wonder if this is necessary or if PL ignores in-camera settings made for the benefit of JPEGs?

(Xavier Rival) #2

It is hard to give a 100% sure answer without knowing your camera model, but the point of RAW is usually to have a digital file that is as pure as possible, and not affected by camera settings. So, I expect that the PL will indeed be unaffected by any of the in-camera digital settings such as sharpening, saturation, picture styles, etc. All the cameras I ever used (Canon and Leica) work like this.
On the other hand, white balance is written in the EXIF tags, and used by all RAW processors; you can still override it, but if it is already done correctly by the camera, you have less post-processing work left to do.

1 Like

Thank you. I should have mentioned that my camera is a Canon M5.
Does this make any difference to your reply?

(Alec Kinnear) #4

As Xavier mentioned, there can be some variance. Sony’s Picture Styles (like Autumn Leaves) do affect the colours out of the box when you open the RAW file but the RAW file underneath remains intact. I.e. you should experiment with your camera and these settings and see if it makes a difference with your RAW files. It would be great if you could come back and report the specific results.

I too turn off any lens correction in camera and leave it for PhotoLab. I also try to avoid any lenses which have serious distortion or corner flaws which are software masked. Most of Fuji APS-C lenses are designed for software correction, as are the MFT lenses. Canon EF and Nikon F have traditionally been much better about not requiring lens correction to look good. Not sure about Sony full frame – though I shoot Sony I’m still working on what auto-correction happens with their own lenses. With my Canon and vintage lenses I can throw them on Sony full frame mirrorless and see what happens without correction. Turning off all lens correction on a Fuji, Sony or Panasonic camera with its own lenses is considerably more challenging. DPReview even has trouble figuring out these issues sometimes.

Roger Cicala over at is about the only guy in town setup to do optical level tests on all lenses regardless of built-in lens correction. Over at (formerly, the reviewers do their best to separate auto-correction from native qualities but they don’t have a magic wand to turn off auto-correction on Fuji, Sony, Panasonic or Olympus lenses. With Canon EF/Nikon F, they can throw the lens onto a mirrorless body to see its optical qualities.

Looking forward to your field report!

1 Like


Thank you for sharing the links. I was not aware of the effort by Mr. Cicala to carry out optical level testing on lenses. Very interesting!

1 Like
(Marie) #6


in-camera settings like sharpening level, picture style, etc. are stored in RAW with Exif or MakerNote tags but only RAW converters from makers (like DPP for canon) will apply the same adjustments as it is done on JPG images from the camera. It’s not specific to DxO and this is because there are numerous in-camera settings with so many possibilities so it would take much too long to try to mimic them in softwares like DxO PhotoLab. Also, purpose of working in RAW is to be able to create your own rendering.

About lens correction in camera, in DxO PhotoLab there are 2 things

  • if they can be turned off DxO lens profile (module) will be generated with it off
  • if they can’t be turned off then we will have 2 lens profiles, one for JPG images (because there is always some default left) and one for RAW images (because when we process RAW we don’t apply in-camera correction).

I hope this answer to your question.


(Alec Kinnear) #7

Marie, thanks for sharing the official position. The issue many of us are facing is that camera makers to save money on building properly corrected images in the first place are forcing software correction even on RAW. But finding out which lenses and manufacturers bake optical correction into RAW is outside the scope of DxO’s work. Your stuck with whatever information the manufacturer includes in the RAW. It’s certain that some manufacturers (including Fuji which I sometimes shoot) embed optical corrections in the RAW file.

DPreview talks often about this issue, magicia. It would be great if you were able to share the results of your independent research on software correction with us. Baked in lens correction is definitely a fascinating issue. I choose to vote with my money: avoid software corrected lenses and manufacturers who burn lens correction into RAW.


The situation is evidently not as simple as thought. If I understand correctly for most cameras PL ignores the in-camera settings except White Balance. As Marie points out, built-in lens corrections are an exception. The review in Optical Limits (thanks for the link, uncoy) confirms that this is not the case with my Canon EF M 15mm-45mm lens, so one less thing to worry about.

A quick check, under less than ideal conditions for a “scientific” experiment showed that raw images developed in PL from my M5 with and without the in-camera lens corrections switched on looked the same. Shots taken with and without the Picture Style on in-camera also looked the same, including the Monochrome Style. Monochrome style was displayed for about 1/4 second the first time the image was opened in LP, then changed to colour. Perhaps the jpeg thumbnail in the raw file was displayed briefly? It is possible that the same happened with the other Styles, but I did not notice because the differences in colour were so small. While certainly not as reliable as they could be these results give me no reason to believe that raw files from the Canon M5 with EF M15mm-45mm lens would behave any differently from the manner described by Xavier and Marie.

Thank you all for your detailed answers to my question.


1 Like
(Xavier Rival) #9

No, it does not.
I never had an M5 in my hands, so I cannot say for sure. But I expect it not to apply any of the in-camera settings to the RAW file in any non-reversible way. The best would be of course if an M5 user could confirm.
As mentioned in other replies, you can also do some testing by comparing results but if you do so, you need to control the environment (e.g., no changing light due to passing clouds, etc).

(Marie) #10


in order to provide a correct image when you put a RAW in PhotoLab we obviously need to consider some parameters as any RAW converter otherwise image you’d have at first wouldn’t be usable :slight_smile:
Parameters we use and you have some kind of access are:

  • White balance as shot, but you can change it the way you want in WB pannel
  • ISO to denoise image correctly but you can adapt denoising
  • Crop to fit dimension of JPG from camera when shoot in maximum nativ resolution so you loose nothing

@magicia we can display briefly the JPG embedded in the RAW file at first import of the image that’s why you may have seen a moment capture style of JPGs.




My concern was that parameters set in the camera such as Picture Style, sharpness, contrast etc. might used by PL. Thank you for your clear statement that this is not the case.




The changing environment describes the “less than ideal” conditions I mentioned as limiting the value of the quick test I did yesterday.

Your original statement and Marie’s confirmation would seem to
make further testing unnecessary.