Sony 24-70 II and 70-200 II different colour casts

Sony A1 ARW files through PureRAW (now version 3) before importing the DNGs to Lightroom.

Comparing the DNG to the ARW. I set them all to the same white balance settings which look good on both lenses for the ARW files

The same white balance setting on the DXO files shows the 24-70 DNG is greener, and the 70-200 DNG is pinker.

examples here

It causes a headache when trying to manage white balances across a set, even more so if I only run DXO on the high ISO files and have a mixture or ARW and DNG with both lenses!

I’ve sent a support ticket to DXO but haven’t not yet had a response, I wonder if there is a solution already out there?

Hi Dan and welcome,

Can you post the original RAW ARW files on Imgur or some other sharing service so that we can download them and test them ourselves. This way we can better determine what the problem is.

Hi Mark,

Here are two raw files zipped up


This may or may not be helpful, but exporting to DNG via PhotoLab 6 with both DeepPRIME and DeepPRIME XD gives no discernible difference.

Maybe that PureRAW on Windows is subject to the same ill effect that was discovered in PhotoLab on certain versions of macOS?

Anyway - having tested the two ARW files with DPR3 on Mac, I can’t see or measure any differences in WB (or any other hue) in the built-in previews. If I let Lightroom calculate full previews, I can find slight shifts of tonality, but no WB differences beyond ± 1/256. Even though two different lenses were used, I see no colour shifts that could not also have been caused by slight variations of lighting, after all, artificial light was used and who knows, what could have changed by the usual flickering of such lights.

Some differences (detail rendering) exist between DP and DPXD, which is to be expected.
But I did not expect to see these differences in the histogram like they show here:

DP (note the “combs” in the midtones part of the histogram)

DPXD (no combs)

Here is a video I just made of the process.

2 ARW go through DXO PureRaw
Import all 4 files to lightroom
Choose a custom white balance and paste it across the set

The 24-70 and 70-200 look the same (so at this point all would be good, if I wasn’t including DXO)
The DXO 70-200 is notably pinker than the 70-200
The DXO 24-70 looks a bit warmer than the 24-70, but typically it’s more obviously greener and would be if I had spent time figuring out a better custom WB to use.

The differences can even be seen in just the thumbnails in the bottom left of LR

This difference between lenses is also consistent across the entire delivered set of 91 photos (and probably all those that didn’t make the selection process) it’s not just a light flicker. It’s also not isolated to this one event.

To get the files to look the same, I need to manage a custom WB for each variation and try to get them to match each other.

It case the video is not clear enough i have done a quick export and uploaded the jpg images as a result of the process in the video

Consider that Sony has created two lenses with slightly different colour characteristics.
Consider that lighting changed (due to flickering and mixed bulbs) while you took the images.
If you want to make sure what’s going on with your lenses

  • capture a colour checker or reference object
  • make sure that you have non-flickering, non-mixed lighting or use a long exposure time or preferably, unaltered daylight.

Nevertheless…colour differences might also be caused by DxO modules. Only DxO can check that though.

Ok but I have shown significant differences between the colour of the files from a single lens and a single source file - which is not subject to different lighting or lens characteristics.

and the (direction of) difference between a 70-200 ARW file and it’s DXO counterpart, is different to the 24-70 and it’s DXO counterpart. Whilst the two 24-70 and 70-200 files match each other well without being DXO processed.

Here is a colour checker card photographed, just now, in natural light with the 24-70, the DXO file is visibly warmer, and the black borders in the original file look a pinker. Both images share the same WB values.

When I check the checkers in imgur, I see mostly differences within that ±1/256 range.
To view, I used Apple Safari Browser version 16.4.1 and that browser is colour managed.

Again, processing in DPR (or any other RAW developer) produces results that can differ from what you see in OOC JPEGs or built-in previews. There seems to be no way around that except and maybe by using the camera manufacturer provided developers.

Maybe you should get in touch with DxO support and see what they make of your perceived issue.

I noticed similar differences with an event shoot this past weekend.
Camera 1 was a Sony A1 with a sigma 24-70 DG DN f2.8 lens and camera 2 was a Sony A9II with a Sony 70-200GM version 1.

Both cameras had set a manual white balance off a white card (not ideal, but the same card in the same location at the same time)…

Bring the images into DXO and switch to manual raw white balance to tweak the results and in order to get the same colors between the two cameras I have to add 30 to the tint on one camera…. It’s weird. I assumed it was some weirdness with the bodies but maybe it’s with the lenses…

Unfortunately I didn’t swap lenses for the whole event so I can’t tell if the higher tint follows the camera or the lens. I need to do more testing.

PS I, using Photolab 6, but I think the issue could be the same.

I was waiting to provide an update when I heard back from DXO Support, but It’s been a month since I first contacted them.

Fortunately Lightroom have now added improved noise reduction, so I’ll just roll with that now.