Since 3 months, I have a strength barrel distortion on my pictures developed by DxO.
The camera jpeg is OK, the draw visualised in the finder or in DxO is OK but the jpeg developed by DxO has a strong barrel distortion.
This appeared recently ; I have shot 50000 pictures with this lens and 10000 with the camera.
I would have noticed…
My questioning :
it could be a problem with the lens ; I encountered one among 10000 pictures with a lens element having its own life for just one picture, but the anomaly appears on RAW ans camera jpeg.
it could be DxO that applies a correction, but why…
I use my own preset and have compared with a standard preset : same result.
What seems to be happening is that none of the optical corrections are being applied to the JPG version and, presumably, this is the image you are exporting to create the _DxO.jpg export.
Have you downloaded the JPEG lens module as well as the RAW module?
Of course ; as I wrote, I have processed tenths of thousand pictures!
Probably but why ?
What settings to verify ?
I have made some tests on older pictures : no anomaly.
I am dry…
have run a test on my picture ? If you have Elite version, you can get the module for lens.body.
I made many more tests.
On my old MBP, with PL4 or 5.1.1, no distortion.
I transferred the dop file on my new MBP, relaunch and reprocess : I have always distortion.
Probably an issue with DxO but what to do ?
I update PL5 to 5.5 version on my old MBP : no distortion.
I verify the version : it is 5.3 => I update to 5.5 and it is now OK.
I verified also on other recent picture where I noticed the issue with geometric lines* : OK.
I am relieved though I lost very much time and I would like to hear DxO about that issue…
I don’t know the lifetime of the 5.3 version ; upgrading to a higher version has probably occurred during my absence ; I feel it is a bug introduced by 5.3…
How many hundredth pictures I will have to reprocess ?
it is difficult to see the distortion on grass but I frequently compare build-in and DxO and I can see the difference.