Tweet
A few of us in my camera club have been investigating an issue with the way the Capture One 23 Pro raw converter handles DNG files created in CO itself and in applications such as Topaz Photo AI and DXO.
This seems to be most serious with Fuji RAF files (which lose their Curve information, preventing choice of film simulations etc, and some Fuji files have major changes to their exposure: this maybe related to compounding problems of de-mosaicing of the unusual X-Trans files), but Olympus ORFs show alteration in the white balance and tint of the DNGs, which is an issue if one's working for example on a set of images or a panel.
To keep this brief, can I suggest anyone who uses CO produces a 3-shot, 2-stop HDR in the programme (which saves the resulting file as a DNG)?
Assuming all 3 base ORFs have the same WB and tint, you'll see that the blended DNG will be different in both parameters. Pano merging in CO has the same results.
The same effect is seen if an ORF is sent to Topaz Photo AI (or any other external app) and the resulting DNG is taken back into CO - the WB and tint will have changed from the original ORF. Since Topaz gives massively better results (in terms for example of noise reduction and artefacts) when used on raw files than on TIFFs, this is very significant for me.
One of our club members is a recently retired digital software image engineer, and he suspects that this problem with so-called 'linear' DNGs (ie those produced in software) remains after CO solved the severe issue it had some months ago also with the DNG files produced natively in some cameras - eg Leica and Ricoh.
The Fuji issue was reported a month or so ago on the CO Community Forum as a bug, and I have added my Olympus experience and provided example ORFs and DNGs. CO have marked the issue as 'in hand', so let's hope so...
This seems to be most serious with Fuji RAF files (which lose their Curve information, preventing choice of film simulations etc, and some Fuji files have major changes to their exposure: this maybe related to compounding problems of de-mosaicing of the unusual X-Trans files), but Olympus ORFs show alteration in the white balance and tint of the DNGs, which is an issue if one's working for example on a set of images or a panel.
To keep this brief, can I suggest anyone who uses CO produces a 3-shot, 2-stop HDR in the programme (which saves the resulting file as a DNG)?
Assuming all 3 base ORFs have the same WB and tint, you'll see that the blended DNG will be different in both parameters. Pano merging in CO has the same results.
The same effect is seen if an ORF is sent to Topaz Photo AI (or any other external app) and the resulting DNG is taken back into CO - the WB and tint will have changed from the original ORF. Since Topaz gives massively better results (in terms for example of noise reduction and artefacts) when used on raw files than on TIFFs, this is very significant for me.
One of our club members is a recently retired digital software image engineer, and he suspects that this problem with so-called 'linear' DNGs (ie those produced in software) remains after CO solved the severe issue it had some months ago also with the DNG files produced natively in some cameras - eg Leica and Ricoh.
The Fuji issue was reported a month or so ago on the CO Community Forum as a bug, and I have added my Olympus experience and provided example ORFs and DNGs. CO have marked the issue as 'in hand', so let's hope so...
Comment