Rationale: modern native raw and DNG files embed optical correction parameters that could enable lens correction, without a lens-specific optics module provided by DxO.
For example, an optics module for a camera body (e.g. Nikon Z7) would encode body sensor- and digital processor-specific information such as noise and color profiles, while the lens-specific parameters would be derived from embedded lens profiles in the image files.
This would avoid long delays by DxO in responding to user needs for lens support. See for example this post:
These optical correction parameters, originating from camera raw files, are embedded as DNG “opcodes” when a DNG file is generated.
These lens correction parameters include, among others, lens distortion and aberrations:
“geometric distortion in rectilinear lenses (using 4th degree polynomial corrections that can correct complex distortions)”
"remapping of essentially spherical projections (fisheye lenses for example) to a rectilinear image. "
The two things that leap out at me as missing from this list are the two things that keep me firmly tied to PhotoLab — sharpness and noise.
What I see in the list is the standard sort of “lens profile” you get in most competing software, which is certainly useful to be able to read those values when a module is not present. But for any lens or camera I would own, I would always want a full module.
I’d consider this very useful for ‘interim’ support for new lenses that have correction information in their RAW metadata. I wouldn’t be in favor of this as a substitute for DxO’s lens corrections, which I believe are more comprehensive than the built-in corrections.
This is more likely to come from the demosaicking algorithm than from the lens correction. Lens correction is predominantly about optical distortion and chromatic aberration. In my opinoin, chromatic aberation is the more important parameter for sharpness perception.