encouraging. Just curious as to your opinion of the new deconvolution capture sharpening as compared to 3rd party tools such as focus magic.
Hi Wayne,
I need to do some more testing, but it will probably be possible (with the right tools) to squeeze a bit more detail out of an unsharpened image in post-processing. But for a one-stop processing with only Capture One, it is possible to create very nice output quality, and the Proofing option for the process recipes is well implemented.
The lens correction option supposedly uses deconvolution sharpening to reverse Diffraction effects, which is a good choice. What I'm not yet sure about is how they do that (based on EXIF if available, or detail analysis like the CA correction), and if Diffraction is all that's taken into account (which looks like it is) or that also sensel aperture blur is corrected (which it doesn't look like), and the effect of an Optical Low-Pass Filter (OLPF or AA-filter). The Point Spread Function of the diffraction pattern per pixel also varies with wavelength and focus distance (magnification factor), so it's not simple if to be done correctly.
Each sensel has a finite physical area over which it integrates the signals which also blurs fine detail (and reduces aliasing), as opposed to a point sampling device. At some (wider) apertures, there will also be residual lens aberrations that will cause softening. I would have liked it if there was some level of user control instead of only an on/off toggle, but who knows if that's something for the future ...
But as it is, it's a good step forward, even if the effect at some apertures may be subtle. It matters if we look at details in the plane of best focus, or elsewhere in the DOF zone. The Diffraction correction will have the most noticeable effect in the plane of best focus (which is only a shallow zone).
I would also have liked it better if the normal sharpening methods had been expanded with an alternative method, instead of only a means to mitigate the halos inherent in the USM like approaches. I prefer prevention of halos over reducing them after they were created. But the current implementation is certainly an improvement, especially because it can also be applied locally in adjustment layers (multiple layers with different settings are possible).
Cheers,
Bart
