For Eric Chan - Panasonic software optics corrections and "induced banding"

Question

Hi Eric,
may be this was discussed already - not software optics correction of course - but whether that will induce a banding that is almost invisible, unless - and here goes a shot from GH2 and 20/1.7 (which is one of the most heavily corrected lenses)... do you see the pattern there ? was it induced by software correction of optics (distortions)? note that to make it visible sharpening was to extreme in ACR and that some HDR processing was done to make it more clear.

Answer 1

function(){return A.apply(null,[this].concat($A(arguments)))}
MadManChan2000 wrote:
The bands are introduced because uniform image noise becomes non-uniformly distributed upon image resampling (e.g., when spatially-varying optical corrections are applied).
With all due respect, I'm not sure I'm buying this...
If the resampling is done with a high quality algorithm I just don't see how a pattern would emerge from "uniform image noise".  I can imagine that the noise might look more coarse in some parts of the image and finer than others, but that's it.
Resampling only adds patterns due to inaccuracies.
My money is on someone having made a "speed" vs. "quality" decision in the algorithm somewhere.
-Noel