There is nothing quite like the ability of a marketing group to come up with pseudo-scientific bullshit to describe a feature. Intelligent sharpening and intelligent digital zoom skirt the border there, but they seem to avoid crossing the line entirely.
After all, so far my tests have shown that intelligent digital zoom does a creditable job, so what’s the problem? Well, it might be useful to examine the difference between a simple crop and upsize versus the intelligent 2x zoom. And where does intelligent sharpness fit in?
This simple test is not quite comprehensive, but I think it gives enough info to decide how to shoot.
The setup is trivial. The HS25 sits on a table and shoots the coin sitting in front of it a dew inches away. Then, I switch on intelligent sharpening and shoot again, Finally, I switch on intelligent 2x digital zoom and shoot one more.
The processing consists only of cropping and upsizing using the new bicubic auto in CS6. Nothing more, No extra sharpening for anyone.
The results are fascinating …
Please click through each one to get the 800px versions. Even at this small web size, the first image clearly has better edge definition on the coin. In other words, intelligent sharpening plus zoom is actually less sharp when downsized.
And the crops prove that intelligent sharpening is less sharp at 100% too.
More importantly, the internal digital zoom is bested by the upsized crop as well. The digital zoom without intelligent sharpening would no doubt have faired better. But the best it can hope for is a tie as far as I am concerned.
So what this boils down to is that intelligent sharpening does not seem to sharpening at 100ISO for a long exposure like this. No idea why.
And crop with upsize seems to have performed very well, besting the in camera zoom, although that is probably down to leaving intelligent sharpening on.
In all conscience, I can’t say that you want to use either feature unless you absolutely must use the digital 2x zoom for framing. Still, I think it works better to crop and upsize anyway.
