Diffraction f/Entekaphobia

Started 10 months ago | Discussions thread
_sem_
Veteran MemberPosts: 4,313
Like?
Re: Roger is almost correct
In reply to falconeyes, 10 months ago

falconeyes wrote:

The problem is people are confused by what everybody writes and it is hard to really understand.

Even Roger Cicala, as much as I respect his work, suffers from this sometimes. He way too much uses MTF50 as his only tool.

As a matter of fact (and as you will learn when really understanding MTF) information (detail) isn't fully lost until you reach about the Rayleigh criterion (about f/13 for the D800E pixel pitch). It then depends on other factors (such as noise) how much detail you really captured. Certainly much more than suggested by MTF50. The detail is recovered via sharpening or smart sharpening.

In the good old days, it was good practice to compare MTF5 or MTF10 when speaking about resolution (the Rayleigh criterion corresponds to 9% contrast or MTF9).

I don't say Roger is wrong. But unfortunately, this particular article doesn't mean much as he missed to measure and compare MTF10.

I think Roger's article is fairly OK. Not something groundbreaking, as this has been demonstrated before. Pure diffraction blur is known to respond quite well to convolution-based sharpening. The higher the sensor resolution the better (so that the rasterisation interferes less with restoring the unblurred image). Works better on flat targets than in real-life 3D scenery (3D deconvolution would be required for something better...).

One classical issue Roger has forgotten is to measure the actual apertures. The declared numbers are often off. So much that it is still an open question whether some lenses are better than others at very small apertures.

Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow