Can some one clarify what this paragraph means?

[BehindEnemyLines]

Diffraction

ANALYSIS and RECOMENDATIONS

I see sharpness declining at f/8 or f/11. It looks awful at f/32. I see this on my other lenses, too.

Unless you absolutely need depth of field, avoid apertures smaller than f/8 on modern digital SLRs. Their resolving power is so great that you will soften your images by stopping down unnecessarily. This is why many point-and-shoot cameras don't stop down past f/8.

The confusing part is in bold letter. I understand the inverse relationship between aperture size and f-number, but that sentence seems confusing. Is Ken saying that, in general, the sharp f-number range is f/8...f/5...f/4... or ...f/16...f/11...f/8. Which one?

 

[troytime]

smaller than f/8 means smaller hole, bigger number
the smaller the hole, the bigger the DOF

the sharp aperture range is f8, f5, f3, etc

 

[soydios]

Lenses have a sweet spot. At larger apertures lenses tend to be less sharp and have light falloff in the corners; these detrimental effects are reduced by using a smaller aperture. However, at very small apertures diffraction starts to set in, reducing the sharpness of the lens instead of increasing it as one would otherwise expect.

diffraction @ Wikipedia: http://en.wikipedia.org/wiki/Diffraction (note the sections on "Diffraction by a circular aperture" and "Diffraction-limited imaging")

I tested all my lenses a few months ago, and I put the results of that test in a table: http://img101.imageshack.us/img101/594/tifwd2.png
I discussed it in my post: http://forums.anandtech.com/me...id=65&threadid=2231679

Every lens has a sweet spot between the effects of large-aperture problems and small-aperture diffraction. Generally that is at f/8 to f/11.

 

[Flipped Gazelle]

Originally posted by: soydiosI tested all my lenses a few months ago, and I put the results of that test in a table: http://img101.imageshack.us/img101/594/tifwd2.png
I discussed it in my post: http://forums.anandtech.com/me...id=65&threadid=2231679

Every lens has a sweet spot between the effects of large-aperture problems and small-aperture diffraction. Generally that is at f/8 to f/11.

That's a neat table! Well organized and easy to read. :thumbsup:

Regarding Ken Rockwell, the guy's a wingnut. Sometimes he's right on, other times I strongly disagree with him.

The f/32 example he posted looks really bad. I've shot f/32 using a cheap zoom with a Raynox macro adapter hanging off of it, and the resultant image looks far sharper than his example.

 

[soydios]

Originally posted by: Flipped Gazelle
Regarding Ken Rockwell, the guy's a wingnut. Sometimes he's right on, other times I strongly disagree with him.

My exact sentiments, as well.

 

[fuzzybabybunny]

Originally posted by: Flipped Gazelle

Originally posted by: soydiosI tested all my lenses a few months ago, and I put the results of that test in a table: http://img101.imageshack.us/img101/594/tifwd2.png
I discussed it in my post: http://forums.anandtech.com/me...id=65&threadid=2231679

Every lens has a sweet spot between the effects of large-aperture problems and small-aperture diffraction. Generally that is at f/8 to f/11.

That's a neat table! Well organized and easy to read. :thumbsup:

Regarding Ken Rockwell, the guy's a wingnut. Sometimes he's right on, other times I strongly disagree with him.

The f/32 example he posted looks really bad. I've shot f/32 using a cheap zoom with a Raynox macro adapter hanging off of it, and the resultant image looks far sharper than his example.

It should be noted that the sharpness characteristic of lenses differs from lens to lens. My Sigma 10-20mm turned to crap at small apertures past f/16, and it got noticeably softer from f/8 to f/16. I generally stick to f/8 or f/11 and try not to go past them. Other people seem to pull off f/16 pretty well with other lenses, especially macros.