Anyone want to explain this graph?

[sothoth]

Anyone want to explain this one? I scanned it from Digital Photographer magazine (issue #50), a good UK photo mag. It's part of a review on a sigma lens, the 70-200, and they did a similar review on the Canon version (also 70-200). It's funny though, they have a different graph on the Canon review, showing chromatographic aberration (in pixels) for the three RGB colors. I get that graph but not this one. I also don't get why they wouldn't test the same things for each lens for a real comparison?

Maybe another reader can enlighten me on what this graph means.

http://good-times.webshots.com/photo/2349418340100238911MSQOcj

 

[Iron Flatline]

Good question... what is a "Cycle per Pixel"?

 

[markc]

I found this: http://www.sigmauser.eu/content/view/81/9/

 

[sothoth]

I found this: http://www.sigmauser.eu/content/view/81/9/

Thanks for this info... that explains it.

What this doesn't do is explain why they'd use this data/graph in the lens review. The article you linked says that this tells you where the sweet spot is for the lens, it says little of the performance, esp since the absolute number of cycles per pixel have to do with the sensor as much as anything. They say the lens dictates how much difference you get between the three lines (the closer together the better) but they don't have any standard to compare to, so it means nothing to me whatsoever.

Maybe I'm just missing the point but to me this seems like a silly waste of page space in a 1 page review.

 

[ksmattfish]

It's all hoo-haa designed to stimulate the gear masturbation gland. It has more to do with buying and selling lenses, than photography.

KSmattfish's sure fire lens test (no graphs, no charts, no laboratories needed): Take 2 lenses, and take a variety of real world (meaning things you normally photograph) test shots with each. Compare 12"x18" prints with your own eyes.

It's simple, and since you're doing your own testing, you don't have to worry if the reviewer is getting paid off.

 

[fmw]

There are only three optical perfromance parameters that matter in camera lenses - resolution, contrast and distortion. Everything else is scientific mumbo jumbo. All the the other things they talk about are things that affect these three things. The only lens tests that I've seen on the net that I like are those done by DPreview. They photograph things and then compare them to those things photographed by something else.

Don't judge a lens by a graph of scientific mumbo jumbo. It will confuse you and won't help in any way. The magazine writer is just trying to impress you for some reason.

 

[sothoth]

It's all hoo-haa designed to stimulate the gear masturbation gland.

It seems the gland you speak of is very active in that review...

My favorite camera of all time is the Pentax K1000. Cost me $100 in 1986 (used) and I bought two used lenses ($50 each) for it and did most of my best photography with it. All for less money than I just paid for a single wide-angle DSLR lens last week.