[ElJamoquio]

Dr. Camera: Have the recent improvements in sensors changed the optimum sensor size?

Even with most expensive lenses and no anti-aliasing filter, the best practical APS-C system has 12% worse resolution than full-frame with a $120 lens.

Never trust anybody on the internet, but...

[JinDesu]

Why not use the Zeiss Otus on the APSC camera?

[Na Horuk]

No. This graph is all wrong. It makes the mistake of setting "equivalent aperture" on the abscissa (constant) axis, as if this makes any kind of real world sense.
And then there is "normalized resolution" which is suspicious at best, but I won't even touch that one

I respect that FF sensors might have an advantage in some situations (like the D4, with 16MP FF, which guarantees great noise performance), but this kind of "analysis" is skewed.

[elliott]

Must be true, he has a PhD.

[panoguy]

Must be true, he has a PhD.

In music!

[ElJamoquio]

Why not use the Zeiss Otus on the APSC camera?

Because it's 50mm. Send him a note and ask him if he has the data to put a Zeiss 35mm on the APS-C graph.

[carrrlangas]

Why not use the Zeiss Otus on the APSC camera?

If he was using a 75mm lens on the 135 format camera, then it would make sense to use 50mm lens on the APS-C camera but he stated he was comparing normal field of view.

It makes the mistake of setting "equivalent aperture" on the abscissa (constant) axis, as if this makes any kind of real world sense.

It makes a lot of sense for testing output resolution: Depth of field depends on the format (as in, sensor size) and defining the circle of confusion sets reference for resolution measure. So, you need the same depth of field in both systems to get a commesurable number... What looks weird is it seems he shoot both systems wide open.. That would give an advantage to APS-C..

And then there is "normalized resolution" which also suspicious at best, but I won't even touch that one

Normalizing just means he is using a standard scale..

Disclaimer: I don´t car for any format. Very happy with my current system. Would like to try Medium format

[ElJamoquio]

Must be true, he has a PhD.

The most pretentious title with the least meaning...

[clackers]

Must be true, he has a PhD.

[ElJamoquio]

No. This graph is all wrong. It makes the mistake of setting "equivalent aperture" on the abscissa (constant) axis, as if this makes any kind of real world sense.
And then there is "normalized resolution" which is suspicious at best, but I won't even touch that one

I respect that FF sensors might have an advantage in some situations (like the D4, with 16MP FF, which guarantees great noise performance), but this kind of "analysis" is skewed.

"X"-axes do not need to be a "constant" axis. In this chart you clearly want to be at the upper left. This is similar to a lot of cost/benefit charts.

The 'normalized resolution' is indeed eyebrow-raising, though. But the data is similar to other data on the web, even just comparing center resolution in lp/ph from lens/camera to lens/camera.

[Kayaker-J]

That new Sigma 50mm/1.4 Art is not going to be any $800 lens ($1700 is a more believable fresh projection online). If I were selling homeopathic cures, "male enhancement" pills; or perhaps, "The Ex-Playmate of the Month's Guide to Anti-Vaccination Science", I'd be sending my special offer solicitations off to Dr. Camera pronto!

And which Nikon FF are we comparing to APS-C? A D800, surely (?), given the posted result for the Zeiss Otus. Maybe "green diamond" is something else, however. I'll also note that the "$120 lens" is an almost flat field 50, about as sharp on rez as any nifty 50 offered by Nikon until very recently; and we all know the 50's are sharp as a class.

So how exactly is noAA, modern 24mp APS-C [K-3] losing out in a meaningful way to FF here, as a class, on rez alone? I'm not going to quibble about within Dr.PhD's "12%" range. See digitalcamerareview.com (D600 v. K-3, FWIW) and cameralabs.com (D600/610 v. D7100, FWIW) for two shootout comparisons, and decide what you're seeing for yourself. I'd think the latter is the more credible test; you can safely extrapolate results to the K-3, in my estimation.

D600 v. K-3? Personally, I bought a K-3, and I have Nikon lenses. For more feel-good stuff, with caveats, check out the K-3 v. brand spankin' new 24mp Sony A6000 on the "comparator" at imagingresource.com.

[cali92rs]

Doesn't matter...pentax doesn't sell a FF camera, so people will dismiss the results as being invalid...until of course pentax does release a FF camera then those same people will be first in line to buy one.

[waterfall]

Dr. Camera: Have the recent improvements in sensors changed the optimum sensor size?



Never trust anybody on the internet, but...

Dr. camera,

Please define your terms and tell me How You Know, I.e. data is useful. Has your analysis been Peer reviewed? If so, who are the Peers? This entire article appears to be junk science intended as a marketing tool.

Dr. Camera probably has a mail order PhD from a Southeast Asian corporation.

Again: what do you mean? How do you know? The colorful bar graph is as useful as CNN's analysis of Malaysia 370.

IMO this is a piece of garbage, with no actual scientific data presented or cited.

[cali92rs]

That new Sigma 50mm/1.4 Art is not going to be any $800 lens ($1700 is a more believable fresh projection online).

The sigma 35mm f1.4 art sells for $899.
The canon 50mm f1.2 L sells for $1700.

I guarantee it won't sell for $1700. It will be much closer to the Sigma 35mm in price.

[Kayaker-J]

The sigma 35mm f1.4 art sells for $899.
The canon 50mm f1.2 L sells for $1700.

I guarantee it won't sell for $1700. It will be much closer to the Sigma 35mm in price.

Well, Yamaki-san & Sigma made no bones that they were shooting for Zeiss Otus territory with that lens. So, Santa comes early this year. I guess we'll all see pretty soon. Gathering together the 900 clams for yours now? I sure hope Zeiss executives won't be jumping from the ledges.

[Addendum: The latest rumor I've seen -- 4/9/14, ultimate source uncited -- suggests that the retail price will be "one quarter" that of the Zeiss Otus. This game is almost as much fun as following the indulgences and perils of certain notable train wreck celebs. So, I wonder what might "give" on Sigma's side of the balance sheet as the inevitable comparisons are played out? Well, one might just speculatively infer from recent history: properly working auto-focus?]

[Addendum: petapixel.com is now -- 4/12/14 -- reporting that Sigma has pegged the price of the new 50mm Art at $950. Whew, really? The article, citing "shocked" observers, speculates about "sweating... Zeiss executives." Ha, ha... I can't really see the rational for Sigma here, if it lives up to most of the Otus-y hype; unless the competitive target is the new Zeiss 55mm for Sony FE (FF alpha). Well, if this is really it, you win the guessing game, cali92rs! I guess we need to see the concordance take, if one develops, on production model quality before we can put this in proper context.]