Nikon D800/E Diffraction Limits

Started Apr 28, 2012 | Discussions thread
SNRatio Regular Member • Posts: 476
Re: Nikon D800/E Diffraction Limits

bobn2 wrote:

And the D3 has less resolution at f/11 than it has at f/5.6. The D3X has more resolution at every f-number. DOF comes at the cost of resolution, that is inevitable. Throwing pixels at an imaging problem improves resolution.

I see, you have an extremely wide definition of "meaningful".

Not at all, I'm just not going along with your meaningless line of reason.

If you can't see the meaning yourself, you might perhaps want to be a little bit more careful in branding others' concerns as "meaningless"? At least if you want to be taken seriously.

A couple of percent resolution improvement at a much used f-stop is, trivially, an improvement. Sure. But it will probably not be considered practically relevant by many.

That is not at all the point. DOF comes at the cost of resolution. Oh I see that I said that before. Sure, if your 'much used f-stops' are f/16 equivalent and so-on, then there really isn't much point using any FX camera, or DX for that matter. You might as well save money and inconvenience and use a compact. These cameras, whatever the pixel count, only come into their own for shallower DOF's. Though, even at deep ones a 36MP camera will be marginally better than, say, a 12MP one.

This also applies to f/11, which may also be a no-go land in your book, the way you argue.. Go communicate your insight to the macro shooters, and some landscapers plus, plus.... You really have a job to educate them. And your "DOF comes at the cost of resolution" mantra easily overlooks one of the real advantages of increasing resolution. For instance, with the D3X, with several lenses, you can stop down one additional stop and have about the same MTF as the D3 - and this may be even much more pronounced with the D800. You are far from the resolution limits of the sensors then, but that may not be the most relevant issue for many applications - like macro and some types of landscape.

The interesting, and practically relevant, phenomenon here is not the trivial truth that all sensors lose resolution to diffraction on stopping down, but the amount of degradation. While the D3 typically loses 20-25% linear resolution upon stepping down from f/5.6 to f/16, the D7000 may lose 40-45%. And the D3200 possibly a good bit more, when an adequate lens is used.

That's a silly 'glass half empty' argument. Who is interested in how much you've lost ? What matters is how much you have .

You are clearly not very interested in why such considerations may be very important to quite a few shooters, so I just urge you to keep your silliness characterizations to yourself. It's best for you, I think.

Also, from f/5.6 -> f/16, linear resolution is reduced by 30% on the D3X, and 20% on the D3. How will these curves look with the D800?

Much the same but with more resolution.

Three likely features (guessed from the D7000+50/1.8G MTFs): First, top won't be that much higher with the 50/1.8G, probably 10% or less resolution increase. Lens limitation, but this will apply to lots of current lenses, even more than for the 50/1.8G. Second, steeper fall - but this depends on lens max resolution. Third, percent improvements stopped far down decreases with increasing pixel density.

First one you will be shown to be wrong, the 50/1.8G is a pretty good lens. Second two, sure, but still an improvement.
What's certainly true is that there is no fixed 'limit'.

Preliminary data don't seem to support your prophethic abilities, however much you trust them yourself. At lensrentals, they tested the 50/1.8G to 1045 centrally, well below the 70-200 II (1080) and the ZF.2 50/1.4 (1137).

And I think I give up on trying to bring up/point to the phenomenon of "asymptotic".

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