gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
|
Nice work. Your attack on Ross’s helpful post seems to have persuaded him to delete it. Proud of yourself?
- - -
As difficult as it was, I actually waded through your voluminous post with its 14 (!) bullet points and extended riff on MTF. Or something.
What in the world is the point of this?
Ross helpfully provided real-world examples that demonstrate that the effects of diffraction blur in actual photographs are often quite small, an observation that is known to and obvious to virtually every photographer I know. I suspect that he’s had enough of responses like this one, which comes off as “teaching the child a thing or two.”
Ross knows “a thing or two” about making photographs, and his perspective is founded not on reams of abstract theory but on making and analyzing photographs.
All of these points, quite a few of which seem irrelevant or even misleading, don’t change the fact the matter: You can produce very sharp photographs, including very large prints, using apertures considerably smaller than the so-called diffraction-limited aperture.
Yes, diffraction blur increases as you stop down. No, the take-away from that is most certainly NOT that you should always shoot at the diffraction-limited aperture. In fact, for many intentional effects in photographs intelligent and experienced photographers understand that sometimes things like depth of field trump supposed perfect test-bench sharpness. In fact, to a viewer, sometimes a photograph that uses a mathematically less-sharp aperture creates a subjective effect of greater sharpness, for example by increasing the sharpness of more important objects in the frame.
There is no one correct answer to the question: what is the best aperture? In some cases, it might well be the diffraction limited aperture. In some cases it might be. a larger aperture, even with a lens that increasingly vignettes and loses corner sharpness. In other cases it maybe be a smaller aperture that decreases the maximum sharpness at the sharpest point in the frame while increasing the sharpness of other important elements of the photograph.
While you are pondering, ask yourself why Ross may have deleted his useful post after your unfortunate reply? Is it your intent to silence other posters, in particularly talented and experienced photographers like Ross?
Perhaps take a step or two back from the test bench, articles about optics, and MTF charts and consider how this all applies to making photographs…
philip_pj wrote:
A few notes, Ross, and of course thanks for doing this.
. there is no tougher subject matter than busy natural scenes, by their nature books are all about broad edges; it is simply unrepresentative of typical use cases, if only because there is little to no depth. They are easy to find, and lazy reviewers like them. But they tell you little.
. few use a fast 135mm for any form of landscapes, which is the prime intended usage of Sony's R cameras;
. no one shoots regular natural subject matter that matters to an R Sony at this focal distance with a 135mm; many portrait lenses (like the V here) are 'softened' in the design stage at close focus (a skin tone issue); medium distance focusing will elicit higher levels of performance at optimum apertures, widening the gap to diffraction impacted apertures;
. your recovery workflow is time-consuming, fussy and extensive; and very probably image-specific making it an onerous overhead for putting lipstick on a pig, so to speak;
. the left side images are still clearly 'crisper';
, focus stacking is stymied by even small shifts in focus breathing and longitudinal changes in field curvature;
. by impacting (flattening) MTF, diffraction apertures retard depth perception by reducing lens contrast - the differences between tones (both brightness and color) are diminished; the 'sparkle' is gone.
. modern lens design is overwhelmingly dominated by the market need for greater MTF at wider apertures, it can be done better, as the Leica SL APOs show;
. lenses with more gradual focus fade are better open field options (they model human vision more realistically), and they tend to be slower lenses;
. Albert's very stylized work generally features 'near-far' compositions, using a lot of saturation, little gradual focus fade, and plenty of darker tones - all these help mask deficiencies in sharpness, not that his dramatic work depends much on that aspect of imaging.
. we don't lose everything with diffraction and the fear of it is overdone, and one's specific work may not need much final peak MTF, we all shoot differently.
. nevertheless, it's real and it's best to avoid your particular threshold value for image deterioration;
, it varies by many aspects of lens design, most older spherical lenses were wonderful for DOF presentation even near the generally accepted threshold of f8-f11; others - in particular APOs - have so much more to lose, being designed for peak work near wide open.
Not many people have seen a full series of MTF aperture as the makers try to keep it out of sight, as with so much else they do. Here is a link to an all-aperture series given to Lloyd Chambers by Carl Zeiss for the Batis 85/1.8 - an 11/8 modern design with 3 APD elements and FLE. It shows, using a nifty panel below the chart, MTF for all apertures from f1.8 to f16.
We always look first at axial performance, the on-axis MTF for 40lpmm are as follows:
f1.8 - 66; f2 - 69; f2.8 - 76; f4 - 78; f5.6 - 79; f8 - 75; f11 - 70; f16 - 56.
That simple step from f8 to f11 loses the user 7% and f16 loses him 25%. From peak (f5.6), f11 'costs' 11% and f16 costs 29%. We can quite validly extrapolate f13 as ~midway between f11 and f16, and find that it costs us 75-63 = 12/75 or 16% from f8, and 79-63 = 16/79 or 20% from f5.6.
But do visit this link and use the widget to get a solid idea of what is happening, it's one case where a picture tells us plenty:
https://diglloyd.com/blog/2015/20150429_1044-ZeissBatis-85f1_8-MTF.html
Good advice in the digital age is to aperture-bracket f8 and f11, for most lenses for most compositions. If your image foreground is 50m away, f5.6 is indicated for 85mm or so FLs. ...Show more →
Edited on Jul 15, 2026 at 11:16 PM · View previous versions
|
