Demo of diffraction effects in MFD

[ErikKaffehr]

Hi,

It has been suggested that f/22 is usable on MFD. Obviously, it is necessary to stop down to achieve sufficient DoF, but it doesn't com free in my experience.

Here is a link to three actual pixel crops (with links to raw files) http://echophoto.dnsalias.net/ekr/Articles/MFDiffraction/

The same samples were than sharpened in FocusMagic, essentially using default settings and letting FM choose radius.

Here is the link to those images: http://echophoto.dnsalias.net/ekr/Articles/MFDiffraction/FocusMagic

I also included the images sharpened in FocusMagic as a screen dump. I would say that the loss of image quality is quite obvious at this pixel peeping resolution, but it is quite possible that it wouldn't matter in small size prints. What I finally did was to stack all my five images, three different points of focus at f/11 and the f/22 and the f/32 images.

The Sonnar 180/4 used here is quite sharp, I guess it performs optimally at f/5.6 or f/8.
Equipment used: Hasselblad 555/ELD, Sonnar 180/4, P45+

Best regards
Erik

[Hank Keeton]

Hello Erik,

Yes, we shoot with similar gear (same lens+body....Leaf, for me...), and have similar experiences. That 180mm is one of my favs, but I rarely stop-it-down any further than f16. You have demonstrated what I have found in my more stopped-down images.

Cheers,

Hank

[Paul2660]

Erik

Thanks for the info. I have found with the 40mm rodenstock and 60mm Schneider, that past F11 the effects of diffraction become apparent.  This is on the 60MP dslsa.

However on the 36mp Sony chip and the same lenses, I felt I was able to get to F16 before the diffraction issues were showing. I used the Arca DSLR2 and a Sony A7r.

Paul

[torger]

Ideally I'd like the sensor to outresolve the optical system. The advantage of a soft 100% is that you're then free of aliasing artifacts and no matter how large print you make it will not look "digital" up close. So in a way, diffraction is your friend, it smoothes away aliasing.

However, with "only" 39 megapixels or 33 as I have a soft 100% means that the total resolution is a bit low if striving for "large format" resolution. If I had say 120 megapixels I'd still shoot at f/11 get a soft 100% and be happy with a file free from digital artifacts that resolves the optical system almost in full. Today megapixels set pricing though and there are some technical disadvantages of small pixels, but in the future I hope sensors are all so high resolution that it will just be seen as "dumb film" and what we need to care about is only the optical system.

When I started with MFD landscape photography I shot mostly at f/11 with my 33 megapixel back. But today I shoot more often f/16, just because I've started to dislike aliasing more than lack of resolution. My intention is to upgrade to more megapixels though so I can shoot at f/11 with less aliasing than today. I have 7.2um pixels today, and aim for 6um. Not sure if it will change that much on the aliasing front though, the difference from 9um (22 megapixels) to 7.2um was larger. A comparative test for that would be interesting.

I guess the look of f/11 with 6um pixels should be about the same as f/14 on 7.2um pixels. Better, but a little too much aliasing still for my taste, but from a cost and tech wide compatibility aspect I can't go to smaller pixels now.

[ErikKaffehr]

Hello Hank and Paul,

Thanks for feedback. Paul's observation that he can stop down to f/16 on the Sony vs. f/11 on the 60 MP back is interesting, but I don't figure how to make sense of it. It may be that the smaller pixels on the Sony A7r my respond better to sharpening, 6 micron pitch vs. 4.8 microns.

Best regards
Erik

Hello Erik,

Yes, we shoot with similar gear (same lens+body....Leaf, for me...), and have similar experiences. That 180mm is one of my favs, but I rarely stop-it-down any further than f16. You have demonstrated what I have found in my more stopped-down images.

Cheers,

Hank

[torger]

Hello Hank and Paul,

Thanks for feedback. Paul's observation that he can stop down to f/16 on the Sony vs. f/11 on the 60 MP back is interesting, but I don't figure how to make sense of it. It may be that the smaller pixels on the Sony A7r my respond better to sharpening, 6 micron pitch vs. 4.8 microns.

Differences in microlens efficiency maybe has some effect too? I assume that the Sony sensor is more efficient in capturing light from the whole pixel surface than the Dalsa.

The Kodak 39 megapixel has no microlenses at all, just as the Kodak 50 megapixels I'm looking at upgrading to. Should make aliasing worse, at least in theory. Not sure how much difference there is in practice though, and no idea how it would affect the look of diffraction...

[Bart_van_der_Wolf]

Paul's observation that he can stop down to f/16 on the Sony vs. f/11 on the 60 MP back is interesting, but I don't figure how to make sense of it. It may be that the smaller pixels on the Sony A7r my respond better to sharpening, 6 micron pitch vs. 4.8 microns.

Hi Erik,

Assuming the same focal length was used, it must be that, plus the fact that the denser sampling of the projected image by the smaller sensels extracts a higher resolution, if everything else remains the same. Of course, the cycles per pixel resolution is lower, but the cycles per mm resolution is usually higher.  We we however convert to the same output dimensions, then things may also depend on other factors, like MTF which gets hit by diffraction.

Cheers,
Bart

[ErikKaffehr]

Hi,

It has been also suggested that multishot backs are more tolerant of diffraction. As I don't own a multishot back, I obviously cannot say anything on the issue.

Anyway, I would say that diffraction effects need to be seen in the context of diffraction and sharpening. Just to add, my guess is that low contrast texture is lost first as diffraction is setting in.

Best regards
Erik

Hi Erik,

Assuming the same focal length was used, it must be that, plus the fact that the denser sampling of the projected image by the smaller sensels extracts a higher resolution, if everything else remains the same. Of course, the cycles per pixel resolution is lower, but the cycles per mm resolution is usually higher.  We we however convert to the same output dimensions, then things may also depend on other factors, like MTF which gets hit by diffraction.

Cheers,
Bart

[EricWHiss]

It's absolutely true that you can utilize the multishot backs with smaller apertures.  It's maybe 1 or 2 more stops when using M/S.  Unfortunately this difference in DOF isn't that much. That last stop only gets you just a little more DOF.