Lens Diffraction Test, with Hasselblad HC lenses

[David Grover / Capture One]

David,

In all seriousness, let's say you wanted to go Gursky and shoot a scene in times square where you pulled tight detailed 50mpx focus from front subject to background on a very large print, (think moving people that are sharp.

How would you do this with a 80mm or even a 50mm lens in digital?

B

Without setting up the shot, it would be hard to guess what the limits are on DOF.  

I actually don't think it would be too much of a struggle with the 50mm.

Depending on the shot you could use the HTS with a bit of tilt to improve the DOF, IF you weren't too worried about tops of Time Square sky scrapers being in Focus.

Otherwise, you can't...

D

EDIT...  Just to show it is not all bad, this shot with the 28mm is at f16.  Everything is sharp from front to back.  (Front being the chest of draws in the FG, back being the small Lion statue in the back left corner)

[Dick Roadnight]

After testing, I could make a case to just have TWO fstops on a "serious" lens -- one of the fstops would be wide open, for focusing and composing,

We like to have the option to make our own decisions, and be able to expand DOF to the point where we consider diffraction becomes the limiting factor, and (sometimes) avoid bracketing for DOF merge. (... and sometimes we might not have a view camera, or we are unable to get ideal DOF even with movements.)

I'm also filing suit against my college professor that taught me, "just shoot at f22 if you want to carry focus throughout the entire frame, for maximum depth". After that, I'm heading to Retouching University, to learn how to shoot Foreground Focus Layers, and Background Focus Layers, and then stitch them together.

Not many I mean months ago, when the only way to get a good, A2+ single shot print was to use Large Format, that was good advice.

If you want foreground and background (and main subject) in focus read Merklinger's "Focusing The View Camera" and put a view camera on the front of your Hasselblad DCU... or you might try to make do with one of those HTS T/S adaptors.

Before posting this, I see that David has made comments which, I think, agree with what I have said above, and illustrated a compromise between DOF and diffraction.

[Carsten W]

Depending on the shot you could use the HTS with a bit of tilt to improve the DOF, IF you weren't too worried about tops of Time Square sky scrapers being in Focus.

Otherwise, you can't...

Correct me if I am wrong, but it would seem to me that if you can orient the plane of sharpest focus so that it covers all moving elements, cars, people, etc., then you could do a second shot where you got the tops of the skyscrapers in focus, and use Helicon to combine them. Wouldn't this work? I don't see any particular reason to have to work front-to-back with Helicon. Bottom-to-top should work too, I would hope.

[gwhitf]

I have forwarded your firmware ideas to R&D for comment.

Best,

David

David,

Kudos and Respect to you for responding on this thread.

I have certainly learned some things about the limitations of my HC lenses, yet I continue to love them more every day, as I get to know them. (I have the same relationships like this with women -- at first I love ever trait, then I discover their shortcomings, while I, of course, remain perfect).

No disrespect meant toward Hasselblad whatsoever, but I must admit, my bubble was burst a small bit when I saw the small-aperture results. But mostly I felt like I had egg on my face for not doing this test about twenty-five years ago. I feel like a chump. (But maybe it was not so much an issue with EPR or any other 120 film).

I will zip the files and send them to you. There are five tests.

I've owned a million Hasselblads in my life, both 500 and 200, (and shot the 680 with Fuji lenses for twenty years too), so my respect for Hasselblad and Fuji is VERY high, (and continues to be).

I want to ask one final question here -- What about The F64 Group? Seriously. They prided themselves on tiny apertures, and unlimited depth of field, (pre-Photoshop). Were they the most misguided group ever, or was it truly a non-issue with film?

[Doug Peterson]

I want to ask one final question here -- What about The F64 Group? Seriously. They prided themselves on tiny apertures, and unlimited depth of field, (pre-Photoshop). Were they the most misguided group ever, or was it truly a non-issue with film?

Diffraction limitation:
 - kicks in at higher number f-stops wise with larger formats
 - kicks in at higher number f-stops the higher the capture resolution
 - is noticeable only if the print is large enough to show it
 - is important only if the other elements of quality would have allowed a sharper image*

Examples at the extremes of both ends
 - Large prints from a G10 may show diffraction at extremely low f-stop numbers.
 - A 3x4" reproduction in a text book from a 16x20 Polaroid camera may not show diffraction at ANY f-stop

The idea of judging quality based on a view which shows the maximum theoretical quality (100% pixel-to-pixel view) is very very recent in the life of photography.

I think the f/64 folks would have been the first to tell you that allowing a *slight* amount of sharpness degradation (visible only in large prints) is sometimes the price you pay to have all parts of the image equally sharp (if that is your artistic intention) since I remember more than one image from my studies where members of that group stopped down past the theoretical diffraction limit for their format.

*e.g. if you're hand holding a 1" exposure than diffraction is not going to be a meaningful limiting factor at any f-stop

Doug Peterson (e-mail Me)
__________________
Head of Technical Services, Capture Integration
Phase One, Canon, Apple, Profoto, Eizo & More
National: 877.217.9870  |  Cell: 740.707.2183
Newsletter: Read Latest or Sign Up

[gwhitf]

I think the f/64 folks would have been the first to tell you that allowing a *slight* amount of sharpness degradation (visible only in large prints) is sometimes the price you pay to have all parts of the image equally sharp ...

My tests would indicate that instead of "all parts of the image equally sharp", that the more accurate term would be "all parts of the image would be equally OUT OF FOCUS".

That is the key thing here. You stop down, thinking you're going to carry focus deeper into the frame, when in truth, it's like you're laying a Gaussian Blur layer over the entire photograph. Even the subject area that you wanted tack sharp.

(Edit: Actually, after thinking about this topic, I could almost make the case for shooting a 1ds3 for landscapes, instead of MF, where you wanted to carry focus deeper into the frame, using one frame, and not doing focus brackets. Due to the smaller sensor, and the more inherent depth of focus in the smaller sensor. But that would mean another test, and I don't have it in me right now).

[Jack Flesher]

I want to ask one final question here -- What about The F64 Group? Seriously. They prided themselves on tiny apertures, and unlimited depth of field, (pre-Photoshop). Were they the most misguided group ever, or was it truly a non-issue with film?

They shot with 8x10 view cameras.  And it did not supply "unlimited depth of field" for them, just eliminated many aberrations and provided a "comfortable" DoF.  The lenses of that period were nowhere near as good as what we have today and needed to be stopped down to remove heavy aberrations.  More to the point, f64 on an 8"x10" (200mm x 250mm) piece of film where a 300mm lens is your "normal" is very different than f64 on a 37mm x 49mm sensor where your normal lens is 70mm or so.  8x10 with a 300 lens will show more serious diffraction at around f90 or 128.  

Finally, film has two traits a digital sensor does not: One is it has physical depth to emulsion, so captures an image in those 3 dimensions, albeit a dimension that is very shallow; two is halation, where the silver-halide crystals reflect light into neighboring crystals, slightly "fogging" them, which in turn creates edge effects similar to those from -- drum roll -- diffraction.  Taken a step further, the fogging causes loss of edge contrast and can cover up some of the diffraction effects -- just like using a "poor" lens can -- so you end up getting a better net image using smaller-than-ideal apertures...

[Jack Flesher]

Fortunately corrected on all H cameras.

Nice! Can you post an example to show us how well it works --- with say the 100 focused on something relatively close, like at about 1 to 1-1/2 meters away, wide open and f5.6 or 8?

[tho_mas]

(Edit: Actually, after thinking about this topic, I could almost make the case for shooting a 1ds3 for landscapes, instead of MF, where you wanted to carry focus deeper into the frame, using one frame, and not doing focus brackets. Due to the smaller sensor, and the more inherent depth of focus in the smaller sensor. But that would mean another test, and I don't have it in me right now).

I'd assume the AA filter degrades sharpness in fine details (esp. in low contrast details) more than the diffraction that is visible at f16 compared to f11 on your P45+. But I never tested that (and never will).

[MichaelAlanBielat]

I'd assume the AA filter degrades sharpness in fine details (esp. in low contrast details) more than the diffraction that is visible at f16 compared to f11 on your P45+. But I never tested that (and never will).

It is a known fact that the AA filter does degrade overall image sharpness out of camera. Proper sharpening techniques in post-processing do help correct for that however and should get your image back on track to greatness. The P45+ back does not have an AA filter nor do the Hassy's unlike the Canon's, Nikons and various 35mm DSLRs. Any camera that doesnt have the AA filter over the sensor will show a significant sharpness boost right out of camera. The only con to that is that the AA filter isn't there to handle moire correction. This can be done in post however. The H3D Hassy bodies, their closed system works seamlessly from lens to body to back to post production (via Phocus). They know their hardware and software inside and out and can take into consideration a lot of the little nuances that they couldn't do before with other vendor's backs.
For example, any protective IR filter/glass protecting the H3D backs is taken into account along with APO correction and the like. Numbers are crunched milliseconds before time of capture to account for these nuances to provide you with the best image possible out of camera. In Phocus, lens distortion correction is all done and just apply a sharpening of like 200-300 and then you are pretty much all set with a nice image. Tweak as necessary and proper image capture techniques also help considerably.

Get your hands on the Multi-Shot Hasselblad back and watch that out of camera image detail and sharpness just sing!

[ThierryH]

To answer your question is very easy:

no, the "f64" group was not wrong all their years: the f64 group was mainly shooting outdoor landscapes, as such shooting at infinity.

Diffraction is minimum at the infinity scale: when the reproduction size increases, the effect of diffraction becomes more important (see my other post explaining it in details). In additon, this group was shooting on large format, 8x10" or up if I am not mistaken: the diffraction limit, respectively the critical aperture at 4x5" at infinity is theoretically f45, resp. 2 stops more for 8x10" = f90.
Therefore, their claim that f64 was an optimal aperture for them was absolutely right.

Best regards,
Thierry

I want to ask one final question here -- What about The F64 Group? Seriously. They prided themselves on tiny apertures, and unlimited depth of field, (pre-Photoshop). Were they the most misguided group ever, or was it truly a non-issue with film?

[jing q]

sorry but what's the big deal with all this diffraction? You stop down till necessary for the depth of field that you need for the picture, at a certain print size.
with the resolutions available today chances are you won't even notice these degradation issues till you print huge.

[David Grover / Capture One]

Correct me if I am wrong, but it would seem to me that if you can orient the plane of sharpest focus so that it covers all moving elements, cars, people, etc., then you could do a second shot where you got the tops of the skyscrapers in focus, and use Helicon to combine them. Wouldn't this work? I don't see any particular reason to have to work front-to-back with Helicon. Bottom-to-top should work too, I would hope.

Probably worth a bash!  Have yet to try it myself though.

[David Grover / Capture One]

David,

Kudos and Respect to you for responding on this thread.

I have certainly learned some things about the limitations of my HC lenses, yet I continue to love them more every day, as I get to know them. (I have the same relationships like this with women -- at first I love ever trait, then I discover their shortcomings, while I, of course, remain perfect).

No disrespect meant toward Hasselblad whatsoever, but I must admit, my bubble was burst a small bit when I saw the small-aperture results. But mostly I felt like I had egg on my face for not doing this test about twenty-five years ago. I feel like a chump. (But maybe it was not so much an issue with EPR or any other 120 film).

I will zip the files and send them to you. There are five tests.

I've owned a million Hasselblads in my life, both 500 and 200, (and shot the 680 with Fuji lenses for twenty years too), so my respect for Hasselblad and Fuji is VERY high, (and continues to be).

If you only realised the issue after a 'test' was it really an issue at all?

...and I aim to please.  ;-)

[Nick_T]

I want to ask one final question here -- What about The F64 Group? Seriously. They prided themselves on tiny apertures, and unlimited depth of field, (pre-Photoshop). Were they the most misguided group ever, or was it truly a non-issue with film?

I think this is the crux of the matter. The lenses have actually got a lot better when you compare H system to V system. I shot V system for 15 years and loved it but the fact is I never looked at those (film) files at 100%. We were looking at film with what a 16X loupe? I cannot face the maths but would be willing to bet that looking at a 120 chrome with a 16X Schneider doesn't even come close to looking at a 30 something MP file at 100%. I pulled up a drum scan of a frame of EPP the other day and was shocked by the (poor) quality. I just don't think we ever knew.
Nick-T

[gwhitf]

If you only realised the issue after a 'test' was it really an issue at all?

...and I aim to please.  ;-)

Yes, it was an issue with a 120 Macro HC last month, on a tabletop shot.

If I ever track down your email address, I'll give you the details.

In the end, with enough Smart Sharpen it was fine, but I was nervous there for a while; thought I had a defective lens.

So yes, it's been an issue.

[jmvdigital]

I think this is the crux of the matter. The lenses have actually got a lot better when you compare H system to V system. I shot V system for 15 years and loved it but the fact is I never looked at those (film) files at 100%. We were looking at film with what a 16X loupe? I cannot face the maths but would be willing to bet that looking at a 120 chrome with a 16X Schneider doesn't even come close to looking at a 30 something MP file at 100%. I pulled up a drum scan of a frame of EPP the other day and was shocked by the (poor) quality. I just don't think we ever knew.
Nick-T

True. Expectations were probably just a lot lower and more realistic given the technology at the time. It simply wasn't possible to capture fine detail like we can today, diffraction or not. And like Jack said, things like halation and the quality of lenses at the time added up to a much lower image quality than what we're used to with modern photography, despite the fact that it was phenomenal for the time. They/we just didn't know any better.

Happens all the time. Vinyl vs. CD, 16mm vs. 4k RED ONE, CRT vs. LCD, and on and on. That mighty P65+ won't hold a candle to technology in 30-40 years, despite the fact that as professionals, we spend every waking minute improving our technique and pushing the quality limit.

EDIT: I realize some folks prefer film vs. digital, and vinyl vs. digital. There's always room for nostalgia and that little bit of unexplainable "style."

[Dick Roadnight]

Correct me if I am wrong, but it would seem to me that if you can orient the plane of sharpest focus so that it covers all moving elements, cars, people, etc., then you could do a second shot where you got the tops of the skyscrapers in focus, and use Helicon to combine them. Wouldn't this work? I don't see any particular reason to have to work front-to-back with Helicon. Bottom-to-top should work too, I would hope.

You can position a plane through any three points - but the limitations of the movements of any specific camera will limit how much you can move the Plane of sharpest focus - most of the modern compact view cameras limit you to 5 degrees tilt, and you need about seven to get the ground you are standing on (and the horizon) in focus with a 150 mm lens.

Where you need a short exposure to freeze waves in a seascape, (or cars, pedestrians etc.) and another shot for the cliffs beyond, (or buildings) yes, I am planing to combine shots vertically, or by differential focus using tilt rather than extension.

As you get a wedge of acceptable focus, you can often get tall, distant objects in focus, but, as I believe you are saying, you can use movements to allow you to use shorter exposures for moving objects near a plane.

This I took with the Hasselblad 50-110 zoom @ 65 mm, 1/20th f11... if the surf had been breaking, or I had wanted to freeze the people on the beech, or the vehicles, a view camera would have been useful.
[attachment=14403:Cornwall...Object_1.jpg]

[stewarthemley]

Dick, Port Izzac (sp?) looks as interesting as ever. Taken from the Doc's house?

[capital]

I want to ask one final question here -- What about The F64 Group? Seriously. They prided themselves on tiny apertures, and unlimited depth of field, (pre-Photoshop). Were they the most misguided group ever, or was it truly a non-issue with film?

Hi gwhitf, please see my response on post #37 of your thread. I show just what you ask about film.