HC 80mm becomes less sharp with apertures above 8.0... Any ideas why?

[elcamarada]

Hi,

I gave my H1D-22 camera to a fellow photographer yesterday for a studio work. He almost immediately discovered a strange behaviour: when stopping down above f/8.0 (e.g. f/22) and/or making shots with exposure faster than 1/250 (e.g. 1/500, 1/800), the picture becomes visibly less sharp, even in the point of focus. I cannot understand why..

Did anyone face similar behaviour before?

Thanks for your replies,
Roman

[jeremypayne]

Likely you are seeing the effects of diffraction ...

http://www.luminous-landscape.com/tutorials/understanding-series/u-diffraction.shtml

http://en.wikipedia.org/wiki/Diffraction_limit

[elcamarada]

Thanks a lot!

The other thing which I do not understand is that 5D-II produces nice sharp pictures with an old 50 1.4 glass with f/16, while Hass makes blurred images with the same settings.. Is that something expected?

Also, (please forgive me, as I am quite amateur in MF world) how do I achieve both acceptable depth of field and sharpness with my Hass? It makes crisp images with f/2.8-8.0, but I cannot fit the whole head of a model into DOF..

Another thing which bothers me is that images become less sharp at short exposures. Hass boasts that their cameras can sync up to 1/800, but the quality of anything achieved at sync above 1/250 looks.. well.. questionable.

Likely you are seeing the effects of diffraction ...

http://www.luminous-landscape.com/tutorials/understanding-series/u-diffraction.shtml

http://en.wikipedia.org/wiki/Diffraction_limit

[UlfKrentz]

Thanks a lot!

The other thing which I do not understand is that 5D-II produces nice sharp pictures with an old 50 1.4 glass with f/16, while Hass makes blurred images with the same settings.. Is that something expected?

Also, (please forgive me, as I am quite amateur in MF world) how do I achieve both acceptable depth of field and sharpness with my Hass? It makes crisp images with f/2.8-8.0, but I cannot fit the whole head of a model into DOF..

Another thing which bothers me is that images become less sharp at short exposures. Hass boasts that their cameras can sync up to 1/800, but the quality of anything achieved at sync above 1/250 looks.. well.. questionable.

Hi,

you should discover your Canon glass at f8 or wider open as well...
Regarding the fast shutter speeds, the H is equipped with a leaf shutter. This shutter starts opening with a small hole (like a small aperture) and is closing the same way, so it probably is diffraction as well. Using the fast sync speeds will also require a short flash and the use of a fast transmitting system / hardwired sync.

Cheers, Ulf

[elcamarada]

Ulf, thanks for the explanation..

I am getting a feeling that I need to re-study photography, as MF is something which lies in different dimension where my past skills are not applicable..

Does anybody know by chance any good (online?) courses worth attending?

Cheers,
Roman

Hi,

you should discover your Canon glass at f8 or wider open as well...
Regarding the fast shutter speeds, the H is equipped with a leaf shutter. This shutter starts opening with a small hole (like a small aperture) and is closing the same way, so it probably is diffraction as well. Using the fast sync speeds will also require a short flash and the use of a fast transmitting system / hardwired sync.

Cheers, Ulf

[Doug Peterson]

I am getting a feeling that I need to re-study photography, as MF is something which lies in different dimension where my past skills are not applicable..

If you buy your medium format from a good value added dealer (rather than a big box store or used from an individual) they typically provide you all the support you need learning about things like diffraction.

Otherwise Cambridge in Color is the best (single) place I've found:
http://www.cambridgeincolour.com/

Doug Peterson (e-mail Me)
__________________

Head of Technical Services, Capture Integration
Phase One Partner of the Year
Leaf, Leica, Cambo, Arca Swiss, Canon, Apple, Profoto, Broncolor, Eizo & More

National: 877.217.9870  |  Cell: 740.707.2183
Newsletter | RSS Feed
Buy Capture One 6 at 10% off

[EgillBjarki]

I used to have a H2 body with a phase one back, I had HC35mm an HC50-110mm and had the same problem. I found the images unusable at f/22 and soft at f/16.

I agree with Jeremy:

Likely you are seeing the effects of diffraction ...

http://www.luminous-landscape.com/tutorials/understanding-series/u-diffraction.shtml

http://en.wikipedia.org/wiki/Diffraction_limit

I switched back to 35mm, first to the Nikon D700 (right after it came out) but found my self going back to Canon shortly after.

I do not know other medium format systems than Hasselblad on personal experience, but I hope that the other makers are better at making lenses!

[David Watson]

I wouldn't be so quick to assume that it is diffraction.  MF lenses have a larger diaphragm opening at small apertures than 35mm and can therefore be used at these smaller apertures with less diffraction.  Diffraction itself is not a function of the camera or the lens specifically but merely a function of the aperture size. 

I would suggest that you need to send your camera and lens back to Hasselblad for a service and recalibration.

I have attached an image taken on my Hasselblad H3D with a 28mm lens at f22.

[ErikKaffehr]

Hi,

Sorry, you are wrong on the issue. The reason is that although the aperture is larger the lens is also longer (same angle view presumed), so magnification will be larger. The two effects actually cancel out. So for photography diffraction is only dependent on f-stop and the wavelength of light.

The reason you can stop down more on MF is that magnification is less, but nevertheless, you throw away resolution and edge contrast anyway when stopping down.

These two image were shot on an Hasselblad H1 and a P45 back in the "Great 2006 MFD shoot out, by Michael Reichmann, Bill Atkinson and Charlie Cramer), left image is at f/8 and the right one at f/22. The differences are quite visible in large prints.

Best regards
Erik

I wouldn't be so quick to assume that it is diffraction.  MF lenses have a larger diaphragm opening at small apertures than 35mm and can therefore be used at these smaller apertures with less diffraction.  Diffraction itself is not a function of the camera or the lens specifically but merely a function of the aperture size.  

I would suggest that you need to send your camera and lens back to Hasselblad for a service and recalibration.

I have attached an image taken on my Hasselblad H3D with a 28mm lens at f22.

[ErikKaffehr]

Hi,

Laws of physics apply equally to all citizens. Diffraction for small apertures is the same for all lenses.

Best regards
Erik

I used to have a H2 body with a phase one back, I had HC35mm an HC50-110mm and had the same problem. I found the images unusable at f/22 and soft at f/16.

I agree with Jeremy:
I switched back to 35mm, first to the Nikon D700 (right after it came out) but found my self going back to Canon shortly after.

I do not know other medium format systems than Hasselblad on personal experience, but I hope that the other makers are better at making lenses!

[EgillBjarki]

Hi,

Laws of physics apply equally to all citizens. Diffraction for small apertures is the same for all lenses.

Best regards
Erik

That is true Erik, I am not arguing that, how ever some lenses tend to get softer than others stopped down. The laws of physics apply to all lenses BUT some lens DESIGNS are better than others, or optimized for certain areas, I was not happy with the results of my old setup compared to what I have today.

[ErikKaffehr]

Hi,

Sorry I was just discussing diffraction. Some lenses are certainly better in some respects than others.

Best regards
Erik

That is true Erik, I am not arguing that, how ever some lenses tend to get softer than others stopped down. The laws of physics apply to all lenses BUT some lens DESIGNS are better than others, or optimized for certain areas, I was not happy with the results of my old setup compared to what I have today.

[David Watson]

Erik

No Erik you are wrong.

Read this article please:

http://www.cambridgeincolour.com/tutorials/digital-camera-sensor-size.htm

Using the embedded calculator you can see that the diffraction limited aperture of a 22mp 6x4.5 sensor as in the H3D-22 is 20.4 whilst the comparable aperture on a Canon 5D Mk 2 or 1DsMk3 would be 12. I do not disagree that other factors come into play when comparing standard lenses on each system e.g. 50mm on 35mm versus 80mm on MF, that different lenses exhibit different characteristics, and that in any system f8 is almost certainly going to be better than f22.  I cannot accept that an image captured at f22 on a 22mp MFD back will be poorer than an image capture on a 21mp DSLR, in terms of diffraction, all other things (focal length, lens quality etc) being equal.  

It may be that the original poster had a badly setup lens and the camera needs calibration or perhaps the lens was exhibiting some focus shift.

[Slobodan Blagojevic]

... I am getting a feeling that I need to re-study photography...

Good.

(Like with any problem, the true beginning of solving it is always in admitting you have it in the first place).

So let's start with your education. The first and most important word you need to know is:

Measurbators

To become a card-carrying member of that elite club, all you need to do is pixel-peep your images on screen at 100% and compare and measure the difference between brands, lenses, sensors, apertures, etc. But do not worry, you'll be in a good company: most of us on these forums are measurbators to some degree, some more, some less. If you do not believe me, just wait till this thread turns into an epic battle between measurbators with opposing views, until geeks get involved and completely highjack it.

To continue your education, remember this advice Miles Davis gave to young jazz musicians (paraphrasing): "First, learn everything there is to learn about jazz... then forget it all and play until you are dizzy"

So, go ahead, read all those links above and learn all there is to lear about diffraction, Airy disks and circles of confusion (no, it is not a local book club or rock band). Then, forget it all and take pictures and judge them by the feeling they create in you and your audience. If you need to stop down, do it. Yes, it will be less sharp in pixel-peeping at 100%, and it might even be less sharp when printed large enough (as Erik showed in another thread), but only if there is a comparison print next to it. In absence of that, whatever you final product is, billboard, double-spread, or gallery print, it will be sufficiently sharp. If your prints are still not sharp enough, chances are something is broken, or other elements of your technique are at fault.

For those who need links and a higher authority than me, I offer none less than the owner of this site, Mr. Michael Reichmann:

http://www.luminous-landscape.com/essays/stop-d.shtml

Mr. Reichmann, like Miles Davis, and like all true artists and intellectuals, is perfectly capable of being a measurbator when it matters, i.e., in the first, learning phase (as indicated in his articles others quoted above), and then forgetting it and shooting for the sake of shooting, not comparing.

P.S. The above is written in a satirical tone, but, of course, only half-so

[Rob C]

Or, you could be like me in my digital incarnation: learn what you need in order to make the camera and printer (if you want one) work and then stop right there, and go make pictures when something strikes you as worth shooting.

Or not. It may be more fun - I base this idea on frequent reading - to make absolutely no images whatsoever, but to agonise over how sharp they might be if you ever get around to making them. Sounds cool... come to think of it, you could even dream up ways to do that without having to beg, borrow, rent, steal or buy a camera at all!

Perhaps I should have had this thought some time ago.

;-(

Rob C

[ErikKaffehr]

Hi,

I missed on the fact that we were discussing a 22 MP sensor. On the other hand, is your 22 MP sensor really full 645 size?! Most sensors around 22 MP are cropped. (Update: it's 36.7 x 49 mm, so 9 micron pitch)

The samples I have published on the forum clearly show that stopping down to f/22 is an issue on P45,  a 39 MP back. I made large prints from these. I unfortunately don't recall the print size but I guess it was around 100x70 cm. I did not make full size prints just A4 crops. The f/8 and f/22 images are easily told apart at 1.5 m viewing distance and the f/22 image is about on par with a 24.5 MP DSLR image shot at f/8. Back in 2006 I made a comparison between my Konica Minolta Dynax 7D and my Sony Alpha 100, the 7d has a pixel pitch of 7.8 microns while the Sony had 6.08 microns. Resolution on both peaked at f/8 and both lost significantly at f/22.Resolution was measured with imatest, but the results matched visual observations.


There is a discussion on how diffraction affects pixels, as four pixels are needed to fully resolve color information, but all pixels are actually used for luminance information. My experiments and observations really indicate that the "Cambridgecolour site is a bit on the optimistic side.

The original poster have seen loss of sharpness at f/22, in my view diffraction is the probable cause of that.

As it seems that you have access to a 22 MP MFDB, why don't you just post one image shot at f/8 and one at f/22, using same sharpening and optimal focus?

Best regards
Erik

Erik

No Erik you are wrong.

Read this article please:

http://www.cambridgeincolour.com/tutorials/digital-camera-sensor-size.htm

Using the embedded calculator you can see that the diffraction limited aperture of a 22mp 6x4.5 sensor as in the H3D-22 is 20.4 whilst the comparable aperture on a Canon 5D Mk 2 or 1DsMk3 would be 12. I do not disagree that other factors come into play when comparing standard lenses on each system e.g. 50mm on 35mm versus 80mm on MF, that different lenses exhibit different characteristics, and that in any system f8 is almost certainly going to be better than f22.  I cannot accept that an image captured at f22 on a 22mp MFD back will be poorer than an image capture on a 21mp DSLR, in terms of diffraction, all other things (focal length, lens quality etc) being equal.  

It may be that the original poster had a badly setup lens and the camera needs calibration or perhaps the lens was exhibiting some focus shift.

[ErikKaffehr]

Hi Slobodan.

I appreciate your satirical tone, no issue with that. I just want to point out that I can see these differences in print, even without pixel peeping

I did actually compare P45 images at f/22 with Sony Alpha images at f/8 and they were on par, actually possibly with a small advantage for the Alpha 900. So if you invest 30 kUSD in  a P45 and shoot at f/22 you give away a lot of investment. In my view it's crucial to know the equipment one has, in order to make best use of it.

Best regards
Erik

Good.

(Like with any problem, the true beginning of solving it is always in admitting you have it in the first place).

So let's start with your education. The first and most important word you need to know is:

Measurbators

To become a card-carrying member of that elite club, all you need to do is pixel-peep your images on screen at 100% and compare and measure the difference between brands, lenses, sensors, apertures, etc. But do not worry, you'll be in a good company: most of us on these forums are measurbators to some degree, some more, some less. If you do not believe me, just wait till this thread turns into an epic battle between measurbators with opposing views, until geeks get involved and completely highjack it.

To continue your education, remember this advice Miles Davis gave to young jazz musicians (paraphrasing): "First, learn everything there is to learn about jazz... then forget it all and play until you are dizzy"

So, go ahead, read all those links above and learn all there is to lear about diffraction, Airy disks and circles of confusion (no, it is not a local book club or rock band). Then, forget it all and take pictures and judge them by the feeling they create in you and your audience. If you need to stop down, do it. Yes, it will be less sharp in pixel-peeping at 100%, and it might even be less sharp when printed large enough (as Erik showed in another thread), but only if there is a comparison print next to it. In absence of that, whatever you final product is, billboard, double-spread, or gallery print, it will be sufficiently sharp. If your prints are still not sharp enough, chances are something is broken, or other elements of your technique are at fault.

For those who need links and a higher authority than me, I offer none less than the owner of this site, Mr. Michael Reichmann:

http://www.luminous-landscape.com/essays/stop-d.shtml

Mr. Reichmann, like Miles Davis, and like all true artists and intellectuals, is perfectly capable of being a measurbator when it matters, i.e., in the first, learning phase (as indicated in his articles others quoted above), and then forgetting it and shooting for the sake of shooting, not comparing.

P.S. The above is written in a satirical tone, but, of course, only half-so

[Slobodan Blagojevic]

... I can see these differences in print, even without pixel peeping...

And I acknowledged that, by saying:

...and it might even be less sharp when printed large enough (as Erik showed in another thread)...

However, I stand by my (and Michael Reichmann) position that it might be true, but irrelevant. In absence of a comparing print, the result, even at f/22, will be sharp enough. I also assume there is a sound reason to shoot at f/22 in the first place. No one shoots at f/22 out of total ignorance (or so I hope), just like no one shoots at ISO 25,600 regularly, without absolute necessity.

When I was shooting Hasselblad (film), I mostly ended up using f/16 and f/22 as I needed DOF. The resulting images were still tack-sharp. Would particular areas of the image, properly focused and exposed at f/8, be sharper? Absolutely. But then, in absence of focus stacking, I would lose DOF.

Also, the whole debate hinges on an obsession with absolute, maximum sharpness. If that is the reason behind investing so much money, than it is a foolish investment, as we all know by now how difficult it is to achieve that maximum sharpness with medium and large formats.

[ErikKaffehr]

Hi,

The tone in the original posting indicated pretty much two things:

1) The original poster frequently uses f/22
2) He was complaining that the pictures were much less sharp at f/22 than at f/8

I missed out on the fact that he was using a relatively low resolution MFDB. Obviously, sometimes you need to stop down. How much is lost depends of course on pixel size, with a higher resolution sensor you would loose more.

What I could see on my prints was that P45 at f/45 performed as a 24 MP DSLR at f/8. Now, the P45 is old stuff, it was the best one could buy in 2006. Today the best is the IQ180, but even that camera will perform as a 24 MP DSLR when stopped down to f/22.

My view is that it's OK to stop down to what is needed. Diffraction is also a bit benign to sharpening. So I stop down to f/16 when needed on my DSLRs, past that, rather not.  It's quite true that the effect will be much less visible with large pixels, because there is less to loose.

The article you refer to was in the time we shot film. I'm pretty sure that Michael has reevaluated his opinion. One of the reasons is that we print larger. My standard print size i A2 and Michael prints larger than that.

You are also right that a picture shot at f/22 may look sharp until you compare it to one shot at f/8 that is sharper. Sharpening can regain a lot of edge contrast as shown here: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/49-dof-in-digital-pictures?start=2 .


Best regards
Erik

And I acknowledged that, by saying:

However, I stand by my (and Michael Reichmann) position that it might be true, but irrelevant. In absence of a comparing print, the result, even at f/22, will be sharp enough. I also assume there is a sound reason to shoot at f/22 in the first place. No one shoots at f/22 out of total ignorance (or so I hope), just like no one shoots at ISO 25,600 regularly, without absolute necessity.

When I was shooting Hasselblad (film), I mostly ended up using f/16 and f/22 as I needed DOF. The resulting images were still tack-sharp. Would particular areas of the image, properly focused and exposed at f/8, be sharper? Absolutely. But then, in absence of focus stacking, I would lose DOF.

Also, the whole debate hinges on an obsession with absolute, maximum sharpness. If that is the reason behind investing so much money, than it is a foolish investment, as we all know by now how difficult it is to achieve that maximum sharpness with medium and large formats.

[BJL]

Erik

No Erik you are wrong.

Read this article please:

http://www.cambridgeincolour.com/tutorials/digital-camera-sensor-size.htm

Using the embedded calculator you can see that the diffraction limited aperture of a 22mp 6x4.5 sensor as in the H3D-22 is 20.4 whilst the comparable aperture on a Canon 5D Mk 2 or 1DsMk3 would be 12.

That sensor is a bit smaller than 6x4.5, it is 48x36mm, so the diffraction limited f-stop by that calculator would be about f/17 ... putting f/22 somewhat info the diffraction limited regime.

Perhaps you are both right but saying things differently. Here are a couple of different ways that people might want to compare. The key is the distance the light gets smeared in the focal plane is proportional to the aperture ratio (f-stop), and the wavelength of the light, and nothing else except some minor variations in the effect due to lens design. Thus the threshold F-stop for adequately low diffraction effects scales in proportion to the pixel size, or perhaps it is better to say it scales inversely with the lp/mm of resolution that you require in the image formed at the focal plane.

A. Aiming for the same image resolution, like with equal pixel count (or in the old currency, equally many line pairs per pictur height of resolution) which is what that Cambridge in color calculator is set up for:
The "diffraction limited threshold" f-stop varies in direct proportion to linear format  size (because it scales with pixel size, and the larger sensor has proportionately larger pixels in this scenario). By the way, DOF scales the same way if you judge by equal sized prints, so you get the same maximum DOF when aperture is limited to get the same resolution on same sized prints.

B. With equal pixel size and the larger format being used to get a more detailed image, suitable for larger prints intended to be viewed from equally close:
Then getting the full resolution offered by the pixel count requires keeping the diffraction spot below the same size limit, and that sets the same f-stop limit. By the way, this means that the maximum DOF attainable decreases as the "total image detail" desired increases.