Diffraction Puzzle

[jwrobbo]

I've been taking pictures for 50 years. How come I never heard of diffraction until I started using digital? Almost every lens test warns you not to go passed about f11.
I don't remember this being any sort of a problem when shooting film.
I happily shot at f16-f22 (using Pentax gear mainly) for years. Is there a difference in diffraction between film and digital, or are we now much more analytical?

jwrobbo

[Black Pearl]

No one ever enlarged a negative to the point where it became an issue - in fact very few people enlarged their negative beyond a few inches so while diffraction was there is wasn't a real world issue.

The fact remains that is still isn't a real world issue but we now have a terrible disease in photography, a one that is on the brink of ruining it as a form of creative expression, ultimately turning it into a purely technical process.

PIXEL PEEPING!

What was the first thing you did when you took your first digital photograph? I'll bet it was hit the zoom button to look at it at 100%....then 200.....then 300.....you get the idea. Yes if you enlarge a file far enough you will find all sorts of faults - 'Oh my God I can see noise in my image' is a cry I hear every day and every day I tell the person beating themselves to death over their choice of camera to just print their damn picture and then look for the noise, or the diffraction, or the chromatic aberrations......the list is almost endless. It won't be a problem.

The internet is mostly to blame for this and unfortunately forums are the biggest pot for the moaners to stir and add their poison to.

Simple answer:

STOP PIXEL PEEPING AND PRINT YOUR PICTURES

Complicated answer:

Well stop pixel peeping and print your pictures. I have dozens of 15x10" photographs on the wall at work. Mostly to help sales by showing customers my staff and I can actually use the gear we are trying to sell them, partly to brighten the place up and to some extent because we all like it when someone say 'Ooo, did you take that, it's lovely' They have been taken over the last few years with about ten different DSLR's ranging from my trusty old Nikon D70 to a D3s. From a couple of foot away - still too close to really appreciate a 15x10" print - there is no noise, no diffraction and if there are distortions or other 'problems' they just don't matter to the final image.

I love photography, I always have, but todays pixel peeping has the potential to ruin it as a means for people to express their creativity.

It has to stop.

[Colin Southern]

I agree with Robin.

Photography is always a compromise - if one shoots at a wider aperture to avoid diffraction, one runs the risks of perhaps insufficient depth of field, or too fast a shutter speed - both of which may well have undesirable consequences.

So the big question becomes "which limitation will degrade my image the most"? - in "real world" photography, diffraction has minimal effect on images. Personally, I use apertures like F22 and F32 without hesitation.

I might add that "high ISO noise" is similarly given waaaaaaaay too much press!

[arith]

f22 is only one stop over the diffraction limit on a SD1. About the same as f8 is on mine

[Dave Humphries]

What was the first thing you did when you took your first digital photograph? I'll bet it was hit the zoom button to look at it at 100%....then 200.....then 300.....you get the idea.

I agree with the sentiment Robin,

I do admit the first thing I do is go to 100% so I can delete the mis-focused, subject movement blurred and camera shaken ones in comparison to the rest, but I do that because I shoot multiple frames of birds (or planes) in flight. For 'one off' shots like the first Spitfire one, there's just no point, it's as good as it'll ever be, so I'd better make the most of it

I don't see any point in going beyond 100% for sharpness - in fact the only time I'll do it is to allow 'sloppy brushwork', i.e. less need to be so pixel accurate with mouse (I must try a tablet one day ) if I'm getting tired.

[Black Pearl]

I do exactly the same when taking multiple shots of a subject, much easier to choose which shots to chimp when you can view them at pixel level. What drives me up the wall is people on forums (not this one to be fair) who don't actually take photographs with their cameras instead they seem to spend inordinate amounts of time trying to find faults with them then scream about it to anyone who will listen. I have studied my shots at 100% to see what the noise characteristics are like as I have spent a hour or so running through all the apertures to see when diffraction starts to show but it was idle curiosity and nothing more. What I do is print a large proportion of my pictures (its handy having a Fuji Frontier at work) and these issues are just not field relevant.

[AlanP]

The reason that you read more about diffraction in digital is that diffraction depends on the physical size of the lens opening and on the size of a sensor. A 35mm will produce measurable but not noticable blurring at f22. A DSL with a somewhat smaller sensor will produce the same effect at about a stop less. A point and shoot with a very small sensor and a physically smaller opening for a given f stop will have noticable blurring at all f stops.

[Colin Southern]

The reason that you read more about diffraction in digital is that diffraction depends on the physical size of the lens opening and on the size of a sensor. A 35mm will produce measurable but not noticable blurring at f22. A DSL with a somewhat smaller sensor will produce the same effect at about a stop less. A point and shoot with a very small sensor and a physically smaller opening for a given f stop will have noticable blurring at all f stops.

Hi Alan,

I think it's important to set terms of reference for "noticeable" though ... in this context its pretty much only noticeable when viewed at 100% - and I think the thrust of Robin's & my (and Dave's) point is that in the real world, people don't look at photos at 100%. A typical image displayed online contains only about 3% of the original information (so the diffraction is essentially sampled out), and with prints - unless the print is going to be exceptionally large - the diffraction is too small for the eye to resolve. And in both cases, correct sharpening goes a LONG way towards compensating for the effect of diffraction (in terms of a pleasing image, if not technically).

So not disagreeing with you (just the opposite in fact) -- I just wanted to expand on what you wrote a little.

[Colin Southern]

What drives me up the wall is people on forums (not this one to be fair) who don't actually take photographs with their cameras instead they seem to spend inordinate amounts of time trying to find faults with them then scream about it to anyone who will listen.

I agree. There seems to be a small hard-core group out there who have forgotten that cameras can be used for shooting something other than test targets under controlled conditions!

[jwrobbo]

Thanks for replies. Exactly as I thought. Too much pixel peeping going on.

[Colin Southern]

Thanks for replies. Exactly as I thought. Too much pixel peeping going on.

Absolutely!

Be sure to spread the word

[jiro]

I've been taking pictures for 50 years. How come I never heard of diffraction until I started using digital? Almost every lens test warns you not to go passed about f11.
I don't remember this being any sort of a problem when shooting film.
I happily shot at f16-f22 (using Pentax gear mainly) for years. Is there a difference in diffraction between film and digital, or are we now much more analytical?

jwrobbo

I think it's because a film camera doesn't have micro lenses and pixel sensors to induce diffraction at pixel levels. All the film plane sees are photons reacting chemically to the base elements of the film media. There are just some people who are more entertained discussing very technical terms with regards to their equipment than others. I think we all realize that it is the photographer creating these wonderful images and not the camera. The photographer sees the image, the camera merely records it. I hope I am not too rude to say this but most of the time I have seen an emerging pattern wherein guys that gets too technical about their stuff loses the fun of simply using their camera and look for interesting subjects to shoot. The shots looks "technically perfect" or should I say "pixel-peeped" but I find no emotion coming out from me whenever I view their images. Just a thought. If found offensive please kindly delete by mods.

[arith]

Is it pixel peeping to know that deep red flower that blooms only a week a year will not turn out as sharp at f22, but will be perfect at f7.1?

Is it pixel peeping to know that that bronze statue is going to have a red and blue border set against the sky at 17mm, but 35mm at twice the distance gives the same dof and no CA?

Cor I don't know Most the time I don't even notice faults until I print, but then I print bigger than A3. In fact I never went smaller than 10 x 8 when I was a kid, except when they was medium format contacts, but I can't remember anything about that now.

You used to have to do everything yourself at school, processing, developing and of course taking the picture. You was shown how once only and you was then on your own and had to do it right; or you would find yourself dropped from the photography club and put in rugby

Digital is not so hard; just different.

[Dave Humphries]

Given what you do with your images Steve, you peeping is justified

But for the vast majority of Photographers, well, that's a different story

[arith]

Thankyou Dave; that is very kind, especially from one whose aircraft pics are thousands of times better than my own. I keep trying, can't do Rugby anyway.

[JK6065]

I'm sorry for replying with a question that's different than the original one. But when speaking of diffraction there's something I can't get my head around.
The phenomenon diffraction makes perfect sense to me. I studied it during physics class in secondary school and what that tells me is that full diffraction occurs when the gap the light has to pas through (aperture for instance) is the same size, or smaller than, the wavelength of light.
For partial diffraction the aperture needs to be somewhere near that wavelength (as explained in the CiC tutorial).

But when calculating it shows that for example red light the size of the gap has to be a minimal of 0.00000075mm (1/1333 of a mm) for full diffraction. Even with partial diffraction the aperture wouldn't be anywhere near for example the thickness a human hear 0.00005 (1/20 of a mm), which is one of the smallest things a human can properly see with the naked eye.

Than how come I can even see the smallest aperture with my naked eye if I take a look at the front of a lens when pressing the DoF preview button? Does the lens front element magnify the sight of the aperture that much or am I missing something?

[arith]

Hi Jeroen

Well I certainly can't remember any school physics since I haven't done any for 40 years, but I think the aperture is a tad bigger than those numbers and the conversions are still too small; for instance 50mm f1 is 50mm, and f8 is 50/8 mm and I'm told that is where luminence diffraction starts on a 50D.

But then I rely on the tutorials, but since I occasionally photograph flowers I have noticed a difference with a wider aperture although there is a whole load of things going on like exposure.

[Nihia]

Hurrah someone is taking issue with pixel peeping. I wonder if this pixel peep reflects the recent popularity of minimalism - in interior decorating etc etc. Though I have to watch out for blurriness all the time in my photos, still ...

Cheers
Nihia

[AlanP]

Diffraction also limits how much you can crop a photo.

[Daos]

Hello people.
My first post.
Sorry if it in wrong thread.

Is it true that superzoom camera with 1/2" sensor and 16Mp is diffraction limited at F:2.8 apperture? (online calculator says YES)
If it is true then output 16 Mp RAW/JPEG files probably packed with 8Mp of real image plus 8Mp of useless info similar to noise that have no positive impact on printed image. Am I wrong?