A very good video on "diffraction" and sharpness

[JR1]

https://www.youtube.com/watch?v=N0FX...-S_d6dS31cFCnQ

[NorthernFocus]

That's one of the better such videos that I've seen. He did a pretty good job of explaining and didn't say anything just flat wrong or out of context. Particularly since that's one of those topics where a lot of people do so. Unfortunately the image comparisons didn't really come through very well on the video.

[Mike Buckley]

The problem I find with that video is the same problem I encounter with every discussion about concern for diffraction: it doesn't attempt to explain when diffraction becomes an issue when viewing a print at a reasonable viewing distance. I suspect that's because diffraction is never a problem when viewing a print at a reasonable viewing distance.

[Shadowman]

That's one of the better such videos that I've seen. He did a pretty good job of explaining and didn't say anything just flat wrong or out of context. Particularly since that's one of those topics where a lot of people do so. Unfortunately the image comparisons didn't really come through very well on the video.

I also felt it was a good tutorial, he really got to the point rather than trying to be entertaining us. I also found myself not seeing the difference or finding one or the other image as being sharper or more detailed. I guess it's best to gather our own examples.

[george013]

I also felt it was a good tutorial, he really got to the point rather than trying to be entertaining us. I also found myself not seeing the difference or finding one or the other image as being sharper or more detailed. I guess it's best to gather are own examples.

Somewhere in the beginning there is a link in the video.
http://www.backcountrygallery.com/ph...n-photography/
You can inspect all the images here.
George

[xpatUSA]

The problem I find with that video is the same problem I encounter with every discussion about concern for diffraction: it doesn't attempt to explain when diffraction becomes an issue when viewing a print at a reasonable viewing distance. I suspect that's because diffraction is never a problem when viewing a print at a reasonable viewing distance.

I like that, Mike! Must rush outside with my Sigma 50mm macro set to f/45 [sic] and take a shot of my trees. Then print it 8x10 and hold it the regulation 250mm from my nose. My gut feeling is that diffraction will have become both an issue and a problem with that shot

I'm fresh from a big bash-out on DPR where few people seem to understand that "diffraction" is something that occurs at the aperture and has units of angle. Whereas many people use the word "diffraction" to mean it's effects at the image plane. Thus confusion rules, as is usual in our chosen hobby. Rant over.

[george013]

The problem I find with that video is the same problem I encounter with every discussion about concern for diffraction: it doesn't attempt to explain when diffraction becomes an issue when viewing a print at a reasonable viewing distance. I suspect that's because diffraction is never a problem when viewing a print at a reasonable viewing distance.

In the diffraction limit calculator in the tuthere,the airy disk is compared to the coc. So I think the same rules as for sharpness can be used.
George

[Shadowman]

Somewhere in the beginning there is a link in the video.
http://www.backcountrygallery.com/ph...n-photography/
You can inspect all the images here.
George

George, thanks for the link.

[NorthernFocus]

Somewhere in the beginning there is a link in the video.
http://www.backcountrygallery.com/ph...n-photography/
You can inspect all the images here.
George

Much more demonstrative. Thanks.

It is true that for the VAST majority of photographers, diffraction is one of the last things they need to worry about. It's good information for anyone with a philosophy of continuous improvement. One thing for sure is that the higher quality the image is, the more options there are for what can be done with it. That's useful/meaningful to some, not to others.

[Manfred M]

I remember looking at a few MFT charts and what struck me is that the loss of IQ at the f-stops where diffraction became a factor had similar image quality to shots taken with the lens wide open or stopped down one stop from wide open.

I have absolutely no issue shooting with a lens wide open to get shallow DoF; after all, that's why I spent the extra money to buy "fast glass". In the same vein, I have absolutely no issue shooting at f/16 - f/32, if I need that level of DoF to make the shot.

Like many other photographic considerations, it is something to be taken into accound when setting up the camera to take the shot. We make all sorts of tradeoff decisions when we select a particular ISO, shutter speed and aperture settings to make a specific shot. What is important is that the photographer needs to understand what this means to his or her intended outcome. I have seen very few (certainly not zero) "real life" examples where diffraction has a significant negative impact on image quality.

The real issue is that there are too many photographers out there that haven't got a clue as to what diffraction is and more importantly, how it impacts an image...

[NorthernFocus]

...I have absolutely no issue shooting with a lens wide open to get shallow DoF; after all, that's why I spent the extra money to buy "fast glass". In the same vein, I have absolutely no issue shooting at f/16 - f/32, if I need that level of DoF to make the shot.

Like many other photographic considerations, it is something to be taken into accound when setting up the camera to take the shot...

The real issue is that there are too many photographers out there that haven't got a clue as to what diffraction is and more importantly, how it impacts an image...

Well stated. Too many talking heads with blog space to fill and needing to publish something that makes them sound more knowledgeable than the next person. Otherwise why would anyone tune in after all. So information gets poorly communicated, even more poorly understood, and all of a sudden it becomes way more important in theory than it is in practice.

[Dave Humphries]

I also found it a very good video - I will now check out some of his others.

Thanks Jeremy.


I was watching at 1080 on a 1080 screen (23" diagonal) and for the comparison shots, if I leaned forward to be 8" (20cm) from the screen I could certainly see all the detail differences.

If viewed on any screen that is not the same resolution as the uploaded video at its maximum resolution (1920 x 1080 in this case), I can well believe, in fact I would expect, that the comparison differences would be far more tricky, if not impossible, to see.

This is caused by re-sampling issues.

[NorthernFocus]

...I was watching at 1080 on a 1080 screen...

Well that made a difference. Thanks, Dave. I'm not much of a YouTuber. Had no idea that some are full HD now. Cool...

[Black Pearl]

Well that made a difference. Thanks, Dave. I'm not much of a YouTuber. Had no idea that some are full HD now. Cool...

Some are even 4k but you need the screen to match and a connection with a decent speed to view them - they are impressive though.

[JR1]

I also found it a very good video - I will now check out some of his others.

Thanks Jeremy.


I was watching at 1080 on a 1080 screen (23" diagonal) and for the comparison shots, if I leaned forward to be 8" (20cm) from the screen I could certainly see all the detail differences.

If viewed on any screen that is not the same resolution as the uploaded video at its maximum resolution (1920 x 1080 in this case), I can well believe, in fact I would expect, that the comparison differences would be far more tricky, if not impossible, to see.

This is caused by re-sampling issues.

I have a 28" curved Samsung monitor, amazing value for money and HD is exceedingly good on it

http://www.overclockers.co.uk/showpr...odid=MO-207-SA

[Joan]

https://www.youtube.com/watch?v=N0FX...-S_d6dS31cFCnQ

Very interesting video, but... shouldn't this kind of comparisons be made by taking pictures of test charts with lines placed at different intervals (eg the ones used for MTF charts) instead of taking pictures of trees ?. IMHO the effect of diffraction would be quite more evident and could be even measured and assigned a numerical value (sorry, I'm engineer...)

[ajohnw]

I haven't read all of the posts but if people watch it again and think contrast when ever he says sharpness they will get the general idea of what is actually going on in practice. The 2 things are only loosely related in practice as it's extremely unlikely that any one will ever buy a diffraction limited camera lens. At some point though diffraction is likely to have some effect..

Maybe one day some one will do it correctly and take a raw image without debayering and just scaled according to the filters over the pixels as the light loss in the 3 colours varies - or I suspect it does as colour channel values are scaled. In terms of quality of shots it would just be an academic exorcise but could show pure diffraction effects if the lens was up to it.

Surely a sensor with pixels of half the size of another of the same format will resolve more. Afraid not but it is likely to produce a subjectively sharper result but not for the reasons he mentions. Maybe the ideal would be a cluster of many pixels within the resolution limit of the lens used as debayering would work more effectively. It would be guessing over shorter distances across the sensor - that is why higher pixel densities can give a subjectively sharper image,

Better than most attempts of the same nature though. The debayering effects are described here - surprise surprise. I doubt if similar info is available any where else,

https://www.cambridgeincolour.com/tu...tography-2.htm

On this basis my E-M1 is way way better than Manfred's D800 but is it really ? As they point out it's a very woolly area even though some of my M 4/3 lenses are very likely to resolve to a significantly higher level than Manfred's Nikon lenses. Then move to larger sensor medium format or even film. Lenses are likely to have lower actual resolution but images are very likely to show resolutions well past what can be achieved on full frame even though the camera has grain or even bigger pixels.

I sometimes think Colin had the best attitude - forget it. It's might have some bearing when very small crops are taken out of shots. On the other hand it's not difficult to take a shot that looses detail unless it's displayed at a size well over A4 even with relatively slow apertures even on an m 4/3 camera.

John
-

[xpatUSA]

Very interesting video, but... shouldn't this kind of comparisons be made by taking pictures of test charts with lines placed at different intervals (eg the ones used for MTF charts) instead of taking pictures of trees ?. IMHO the effect of diffraction would be quite more evident and could be even measured and assigned a numerical value (sorry, I'm engineer...)

+1

[george013]

I haven't read all of the posts but if people watch it again and think contrast when ever he says sharpness they will get the general idea of what is actually going on in practice. The 2 things are only loosely related in practice as it's extremely unlikely that any one will ever buy a diffraction limited camera lens. At some point though diffraction is likely to have some effect..
......
John
-

What do you mean with a difrraction limited camera lens?
George

[chauncey]

Would it be fair to opine that, as they use only the sweet spot of the lens, a crop sensor camera
is less susceptible to diffraction than is a full frame camera.