There has been discussion here on diffraction effects and how important they are for sharp shots before but it is difficult to isolate them from flash duration effects.
I did a series of test shots of a feather with a setup that should negate any flash duration effects (holding the camera down on a table with the subject and camera on the same surface). However to make sure what was going on I also did the shots in series of doubling ISOs and halving apertures so the flash durations should have been constant.
As usual there is no correct aperture for any given magnification- it depends on the shooting style, The amount of small detail in the subject and the DOF required. This is shown in the 5:1 shots where there is not much difference between the F8 and F5.6 shots (I would have expected a large difference) but suspect I ran out of fine detail on the feather.
However I think you can see that diffraction does play a significant role in fine detail sharpness and in these shots the changing ISO's didn't seem to affect final image sharpness. The shots are all 1/3rd height crops of the originals and all had exactly the same PP.
This also does not mean that flash duration (ie effective shutter speed) does not play a role in helping to get sharper shots depending on the shooting situation.
You will need to view the full size originals by clicking on the pics set to full screen and then click again to see the effects clearly.
Wow, thanks for posting, Brian! Very interesting. This is with a 40D, right? I suspect you can gain a couple of stops more with a low-pixel density FF body, such as the 5D.
Eyvind Ness wrote:
Wow, thanks for posting, Brian! Very interesting. This is with a 40D, right? I suspect you can gain a couple of stops more with a low-pixel density FF body, such as the 5D.
Yes all taken with a 40D.
I suspect but am not certain that if the shots are taken to give the same print magnification with a FF camera (ie 1.6:1, 4.8:1 and 8:1) the effect might be very similar, but i do get very confused with circles of confusion etc
Just FYI: According to the manual a full flash discharge with the MR lasts 1.4ms or 1/714s. 1/64th of that is 1/46000s, and a typical one should be between these outer limits somewhere, maybe 1/20000s. Motion blur could still be an issue since the magnification is so great. Anyways, I suspect the bigger issue is with getting the focus plane right, since the DOF at these magnifications is ridicilously low, at 3x it is between 0.09mm (f2.8) and 0.50mm (f16), again according to the manual.
I`m completely lost, just wish i knew exactly what your talking about so i can get a greater understand of what goes on when shooting at hi-mags because obviously there's something going on hence the post......
you wrote 'As usual there is no correct aperture for any given magnification- it depends on the shooting style, The amount of small detail in the subject and the DOF required'
see that's were i`m at in my head just wish i understood this diffraction effect thing, its probably quite simple really but so am I ...
Lee Ankers wrote:
I`m completely lost, just wish i knew exactly what your talking about so i can get a greater understand of what goes on when shooting at hi-mags because obviously there's something going on hence the post......
you wrote 'As usual there is no correct aperture for any given magnification- it depends on the shooting style, The amount of small detail in the subject and the DOF required'
see that's were i`m at in my head just wish i understood this diffraction effect thing, its probably quite simple really but so am I ...
Lee- it's not that hard - provided you have the other factors that effect image sharpness under control then there is a physical limit on the image resolution you can achieve caused by diffraction at smallish apertures. So if you really want to be able to see fine detail in a shot you need to shoot at fairly open apertures. However doing this of course reduces the DOF. So you end up trying to balance the DOF against the detail and that depends on the subject and what you are trying to do. This is one of the reasons I use focus stacking a lot- it allows me to break out of this balance and get fine detail and the DOF I want in shots.
Diffraction effects do not just affect macro lenses- most lenses have a sweet spot around F6.3 which is where you will get the highest resolution.
The reason I mentioned shooting style was because in many pictures , fine detail is not an important element of the picture - the composition and colour may be much more important. So for example I normally never shoot macros smaller than about F11 near 1:1 magnification because I'm trying to capture some detail, but I often shoot flowers at F16 because I want more DOF and often flowers do not have fine detail that needs capturing- it's more about light colour and composition (not that I'm any good at flower shots ).
Thanks for posting this Brian, shows the effect well, I've messed up more than once with the MP-E at high mag by accidentally setting too small an aperture and then getting a 'soft' result in the final image.
Amy- just my "Development Scientist" background coming out. You certainly do not need to focus stack though unless you are really interested in fine detail but you can see why I became interested in it as a technique.
Adrian - I first started looking at this when I was disappointed with my first high mag shots with the MPE-65 but I was shooting springtails at the time max length about 3mm.
LordV wrote:
I did a series of test shots of a feather with a setup that should negate any flash duration effects (holding the camera down on a table with the subject and camera on the same surface). However to make sure what was going on I also did the shots in series of doubling ISOs and halving apertures so the flash durations should have been constant.
One of the things that I'm going to do in my field shooting is to see if mirror lockup makes a difference at +3x...
LordV wrote:
As usual there is no correct aperture for any given magnification- it depends on the shooting style, The amount of small detail in the subject and the DOF required. This is shown in the 5:1 shots where there is not much difference between the F8 and F5.6 shots (I would have expected a large difference) but suspect I ran out of fine detail on the feather.
Agreed -some of the < 3mm critters you shoot would show degraded detail at 5x and F8.
LordV wrote:
However I think you can see that diffraction does play a significant role in fine detail sharpness and in these shots the changing ISO's didn't seem to affect final image sharpness. The shots are all 1/3rd height crops of the originals and all had exactly the same PP.
Still curious as to how mirror lockup would effect the outcome...
LordV wrote:
This also does not mean that flash duration (ie effective shutter speed) does not play a role in helping to get sharper shots depending on the shooting situation.
Yet another reason why I try to keep my flash duration as low as possible -I can't always take full control of the movement in the scene.
Well done on taking the time to do the test shots Brian. You know that you and I will always have to agree to disagree on this one
Dalantech wrote:
One of the things that I'm going to do in my field shooting is to see if mirror lockup makes a difference at +3x...
Still curious as to how mirror lockup would effect the outcome...
Well done on taking the time to do the test shots Brian. You know that you and I will always have to agree to disagree on this one
I admit I've never played with MLU mainly because conventional wisdom implies it only affects shots taken with shutter speeds between 3 and 1/30th sec- I'm normally shooting a bit faster than that.
I think the balance between light diffusion and sharpness is more interesting - not because of flash duration but because sharp detail requires contrast changes to be seen. It's the main difference between cloudy day natural light shots and full flash shots. The interesting part is balancing too much contrast from a non diffused flash with how much diffusion do you need to control highlights but give reasonable contrast to show detail.
LordV wrote:
I admit I've never played with MLU mainly because conventional wisdom implies it only affects shots taken with shutter speeds between 3 and 1/30th sec- I'm normally shooting a bit faster than that.
True, but so many of the conventional rules change when shooting macro that I wonder if MLU needs a second look...
LordV wrote:
I think the balance between light diffusion and sharpness is more interesting - not because of flash duration but because sharp detail requires contrast changes to be seen. It's the main difference between cloudy day natural light shots and full flash shots. The interesting part is balancing too much contrast from a non diffused flash with how much diffusion do you need to control highlights but give reasonable contrast to show detail.
Brian V.
I've seen that as well -instances where I'm convinced that a diffuser is giving me more detail due to a contrast boost. I'm also getting specular highlights that are not blown, so even though there is a reflection there is still detail in that area...
In summary: At 5:1 my results are clearly sharpest at f/4 and degrade from there as f number is increased. This is true independent of flash duration even for durations in the 25-150 usec regime with the camera on a tripod and MLU / timed release used. I would hope that motion blur could be ruled out under those circumstances.
In summary: At 5:1 my results are clearly sharpest at f/4 and degrade from there as f number is increased. This is true independent of flash duration even for durations in the 25-150 usec regime with the camera on a tripod and MLU / timed release used. I would hope that motion blur could be ruled out under those circumstances.
First let me say thanks for taking the time to do such an exhaustive test!
I'm a little confused as to why some of the flash durations didn't change between apertures (1:1 ISO 800 @ F8 and F11) and that you seem to be more conservative, by a full stop, than most on the max Fstop at a given magnification. Here's what Juza, Brian, and believe it or not I place the MPE-65mm:
1x: F16 with F11 being noticably sharper.
2x: F11
3x: F8
4x and 5x: F5.6
Not that I pay any attention to that list
Last: If diffraction is the result of light bending as it passes through a small opening then how can you get sharper images at higher effective apertures at higher magnifications -it seems to me that the more you magnify a scene the more you'll magnify the defects in it...
Dalantech wrote:
I'm a little confused as to why some of the flash durations didn't change between apertures (1:1 ISO 800 @ F8 and F11).
John,
The shape of the flash waveform and peak intensity were changing too, so you shouldn't read flash duration as a strict measure of exposure.
Last: If diffraction is the result of light bending as it passes through a small opening then how can you get sharper images at higher effective apertures at higher magnifications -it seems to me that the more you magnify a scene the more you'll magnify the defects in it...
This is true, but magnification includes display magnification as well as capture magnification. If you take an image that is 3000 pixels wide and reduce it in size to display 5 inches wide (~500 pixels) on the computer screen you are compressing 6 pixels of image data into a single pixel on your display. Even fairly severe diffractive blur (say 3 pixels wide) wouldn't show up under those circumstances.
The shape of the flash waveform and peak intensity were changing too, so you shouldn't read flash duration as a strict measure of exposure.
Dalantech wrote:
Last: If diffraction is the result of light bending as it passes through a small opening then how can you get sharper images at higher effective apertures at higher magnifications -it seems to me that the more you magnify a scene the more you'll magnify the defects in it...
vachss wrote:
This is true, but magnification includes display magnification as well as capture magnification. If you take an image that is 3000 pixels wide and reduce it in size to display 5 inches wide (~500 pixels) on the computer screen you are compressing 6 pixels of image data into a single pixel on your display. Even fairly severe diffractive blur (say 3 pixels wide) wouldn't show up under those circumstances.
No, I was making reference to your own "verbal" conclusions in the post and not commenting on the images. I'm fully aware that resizing an image will make it look sharper...
Very interesting. This is with a 40D, right? I suspect you can gain a couple of stops more with a low-pixel density FF body, such as the 5D.
