Small aperture choice

[Pete Gl]

I understand a large aperture give a narrow depth of field, and for landscapes it's usually recommended to go for the "Sweet spot" of a lens, which is around F8, F11 ish, at that aperture normally everything is in focus, so: -

What circumstances would make anybody want to select a very small aperture of , say F29 F32 ish??

Curious

Pete

[sandpiper]

Well, the first thing that came to mind was shooting moving water, waves, waterfalls etc. where you want to smooth it into an ethereal blur. Even on the slowest ISO setting, in order to get an exposure of around a second or more you may well need the smallest aperture available, possibly with ND filters as well.

There are many other situations where long exposures are required, shooting a location with people constantly walking past and not wanting them to show up in the shot for example, again the smallest possible aperture will be required.

I am sure that other people will have other uses for such small apertures.

[Doug Pardee]

The main reasons would be to get more depth of field in a photo (typically landscape), or to allow longer shutter speeds for the (cliché, IMO) shots of blurry flowing water.

However, on a camera with an APS-C sized sensor, those tiny apertures result in significant diffraction softness throughout the entire photo. Most people aren't comfortable going above maybe f/14 on a 1.6x camera due to the softness. Careful sharpening can take a lot of that softness out, and there are a few people who are satisfied with their results at extremely small apertures.

As with anything in photography, it's a tradeoff and a personal choice.

[Dermit]

I often use f/16 - f/22 in shooting landscapes and am quite happy with the results. Sometimes when shooting with a given lens and wanting a given composition that involves shooting something fairly close in the foreground the hyperfocal distance requires that I use a smaller aperture to get it all in focus. Some of the best landscape shots include a foreground in the shot that is in close proximity to the camera.

[DrPablo]

There are only two reasons to stop down past f/11 (or at worst f/16):

1. You need the depth of field
2. You need a longer exposure time (and you don't have an ND filter)

Stopping down limits the effect of lens aberrations. But beyond f/11 (and beyond f/4 on some good lenses) diffraction will limit lens performance more than aberrations will.

[ryant35]

When I want to shoot motor sports at 1/125sec in bright daylight I use upwards of f/14

[Glenn NK]

There are only two reasons to stop down past f/11 (or at worst f/16):

1. You need the depth of field
2. You need a longer exposure time (and you don't have an ND filter)

Stopping down limits the effect of lens aberrations. But beyond f/11 (and beyond f/4 on some good lenses) diffraction will limit lens performance more than aberrations will.

Dr. Pablo:

I have noticed that the tested IQ (as reflected in MTF bar charts) of some otherwise excellent lenses falls off noticeably at f/16 or f/22; an example:

http://www.photozone.de/8Reviews/lenses/canon_100_28/index.htm

It seems to me that the rapid decrease in resolution with smaller apertures is being caused by diffraction, otherwise the resolution should increase by limiting aberrations (at smaller apertures).

If so, since the test of this and all lenses performed by this site are done on an APS-C camera, it would seem to be an excellent indicator of the f/stop range at which one would achieve the sharpest results on this sensor size.

What are your thoughts on this?


Glenn

[DrPablo]

When I want to shoot motor sports at 1/125sec in bright daylight I use upwards of f/14
Is that because 1/125 is better than 1/250 at f/11 for your purposes? Or is it because the DOF is too shallow at f/11?

[sandpiper]

Is that because 1/125 is better than 1/250 at f/11 for your purposes? Or is it because the DOF is too shallow at f/11?

It's presumably to get the necessary degree of motion blur on the background and wheels. Any faster speed is likely to make the shot too 'static'. I use 1/125th for the same reason for this work.

[ryant35]

Is that because 1/125 is better than 1/250 at f/11 for your purposes? Or is it because the DOF is too shallow at f/11?

No for panning I shoot from 1/250 all they way down to 1/40 in Aperture priority. I would shoot at a larger aperature if I could, but the max is over f/10 or 12. I wish my 20D would go down to iso 50. If I could get my aperature at around f/8 that would be perfect, some sensor dust would not be visible. Since I am panning depth of focus beyond the truck isn't important.

I would like a larger aperture since these races are so dusty and I have to change lens so sensor dust by the end of the day becomes a problem. I swab my sensor before the 2nd day.

[ryant35]

Example:
Aperature priority, 1/250 sec, f/10, iso100
http://i88.photobucket.com/albums/k173/ryant35/IMG_0058.jpg

[Pete Gl]

whoooaa Dr Pablo, 'diffraction' 'aberration', Glenn, MTF bar charts ???, bit more explanation please!!.

[Glenn NK]

I will defer to Dr. Pablo on diffraction and aberration, but MTF (in simple terms) is a measure of the "sharpness" of a lens. (That being said, I believe that when one testing a lens, one is also inadvertantly testing the sensor too.)

In any event, the "classical" MTF charts can be hard to decipher, but the "photozone" site plots the results as bar charts for various f/stops. These are easy to comprehend - the "taller" the bar, the better the sharpness or resolving power of the lens.

Another example of one of my lenses (incidentally, I look carefully at these tests before buying a lens, and not to discredit the opinions of others here on this forum, don't ask anyone "which lens is best").

http://www.photozone.de/8Reviews/lenses/canon_1755_28/index.htm

If you scroll down to the heading MTF (Resolution) you'll see bar charts plotted for different focal lengths of this zoom lens; the blue bar indicates how sharp the lens is at the centre of the image, the pink bar the edge sharpness, and the beige bar the sharpness at the corners of the image.

Values between 1,850 and 2,150 are considered EXCELLENT, the next step down is VERY GOOD, etc.

I wouldn't buy a lens that didn't score into the EXCELLENT range (at least in the centre) as this assures me that if something "isn't right" with my images, it's not the fault of the lens, but most likely me (something that I potentially can "fix").

The tests also rate other aspects of the quality of a lens; useful information when considering a purchase.


If anyone really wants to delve into the testing procedure, the following site is interesting (it's referenced at the photozone site). Some of this material may be a bit "heavy".

http://www.imatest.com/docs/sharpness.html

[DrPablo]

Dr. Pablo:

I have noticed that the tested IQ (as reflected in MTF bar charts) of some otherwise excellent lenses falls off noticeably at f/16 or f/22; an example:

http://www.photozone.de/8Reviews/lenses/canon_100_28/index.htm

It seems to me that the rapid decrease in resolution with smaller apertures is being caused by diffraction, otherwise the resolution should increase by limiting aberrations (at smaller apertures).

If so, since the test of this and all lenses performed by this site are done on an APS-C camera, it would seem to be an excellent indicator of the f/stop range at which one would achieve the sharpest results on this sensor size.

What are your thoughts on this?


Glenn
Glenn, you are spot on.

Light passing through a hole always diffracts. The smaller the hole, the more the diffraction. Diffraction will lower the optical resolution of your lenses.

Aberrations are 'errors' in your lenses, of which there are many kinds. Stopping down will limit the effect of aberrations. But when you stop down past a certain point, diffraction becomes more important than aberrations.

As for the APS-C camera, the effect of diffraction is exactly the same on an APS-C camera as on a 20x24" ultralarge format camera at capture size. But the problem is you have to enlarge APS-C images more than 35mm or medium format or large format. So the smaller the sensor, the more an enlargement (of a given output size) will be affected by diffraction.

So my 8x10 camera is probably sharpest at f/8 or f/11 just like any other. But because I don't enlarge it at all (or no more so than 2x), even if I shoot at f/90 or f/128 (which I can) I don't see the loss of resolution.

[freaking102]

No for panning I shoot from 1/250 all they way down to 1/40 in Aperture priority. I would shoot at a larger aperature if I could, but the max is over f/10 or 12. I wish my 20D would go down to iso 50. If I could get my aperature at around f/8 that would be perfect, some sensor dust would not be visible. Since I am panning depth of focus beyond the truck isn't important.

I would like a larger aperture since these races are so dusty and I have to change lens so sensor dust by the end of the day becomes a problem. I swab my sensor before the 2nd day.

Why you wishing for ISO 50? Just slap an ND filter on your lense and you'll be able to use larger aperature. Your panning pictures may look better if you decrease your DOF.

[ryant35]

yeah your right... I'll have to go get one before the races this weekend.

[JeffreyG]

Why you wishing for ISO 50? Just slap an ND filter on your lense and you'll be able to use larger aperature. Your panning pictures may look better if you decrease your DOF

Mainly because ISO50 is free and fits all of my lenses while quality coated ND filters are expensive, I need different ones for each lens and I'm likely to have left them at home on the rare occasion that I need one.

ISO 50 is better than a 1 stop ND filter.

[ryant35]

Mainly because ISO50 is free and fits all of my lenses while quality coated ND filters are expensive, I need different ones for each lens and I'm likely to have left them at home on the rare occasion that I need one.

ISO 50 is better than a 1 stop ND filter.

Or maybe because my 20D doesn't go down to iso 50:rolleyes:

[PhotosGuy]

Here are two reasons:

http://img.photobucket.com/albums/v218/PhotosGuy/Samples%20-%20General/Seagull-blur_01.jpg?t=1193974276



http://img.photobucket.com/albums/v218/PhotosGuy/Samples%20-%20Rides/F1-Blur-06.jpg?t=1193974184

[DrPablo]

Here's a lot of info about diffraction.

http://www.luminous-landscape.com/tutorials/understanding-series/u-diffraction.shtml
http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm
http://en.wikipedia.org/wiki/Diffraction_limit


You can see that on the Phase One 39 megapixel 36x48mm back, sharpness is noticibly lost by stopping down from f/8 to f/11, and slightly lost even from f/5.6 to f/8. Even if this back (or the other 39MP backs) can resolve as much as 4x5" film by some testing methods, this will only remain true at wider apertures. The ~2.6-fold greater enlargement factor from a 36x48mm back creates a 2.6-fold greater lens resolution requirement for equivalent print resolution. So the diffraction limit will be reached much sooner on physically smaller sensors -- which is why they don't even bother to put anything smaller than f/8 on 8x15mm point and shoot cameras.

[Glenn NK]

Here's a lot of info about diffraction.

http://www.luminous-landscape.com/tutorials/understanding-series/u-diffraction.shtml
http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm
http://en.wikipedia.org/wiki/Diffraction_limit


You can see that on the Phase One 39 megapixel 36x48mm back, sharpness is noticibly lost by stopping down from f/8 to f/11, and slightly lost even from f/5.6 to f/8. Even if this back (or the other 39MP backs) can resolve as much as 4x5" film by some testing methods, this will only remain true at wider apertures. The ~2.6-fold greater enlargement factor from a 36x48mm back creates a 2.6-fold greater lens resolution requirement for equivalent print resolution. So the diffraction limit will be reached much sooner on physically smaller sensors -- which is why they don't even bother to put anything smaller than f/8 on 8x15mm point and shoot cameras.

Thanks for the first and third links Paul; I've added them to my "collection". And further to my previous comments and your confirmation, whenever possible, I will be adhering to the results of the bar type MTF charts on photozone for my lenses. Of course there are times when the increased diffraction from stopping down will be at least offset by increased DOF on some critical aspect of the image. This in fact is covered quite well in my oft-cited reference on landscapes (National Audubon Society Guide to Landscape Photography by Tim Fitzharris).

[DrPablo]

Of course there are times when the increased diffraction from stopping down will be at least offset by increased DOF on some critical aspect of the image.
That's exactly right. I'm not a big user of MTF charts, especially for this question -- MTF charts are most useful to me for getting a sense of how lenses perform at the center versus the periphery. Beyond that, there are other variables that don't show up on an MTF test, so in the simplest terms I just use the widest aperture that will give me the exposure and DOF that I need.

[Glenn NK]

. . . . . I just use the widest aperture that will give me the exposure and DOF that I need.

Which as I recall, is what Tim Fitzharris recommends too.:D

But I also try to stay away from the largest aperture on any given lens.

[DrPablo]

Depends on the lens, of course. The best performers are nearly perfect wide open. Of my small format lenses, my 70-200 f/4L and Tokina 12-24 are just as sharp at f/4 as at f/5.6.

[Pete Gl]

Thanks for all the replies folks!!

Pete