No more details/focus past f/16?

Killerbob

A friend and I was talking, about the issues I am having with oil spots on my sensor. I only notice them at f/14 or smaller aperture settings, and it's not often I use that range anyway.

He then says there is some proof that any pic does not get more detail, or DOF I guess, past f/16 or so... Is that true? And if so, then why do most lenses go to f/20 or so'ish, and I even have an 80-400mm which goes to f/57?

So for the perhaps basic question, but I am trying to learn...

JJ_SO

Google "diffraction". It starts at f/8 and is reducing sharpness the higher the f/number is.
Try to sqeeze your eye-lids. You will get a higher DOF in your eyes but also reduced details. I know, it's oversimplified but not far away from what happens optically.

As for your f/57 - that is at 400mm and closest distance? Well, the aperture is useless, I'd say. Wait: For longtime exposures in daylight it's great, if you just show waves of water and maybe some rocks with not much detail in.

After f/8 the sharpness decreases. Good and bad thing: at all lenses. Bad thing: Our D800 is showing this diffraction more than other cams. Which only means, over f/16 the advantage of a high resolution sensor is loosing a bit of details.

One way to get super sharp pictures with a huge DOF is focusstacking (if the subject is, well, dead and not moving) or tilting a lens (Google "Scheimpflug").

PB_PM

The effects of diffraction depend on a lenses optical formula, so there is no hard rule that says diffraction kicks in at F8. For most consumer lenses that is true, but if you look at actual lens data you'll see diffraction starts to have an affect at F11 or F16 on higher end lenses.

If diffraction really did kick in at F8 most macro lenses would be a waste of time, everyone would have to focus stack every single macro shot.

MikeGunter

Hi,

Seeking proof of life? ;-)

I think he really meant that the best quality of any lens's sharpness lies in a sweet spot somewhere around f5.6 or f8 and that depends upon the lens. The sharpness does tend to weaken from diffraction from more narrowing of the aperture.

The depth of field will be greater with the smaller aperture, and micro lenses take advantage of that.

There is much more, too, you can look of circle of confusion and bokeh for starters.

In creating an image, you try to balance speed of shutter with aperture, movement of the camera and movement of the subject and any movement of background and/or foreground in concert with the lens focal length and ISO selected.

My best,

Mike

Vipmediastar

Lots of great info. I used to shoot f16-f22 depending on scene for overal sharpness. It wasn't until I learned about diffraction that I started to shoot at f/8 or faster. Silly internet tutorials didn't have this info. Even if it was a basic question it's a goodod thing you asked because somebody will learn something new.

Msmoto

If we are talking DSLR lenses, the above is true. But as we get larger in format, e.g, an f/8, 90mm Super Angulon by Schneider on 4" x 5" format, the diffraction issue may not occur until f/32. I try to shoot between one and four stops from maximum, unless I am after depth of field,in which case with a slower lens, e.g., 24mm f/3.5 PC Nikkor I may be down as small as f/22 or f/32. And as noted, when we focus closer the effective f/stop becomes smaller as a result of the magnification increasing dramatically resulting in less light on the film plane, oops, sensor....per area... hope this makes sense.

Golf007sd

@Killerbob: There are many shots when shooting at very small apertures (f/16-22) come in handy. What will cause you to shoot at such an aperture will depend on your distance to the subject....more specifically the foreground subject in relation to the background. Here is an example. Lets say you are shooting with a wide-angle lens and you are low to the ground shooting an object 1-3' from you and you want to show the amazing background that follows. The best way to get this type of shot to shoot at @ f/22 or above. Lastly, if you ever want to have nice "sun stars" from any light source, shooting at f/20 or higher will result in a very nice star...the small the aperture the longer the trails will be.

Have a look at this write up and see if it helps in better understanding my point.

Understanding diffraction is one thing, but knowing how to go about getting a shot is another and that is what counts in my book. Don't let all the tech talk get in the way of knowing how to use your gear.

More tech stuff should your care to read.

sevencrossing

No more details/focus past f/16?

I think the Group f.64 might disagree , but they did use large format camera
as Ms moto points out, diffraction is dependent on the size of the aperture not the f number
with long focus lenses on large format camera using f 22 or above can be normal

Killerbob

Thanks to all for the comments(info - I knew the answer couldn't be that simple.

I live in Greenland and a lot of my pictures will be landscapes/mountains, involving a lot of distance, so being able to pick up the details near AND far is important to me.

JJ_SO

@Killerbob shooting landscapes, most of the time distant mist, dust or - in deserts - heat flares are a bigger problem than diffraction. Your question "no more details/focus (?) past f/16" can be answered with "after f/8 diffraction steps in and is visible on high-resolved FX sensor picture". The next sentence then should be "but to see the effect, you need to be at 100% enlargement, which means 3 big 27" displays horizontally and vertically. Then you loose the overview for the whole picture, if you're close enough to check the effect"

But as you asked... a simplified answer could be, don't close the aperture more than you absolutely need to get your DOF, the sharpness/resolution is not becoming better after f/8.

And as for f.64 on one end and a point & shoot or phone cam on the other end: It's only a question of enlarging big enough which is for large format films simply not very handy and also difficult to measure. And for the small cams... ever asked yourself why they can't be stopped down more than say f/8? It's not because it's too difficult to manufacture smaller apertures. It's just, the gain of DOF is against the losses of sensor noise and diffraction of the super tiny lenses. Not to forget: The lenses may become smaller, but the tolerances to manufacture them economically stay the same.

MikeGunter

Hi all,

@ JJ_SO Maybe - depends upon which topic you're aiming at.

@ Killerbob & Msmoto

And one more thing, and it's somewhat complicated to explain, for me anyway,

With lenses (wide, medium and telephoto) and formats (compact, DX and FX - even medium film and cut sheet) you'll find that 'sweet spot' is a movable target.

Thinking in terms of a 'fixed number' is not the way to go. View camera lenses's might have a sweet spot of f16 or smaller.

Also important in consideration of composition and sharpness position of the photographer to the subject (point of focus) and the lens chosen for the camera. A 105mm on a DX will behave dramatically different than a FX at larger 'stand off' with a more narrow f-stop to boot.

Ansel and company were experimenting with larger depth of field in Group f/64 wanted the world as it is - which is kind of funny, since Ansel always portrayed it as it was not. :-)

He nearly always used filters for his B&W shooting, altering the look of the print significantly with dodging and burning, and the skies had rich blacks with delineated white clouds, and multi-hued grays that only existed in imagination. The sense of real wasn't.

My best,

Mike

Ironheart

@Killerbob you should google "hyperfocal distance"
This will allow you to achieve the near/far focus effect without having to stop down below f/16. For example DOFmaster tells me that with a 50mm lens, f/stop of f/11, and my focus set at 25ft, everything from 12ft to infinity is in focus. If you go to f/16 and focus at 18ft, you get everything from 9ft to infinity in focus. Pretty cool for landscapes eh?

Ade

I don't recommend using the "hyperfocal distance" for most landscape situations (although it is another internet forum favorite technique).

If you focus at the hyperfocal distance (25ft in the above example), far objects are at the "edge" of the focus. To put it another way, they are barely in focus, and the lens will not capture them at full resolution. In fact, you could be sacrificing more than 50% of your lens's resolution this way.

Since in landscape photography far objects often dominate the scene, the whole image ends up looking degraded (soft) if taken at the hyperfocal distance.

A better way might be focusing on the far objects (to preserve maximum resolution) then stopping down to include the foreground objects. The aforementioned "sweet-spot" aperture of the lens can be used as a general guide of how much you can stop down to maximize DoF (this sweet-spot differs from lens to lens and can be found experimentally or via MTF charts.)

Another way to maximize DoF is by using a tilt/shift lens to change the plane of focus.

spraynpray

Ade said:

"If you focus at the hyperfocal distance (25ft in the above example), far objects are at the "edge" of the focus. To put it another way, they are barely in focus, and the lens will not capture them at full resolution. In fact, you could be sacrificing more than 50% of your lens's resolution this way.

Since in landscape photography far objects often dominate the scene, the whole image ends up looking degraded (soft) if taken at the hyperfocal distance."

My experience agrees with that - it ends up all soft. I thought it was my lenses.

sevencrossing

hyperfocal distance is a distance beyond which all objects can be brought into an "acceptable" focus.

The hyperfocal distance is entirely dependent upon what level of sharpness is considered to be acceptable. (The criterion for the desired acceptable sharpness is specified through the circle of confusion)

if your image is not acceptably sharp, then you have not set the lens to the hyperfocal distance

sevencrossing

I live in Greenland and a lot of my pictures will be landscapes/mountains, involving a lot of distance, so being able to pick up the details near AND far is important to me.

As a landscapes photographer, it is it worth looking at the work of Ansel Adams and Edward Weston. Do try and look at some original prints, as most online images are very low res, AA and Weston ( inventor of the Weston exposure meter ) are no longer with us . Of living photographers you might like to look at the work of Tom Mackie

JJ_SO

Or Bruce Barnbaum, if I might add him, who is technically excellent and artistically great as well. Just reading his book "art of photography", I love it.

Golf007sd

@sevencrossing: great recommendation. +1

Killerbob

I read and studied Ansel Adams back in Photography 101-301 in college (I took it as a lab elective so as not to take biology or geology:)

Having taken a look at Bruce Barnbaum's "Art of Photography" I wonder if there is a similar book, but with a focus on today's DSLR world?

JJ_SO

@Killerbob yes it is, Barnbaum did also write a chapter or two about DSLR. You can download a part of his book in iTunes Store to get a taste of his writing. i like hin since I've got "Visual Symphonies" and most of the pictures in that book he explains really good in the "Art of photography". Barnbaum also explains in a nice way where Ansel was wrong

What I also like because it's a creative way to look at pictures is "The passionate photographer" of Steve Simon

Ade

hyperfocal distance is a distance beyond which all objects can be brought into an "acceptable" focus.

The hyperfocal distance is entirely dependent upon what level of sharpness is considered to be acceptable. (The criterion for the desired acceptable sharpness is specified through the circle of confusion)

if your image is not acceptably sharp, then you have not set the lens to the hyperfocal distance

That's right (in theory.)

In practice, however, the Circle of Confusion is defined not by "acceptable sharpness" but by convention based on the sensor size: 0.029mm for Full Frame and 0.019mm for APS-C.

These are the CoC numbers used by virtually all hyperfocal distance calculators out there. There is another thread on this forum explaining the convention behind how these CoC numbers are derived.

Unfortunately, this convention was established prior to modern optics & advanced sensor technology. So if you set focus to the hyperfocal distance based on convention, chances are the results will be disappointing, especially for landscape images.

You could try to redefine your own CoC (perhaps through experimentation) -- but the easier way is simply to focus on the far objects (towards infinity) then stop down to include near objects.

For a more complete treatment of this subject, see Norman Koren's excellent article on Depth of field and Diffraction.

Unless you love math you may want to jump ahead to the section in that article entitled "The myth of hyperfocal distance".

sevencrossing

Ade ............So if you set focus to the hyperfocal distance based on convention, chances are the results will be disappointing, especially for landscape images............
.
+1

I think, if you virtually anything with a modern DSLR, based on conventions set years ago, you might be disappointed

Msmoto

When I shot 11" x 14" Ektachrome on a set, I do not think the lens even had a shutter. And the focal lengths were from 300mm to about 450mm. A black velvet covered paddle was held by the assistant in front of the lens, the dark slide was removed, and the signal was given to the shutter man" to expose the film, then cover again after the timed exposure was complete. Generally the color balance of the film was corrected for times of 20-40 seconds. We usually purchased a full emulsion run of film so the color could be set up on each emulsion run and remained constant as our lab was well run.

And, the f/stop....usually 45, but 64, or 90 was possible with the lens.

The f/stop for various formats is different.....

spraynpray

@sevencrossing: In the first of your last three postings, you seem to disagree with my statement in which I stated that my experience agreed with Ades statement (that the hyperfocal distance doesn't work with modern cameras), then in the third of your last three postings you do a 180 and start to agree! A little consistency is needed (or at least leave longer between opinion swings so I forget what you said before :P ).

I loved the hyperfocal distance mark on the distance scale back before these damn 'G' lenses!

sevencrossing

spraynpray

to clarify, if you some else's definition of "acceptable" to calculate the hyperfocal distance and their criteria is less than yours, then yes, you are gong to be disappointed

as ade says You could try to redefine your own CoC (perhaps through experimentation)

suggestion, set the hyperfocal distance for f 8

then shoot some typical landscapes and f 8, f11, f 16 and f22 and see which one looks sharpest

spraynpray

I'm with Ren Kockwell when he says Nikon castrated their lenses with the 'G' range.