Impact of aperture on distortion. I don't get it!

[sanmich]

It is often stated that "lens X shows some distortion at full aperture, but it is gone by f/*"

I don't understand the mechanism here.

let's say I set a camera/lens @ full bore in front of my favorite brick wall, and the wall edge shows barrel distortion.

Now, I stop it down. this cuts the most periferal rays composing a dot on my film/sensor. Understandibely, it will correct some abherations and increase sharpness.

But in order to corect distortion, it should in fact change the actual place where the dot is formed, which seems impossible, given that nothing happens to the central rays (half of the rays at each stop)

So what don't I understand here??😕

OR

is this distortion vs aperture thing only an urban myth?

Thanks!

 

[jonmanjiro]

change in focus point... sure.
change in vignetting... sure.
change in distortion... i don't remember hearing or reading that before.
got a link?

 

[sanmich]

Lot's of places. e.g. from Huff's review of the 50 Nokton:

"The distortion is only noticeable when up close and shooting wide open or close to it. "

 

[Roger Hicks]

Hmmmm... Conceivably the type of subject matter "up close and shooting wide open or close to it" may make the distortion more obvious? Or severe departures from symmetry (effectively changing the optical layout)? Both sound unlikely.

Cheers,

R.

 

[sanmich]

Hmmmm... Conceivably the type of subject matter "up close and shooting wide open or close to it" may make the distortion more obvious? Or severe departures from symmetry (effectively changing the optical layout)? Both sound unlikely.

Cheers,

R.

Only by intuition, I see "up close" affecting distortion perfectly possible.
what I don't understand is the aperture having a say here...

 

[sevo]

is this distortion vs aperture thing only an urban myth?

I suspect the culprits misunderstand spherical aberration as distortion - whose visible effect is aperture dependent.

 

[jonmanjiro]

Ah ok, I don't normally read Huff's reviews, but he does indeed write that. I know that a lens can exhibit differing amounts of distortion at different focus distances (e.g. the Nikkor Ai-S 35/1.4's barrel distortion get progressively worse the closer you focus) and I can see how distortion could look more obvious at wider apertures, but like you I don't see how stopping down would make any difference 😕

 

[sanmich]

...and I can see how distortion could look more obvious at wider apertures, but like you I don't see how stopping down would make any difference 😕

aren't the two things the same?

 

[sanmich]

".I've noticed some distortion at the edge of the field in my copy....seems to go away after stopping down."

from:
http://photo.net/leica-rangefinders-forum/00bjqW

 

[jonmanjiro]

aren't the two things the same?

not really.

the key word there was *look*. let me try to explain what i mean. hopefully Steve doesn't mind me using one of his photos as an example. the post on the left showing the barrel distortion stands out because its surrounded by a blurred background, drawing the eye to it. what i'm suggesting is that the barrel distortion would not *look* so obvious if the background was also sharp (lens stopped down for more DOF and more sharp details for the eye to see). the barrel distortion would be unchanged though.

 

[sanmich]

not really.

the key word there was *look*. let me try to explain what i mean. i hope Steve doesn't mind me using one of his photos as an example. the post on the left showing the barrel distortion stands out because its surrounded by a blurred background, drawing the eye to it. what i'm suggesting is that the barrel distortion would not *look* so obvious if the background was also sharp (lens stopped down for more DOF and more sharp details for the eye to see). the barrel distortion would be unchanged though.

Got it, thanks Jon.

 

[goamules]

Also, some writers use the very different terms, distortion and aberration, interchangeably. And, there are several types of distortion such as barrel/pincushion, astigmatism, etc.

 

[willie_901]

Also, some writers use the very different terms, distortion and aberration, interchangeably. And, there are several types of distortion such as barrel/pincushion, astigmatism, etc.

This is the key.

Make a list of distortions, artifacts and aberrations, and then people can weigh in on those that depend on aperture.

Vignetting is an artifact and it often decreases as aperture decreases.

Longitudinal chromatic aberration is an artifact and it also decreases (or a least is less obvious) as aperture decreases. Several modern Nikon G prime lenses have large and distracting (to my eye) longitudinal CA that becomes less annoying as aperture decreases. Lateral CA also changes with aperture but it can increase or decrease with aperture. Lateral CA could be less obvious with shallow DOFs depending on where the in-focus objects are in the frame.

Optical distortions such as keystoning/converging verticals are independent of aperture unless the offending objects are significantly unfocused. Likewise volume anamorphic distortion does not depend on aperture. Often it is inconvenient or even impossible to eliminate these distortions. Although tilt-shift lenses can often overcome keystoning/converging verticals.

There is no such thing as perspective distortion because perspective (the relative size of two objects at different distances from the camera) depends only on camera location. We all know ill-chosen camera locations can produce aesthetically unpleasant photographs due to perspective. But this is not the lens' fault. It's the photographers fault.

 

[sanmich]

OK then, when a lens distorts, it does independently of the aperture...
Good to know!

 

[sevo]

OK then, when a lens distorts, it does independently of the aperture...
Good to know!

True where spherical lenses are concerned, as regardless of the aperture the angles between glass and light remain the same. With aspheric lenses geometric distortion could at least in theory be aperture dependent - in practice high end lenses seem to avoid these issues, and there is no aperture control (nor much quality) on low end cameras that use aspherics in more vulnerable ways.

 

[tunalegs]

If the aperture had no influence on distortion, then the placement of the aperture would be of no importance in the lens design. Obviously that is not the case.

With a single element lens this is very easily demonstrated. Moving the aperture to the front or rear of the lens will change the distortion from either pincushion to barrel. Early on it was noticed that distortion could be basically cancelled out by putting the aperture between two elements.

Edit: So far as I know - changing the size of the aperture should not change the distortion, but changing its placement certainly has an effect.

 

[sevo]

Sorry, we should have been talking about aperture size - photo terminology is notoriously sloppy there. Placement is a different matter. FWIW the same goes for size vs. placement of elements.

 

[sanmich]

True where spherical lenses are concerned, as regardless of the aperture the angles between glass and light remain the same. With aspheric lenses geometric distortion could at least in theory be aperture dependent - in practice high end lenses seem to avoid these issues, and there is no aperture control (nor much quality) on low end cameras that use aspherics in more vulnerable ways.

Yet, I don't see how removing (half) the light rays forming a spot by stopping down the lens could change the spot location on the film/sensor.

If the aperture had no influence on distortion, then the placement of the aperture would be of no importance in the lens design. Obviously that is not the case.

With a single element lens this is very easily demonstrated. Moving the aperture to the front or rear of the lens will change the distortion from either pincushion to barrel. Early on it was noticed that distortion could be basically cancelled out by putting the aperture between two elements.

I can easily accept that moving the aperture location inside the lens will impact distortion, but once it's placed, changing it's opening shouldn't change that (see my previous comment).

 

[furcafe]

Along similar lines (no pun intended), I would guess changing aperture would affect the extent to which field flatness, or lack thereof, is visible (not sure whether field flatness is technically related to distortion).

. . . I can see how distortion could look more obvious at wider apertures, but like you I don't see how stopping down would make any difference 😕

 

[Photoniac]

Distortion

Distortion

Sanmich:

Change of Aperture has no influence on distortion.

Lenses are usually optimised for infinity - close up, the distortion pattern
can change dramaticly - especially with assymetric lenses.

See attachment.

Regards, Ronald.