distortion on 28mm SLR lenses compared to RF wides

[andrew_schank]

I just got a few rolls back taken with a nice 28mm f2.8 AIS Nikkor that I was given as part of a deal to sell a bunch of stuff for a friend. The color and sharpness are top notch, but I noticed an unpleasant edge distortion on many pictures. It's not that straight lines are noticeably bent, but more that the corners seem to stretch things out much more than the RF wides do. Is this a retro focus limitation? That 28 Nikkor is a highly touted lens, but at this point I feel like selling it and putting the money towards a RF 28 like the new Voigtlander.

[Rob F.]

Andrew, I have the impression that this is pretty much a consequence

of the way perspective is rendered by wide angle lenses. The 38mm

Biogon on my SWC does the same thing. Shapes near the corners seem

to get drawn or sucked toward the corner. I haven't noticed it with

my 28mm Elmarit, but then I haven't done much architectural work with

it yet. Now that you've raised the point, I'll be keeping my eyes

open for it.

[andy_piper2]

The 28 2.8 AIS Nikkor is possibly one of the THE best SLR 28s ever

made (and one of the best Nikkors). According to Photodo.coms MTF

charts it is right up there with the Contax G/Leica/Konica 28s.

 

<p>

 

But MTF and distortion are two different animals. I never noticed much

distortion when I was using the Nikkor, especially compared to the old

24 from the 60's, which made everything near the edges look as wavy as

Poirot's mustache.

 

<p>

 

If you are going to keep some Nikon stuff around, this is one of the

real keepers in their line, but in theory the RF lenses should have

less distortion.

 

<p>

 

But bear in mind that the current Leica and Voigtlander 28 designs are

actually slightly retrofocus designs, to provide clearance for the

metering systems of the M5/CL/M6. So they may actually show some

distortion compared to the '60s era 28's and the Zeiss G-Biogon, which

is a virtually symmetrical design. The Konica Hexanon-M is also very

close to symmetrical, but I've never tried one.

 

<p>

 

So test before dumping the Nikkor!!

 

<p>

 

As an aside, has anyone tried the Konica Hexanon-M 28mm? Comments,

image samples? The Photodo charts make it look slightly better than

either the Zeiss G or Leica M lenses!!(?) I just shot camera-counter

comparisons between the 28 'cron and the latest 28 Elmarit today -

will swap my impressions for Konica impressions as soon as I get home

and process the film.

[andy_piper2]

Here's a shot (no artistic merit) with the 28 Elmarit. Note how the

guy's shirt is much wider towards the corner (makes him look really

heavy!), and how the picture frame at top right is not square. This is

the normal perspective effect of a wide angle lens, but NOT distortion

as defined optically.

 

<p>

 

<center>

<img src="http://www.photo.net/photo/330295" width=500

height=350>

</center>

 

<p>

 

Optical distortion is precisely defined as a part of the image being

further or closer to the center than you would expect from an "ideal"

lens (the "bent lines" idea - or what you see with a fisheye lens).

What you may be seeing (and what this picture shows) is the exaggerated

perspective from a wide-angle, which is common to any WA from RF to SLR

to view camera (Look up Bruce Davidson's book "East 100th Street",

mostly shot with a superwide on a 4x5).

 

<p>

 

Now let's see if my link worked....

[alan_ball1]

Andrew,

 

<p>

 

I've experienced that phenomenon with extremely well corrected wide

angles on rangefinders as well as on SLRs.

 

<p>

 

I'm therefore also quite convinced that this type of "elongation" on

sides and corners is strictly a perspective issue, and not a

distortion/correction issue. It does not appear at all distances with

all subjects: you will notice it more often in close-ups, while

shooting 3 dimensional subjects.

 

<p>

 

The way you hold the camera (strictly parallel to subject plane or

not) also influences the phenomenon. Very noticeable with people

(crowd shoots), and more visible the closer and the wider you go.

 

<p>

 

Try shooting a patterned flat subject(newspaper), with your film

plane parallel to subject plane, and I'm quite sure you will NOT see

that "stretching" of straight lines towards the corners. You will

notice that you have a very well corrected almost distortion free

lens.

 

<p>

 

Alan

[andy_piper2]

Oooops - don't know what happened to my image.

 

<p>

 

<a href="http://www.photo.net/photo/330295">TRY

CLICKING HERE</a>

 

<p>

 

Alan is also right on, as I hope you'll see.

 

<p>

 

Andy

[sait_akkirman]

I use the same 28mm lenses (latest Elmarit 2.8 and Nikkor 2.8

Ais that focuses to 20cm). Both are very nice lenses. I get the

same feel as you do that when I shoot off centre HEADS ( or 3

dimentional round objects), at near distances, off centre, Nikkor

produces slightly more distortion (stretch) than the Elmarit. This

is just a feeling, I haven't actually done any tests.

[victor_randin]

I met very unpleasant sort of aberration named the COMA, when was

shooting with the Nikkor f2.8/28mm AI, and the Rollei 35S f2.8/40mm.

Any round subject (for example, a face from 4-5 m) displaced near the

edge looked like a pear, and it was impossible to improve with

stopping down. Though, I didn�t pay attention to straight lines near

the edges of the frame. Meanwhile the Nikkor f3.5/28 AI and the

Summaron�s f5.6/28mm, are free from such a sort of aberrations.

[andrew_schank]

I just gave a good look at the images and I think it does really only

show up most on closer subjects. I have taken so many images in the

past few years with my 40 Rokkor and 50 Summicron, which of course do

not have the corner perspective distortion of a wide like a 28mm,

that may be why it seemed so noticeable. My 35 f2.8 Summaron doesn't

seem to do this type of corner stretching either, nor does the 35

f2.8 on my Contax T3, but maybe its just because the 35mm is not as

wide as a 28mm.

[jackflesher]

Andy Piper:

 

<p>

 

Nice photo! You should share some more with us!

[Robin Smith]

Andrew

 

<p>

 

Welcome to the world of real wide-angles! This is one of the reasons

why 28mm is about as wide as you can get really if you care about the

look of groups of people (or rather if you are going to share photos

with them). Once you get to 24 or 21mm you can imagine what happens to

people at the edge of the frame! Photojournalists (who often seem to

use them) don't care as the people involved don't talk back and

sometimes the effect is good, but "real" w/a shots of family groups or

groups of friends in which people appear on the edges will not make

you Mr. Popular with those who unhappily chose to stand at the edge of

the frame. You do not notice this effect with a 35mm.

[peter_hughes1]

The effect is similar to that of stretching a globe to fit on a flat

map. It is inherent in the physics of lenses and has nothing whatever

to do with SLR v. RF.

[andrew_schank]

I understand this of course Peter, but it seems to me that like the

elusive "bokeh", some wide angles have a more pleasing way that they

bend a wide 3D world to fit in to that little 24 X 36mm flat

rectangle.

[alan_ball1]

Robin is quite right, and this phenomenon is the price you pay for

the "sleathy" advantage of ultrawides in crowd shots. You do get the

unaware nearby people in the frame, while they think you are

photographing someone/something else.

 

<p>

 

But you are also at risk of inducing a "strangeness" in those

people's features, especially if those at the edge of the frame are

also those who are the nearest to the camera. That strangeness will

or will not spoil the image depending on the informational nature of

the shot: the viewer will gladly relate to the picture if this

information is interesting, rather than be distracted by esthetic

details. On the contrary, if the information content is low or banal,

then the viewer's eye will be attracted by the esthetical "defects"

(in this case unusual perspective phenomenae).

 

<p>

 

That is probably why the ultra-wide is a favourite with

photojournalists covering crowd events (immense depth of field, view

from within the action, etc) but not in most other circumstances

involving people (except of course if you NEED such a "strangeness"

to attract the viewer's attention to your picture).

 

<p>

 

Again, I cannot see why RF wide angles would be a protection against

such perspective phenomenae. They do not in my experience (21mm Apo,

28mm Ricoh). But I am ready to be convinced I am wrong by any solid

argument of course.