24 f/2.8 AF-D - Is this distortion, or something else?

[shawn_rahman]

To All:

 

I just got my first roll back using my new Nikkor 24mm f/2.8 AF-D. I

have little experience with anything wider than 35mm, so perhaps I

need to adjust my perspective a little with this new lens.

 

The above (otherwise non-descript) picture was taken with with this

lens. The flagpole and the trees to the right of it are, in fact,

upright, yet you can see how they appear so distorted.

 

My inexperienced questions for those who are familiar with this lens

and perspective:

 

1) Why is this distortion not so pronounced on the left side? I

think I understand the basic concept of barrel distortion, but that

does not appear to be the problem here, unless I am mistakken.

 

2) Is distoriton of this type more pronounced depending on where I

am situated with the camera and what I am shooting? I have to

imagine it is.

 

3) Isn't this lens supposed to have some corrective property built

in?

 

I am otherwise VERY pleased at how sharp all my pictures really are,

as well as the amazing colors I got this past weekend, but would like

to learn to minimize the distortion, if it is, in fact, preventable.

 

Thanks in advance.

 

FYI - Picture Details: Nikon F100; 24 f/2.8 AF-D; Fujipress 400

Color Print Film; Circular Polarizer.

[dave_schneider]

You weren't holding the camera level, were you. I doubt that you'll notice any realy distortion with this lens. This is the effect of a wide angle lens when you aren't keeping the film plane vertical.

[EricM]

Hopefully Ellis will respond to your question...But me thinks this lens corrects the horizon when held horizontally, turn your camera 90 degrees. You really should look for Ellis on this forum for this plight.

[kam_kozan]

The previous comment is right. I own this lens; it has no barrell distortion. What you are seeing is normal wide angle distortion when you angle the lens upwards. If you want to avoid it you have to hold the camera parallell to the ground. If that does not give you the composition you like you can crop the bottom part later.

[harry_akiyoshi]

Converging verticals. People buy shift-lenses and view cameras (with movements) to avoid this problem, but you could always just hold your camera level and crop out the unnecessary foreground later.

[grant_.]

maybe its the film

[Carl Stone]

This is not the lens for a polarizer, that sky is very uneven, and the polarizer is the

culprit.

[EricM]

maybe dark film for skin tones?

[alex_lofquist]

This phenomenon is referred to as perspective distortion: It is not a true aberration but merely how we perceive a three dimensional object imposed on a two dimensional surface. If you had a flagpole on the left side of the frame you would probably have seen its image bend to the right. I would suggest getting the E-screen (grid) for your F100. It will point up situations as you have just encountered. Until Nikon makes a more modern perspective control lens for wide angles, your only alternative is the AIS PC Nikkor 28mm f/3.5. It is strictly a manual lens and very expensive even on the used market.

[ilkka_nissila]

The PC Nikkor isn't even AI, let alone AI-S. It has a manual diaphragm. It's an excellent lens but quite expensive. Your best bet short of getting a view camera is just getting used to the keystoning effect. Use it to your advantage.

[david_h._hartman]

Just a thought...<br>

<br>

When folks tip a camera up and the sides of the building converge

may call this distortion. If they point a camera down a path,

road, railroad tracks, picket fence and such the convergence

looks very natural, right? <br>

<br>

I call all these "perspective effect." I dont

consider them distortion as such. Am I constant? When a ball or a

persons head is near the corner of the frame with a 24mm or

wider lens and it looks oblong or egg shaped, then I call it

distortion. Nope, guess I not.<br>

<br>

When the word distortion is paired with lens I think of linear

distortion (barrel/pincushion).<br>

<br>

---<br>

<br>

<em>"...but would like to learn to minimize the distortion,

if it is, in fact, preventable." -- Shawn Rahman<br>

</em><br>

Basically you can do three things: hold the camera back

perpendicular to your subject (Lex can explain this), live with

it or exploit it.<br>

<br>

Regards,<br>

<br>

Dave.

[runkel]

It's not the lens. The converging lines phenomenon is unavoidable whenever a camera with film plane parallel to lens plane is angled upward, downward or sideways. Convergence results from projection effects, not distortion, and will occur regardless of the lens type or focal length involved. However, projection and convergence effects are much more pronounced with lenses of short focal length.

 

Think of your lens as a slide projector and the film plane as a screen. If you move the screen closer to the projector, the projected image becomes smaller, and if you move the screen further away, the image becomes larger. What happens when you angle your camera upward to photograph a flag pole is that the bottom of the film plane moves closer to the lens and the top moves further away. Since the image projected on the film plane is inverted, the top of the flag pole is now being projected smaller and the bottom of the flag pole, now projected on the farthest part of the film plane, is being projected larger. Mostly these projection effects are noticeable when the compostition includes lines (vertical or horizontal, as the case may be, such as on the right side of your photo), but they are always present when the camera is angled (including, unnoticeably, on the left side of your photo).

 

The shorter the focal length, the larger the differences in film plane distance become in proportion to the focal length, and the greater the differences in the sizes at which different parts of the image are projected. Imagine an upward camera angle that results in the bottom of the film plane (top of image) being 10mm nearer to the lens than the top of the film plane (bottom of image). If a 200mm lens is mounted, the difference in projection distance between top and bottom amounts to 5% of the focal length (10/200). If a 20mm lens is mounted, it's 50% of the focal length (10/20). Obviously, the convergence "distortion" with the short lens is much more dramatic, even though the same principles are at work in each case.

 

Similar projection geometry explains the "distortion" of objects at the edge of the frame in wide-angle photographs.

[lachaine]

It's perspective distortion, which is normal for any wide angles lens like a 24 depending on how you use it. There is no barrel distortion in that photo. Barrel distortion means that straight lines are rounded, like a barrel. Pincushion distortion is the opposite. The 24mm 2.8D has no such linear distortion. And, I guess you now know why it's not practical to use a polarizer on a super wide angle: the effect on the sky is uneven, due to how much angle the lens covers (the polarizing effect in the sky is dependant on the angle of to the sun, but a 24mm lens just covers too much of it).

[ellis_vener_photography]

keep thecamera level and the back square to the vertical plane. a two way level fitted

to the "hot shoe" of the camera helps

[ShunCheung]

What you are seeing is <B>converging verticals</B>. See <A HREF="http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=004wFm">this other thread</A> on a similar subject. The "brick wall" image I posted there shows some pretty serious barrel distortion from a typical 24-xxx zoom at 24mm. That is a completely different type of problem.

[shawn_rahman]

Thank you, all, for the incredibly informative information. This is, by far, one of the more educational threads I've read in a long time.

 

I will stick to keeping the camera level, and note that on shots that I did, no distortion of this type is visible. And no more polarizer on this lens, for obvious reasons.

[tcb.photo]

Hi Shawn

 

Definately it's the angle of view. By not holding it level you get these "effects". I did the same thing with my 20-35mm, at 20mm. So Now I use a tripod to make sure it's held level. I've seen Leica, and Contax lens do the same thing

 

Best of luck

 

Rob

[ShunCheung]

By the way, this link: <A HREF="http://www.vanwalree.com/optics/distortion.html">http://www.vanwalree.com/optics/distortion.html</A> has a very good description on various types of distortion.

[albert_smith]

With some care and a good look through the finder, you can keep the vertical lines vertical. You need to keep the camera back straight up and down, and don't forget to try to turn the camera vertical if things don't fit in the composition in the "landscape" orientation.<P>

 

Some shots with a 24mm Nikkor where I try to keep everything correct:

<a href="http://www.photo.net/photo/1163139"> photo 1 </a>, <a href="http://www.photo.net/photo/1163147

"> photo 2 </a>, <a href="http://www.photo.net/photo/1163152"> photo 3 </a>...And when I was sloppy and had some angle that I did not catch before tripping the shutter:<a href="http://www.photo.net/photo/1163143

"> photo 4 </a><P>

 

The 24mm Nikkor is no worse than any other wide-angle, it just takes some care to avoid the classic converging vertical error.

[scott_warn]

Like everone else said this is perspective distortion or convergence. What ever label you want to stick on it, it is the result of the film plane not being perpendicular to the subject. The old building falling backwards effect. With a view camera you can put a level on the flim holder and raise the lens for composition so you don't see this effect. With 35mm it's either a Perspective Control lens or Post Processing. The good news is that it can be corrected digitally in an image editing program. In the dark ages before digital, you would have to tilt the paper carrier on the enlarger and swing the lens using the Sheimphlug (I know the spelling is way off - you spell it) principle to maintain sharpness until the lines were square.