21mm vs 25mm distortion on a 1.5 crop factor sensor

Redseele

Established
Local time
1:36 AM
Joined
May 18, 2013
Messages
161
Location
New York City
I wonder if anyone can help me with this.

I have a Fuji X-E1, which has a sensor crop factor of 1.5x. Right now I'm trying to get a 35mm equivalence for my main lens. I do not like the Fuji autofocus lenses, I like manual focusing.

A friend just lent me his beautiful Voigtlander Skopar 21mm lens (31.5mm equivalence), but I noticed that there's a lot of distortion in it. Do you think that a 25mm (a 37.5mm equivalence) would not have this distortion?

I used to have a Fuji X100 (35mm equivalence) and I really liked it because it seemed to me that it didn't distort the image at all (even though in theory it was a 23mm lens).

Anyway, if anyone could help me with this bit of optics theory I'd really appreciate it 🙂
 
Yes, but I also remember reading a lot about how the 21mm was also known for its lack of distortion, and yet it does have quite a bit of it apparently, even on a crop sensor.
 
Yes, but I also remember reading a lot about how the 21mm was also known for its lack of distortion, and yet it does have quite a bit of it apparently, even on a crop sensor.
I use the lens on film and the M8. It has minor barrel distortion, but it never bothered me nor required correction. If you are not shooting a lot of architecture, you should not have a problem.

You may be talking about perspective distortion. You need a longer lens to correct that (alternatively, step back and crop).
 
I use the lens on film and the M8. It has minor barrel distortion, but it never bothered me nor required correction. If you are not shooting a lot of architecture, you should not have a problem.

You may be talking about perspective distortion. You need a longer lens to correct that (alternatively, step back and crop).

Perspective distortion? I think this might be my problem (certain things seem elongated if I take them from certain angles). Can someone explain this a little further? Also, would a jump to 25mm make this significantly better?

The reason for this is that I'd like a 35mm equivalence to be able to shoot indoors.

Thank you
 
In photography, perspective distortion basically refers to how your lens transforms the near and far objects in the frame due to their relative scales vs. the same framing through a normal lens. A wider lens will make you move closer to achieve the same framing. The close object thus gets larger in size. This is extension distortion. It may look painful at the frame edges and even in the center when you get too close. A long-focus lens gives you compression distortion (like a smaller nose in a portrait or a compression of distances in a landscape).

If this is the problem you see, you need to find the longest "35mm equivalent" that you are comformtable with. A 25mm lens may very well do the trick for you.
 
There are several types of "distortion":

1. Perspective Alteration (I prefer this to "Perspective Distortion") - This is not a real distortion as it is actually a faithful reproduction of the scene. What is "odd" is that the perspective (relationship of forground object to background object) is not what the viewer expects. It is causes only by the choice of working distance and is not a property of the lens. The lens' only influence is the framing.

2. Rectilinear Distortion - this is the primary real distortion seen in camera lenses. It is usually either barrel distortion, pincushion distortion, or some odd mix of the two. It generally manifests itself as straight lines not being straight.

3. (name=??) - there is another distortion that comes into play when the film/sensor is flat (rather than spherical) and the exit pupil of the lens is not an infinite distance from the film/sensor, that is "image side telecentric". This is most commonly seens when a large group portrait is taken using a wide angle lens. The faces nearest the corners of the image will seem stretched radially from the center of the image. Modern lenses made specifically for digital camras (e.g. not legacy film lenses) are generally partially image side telecentric designs (exit pupil is very far from the sensor though not necessarily an infinite distance). This is done to reduce/eliminate the issues with light striking the sensor at an angle but has the secondary advantage in that it eliminates or reduces the distortion associated with wide angle film lenses.
 
Olympus Pen-F 25mm F4 half frame lens on my X-E1, and don't notice any problems. But I'm not sure what you're seeing.
 
Perspective distortion? I think this might be my problem (certain things seem elongated if I take them from certain angles). Can someone explain this a little further? Also, would a jump to 25mm make this significantly better?

The reason for this is that I'd like a 35mm equivalence to be able to shoot indoors.

Thank you

This the kind of "distortion" that you get any time you point the lens up or down. This happens with any kind of lens, except shift lenses that can correct for this to an extent. As others have pointed out, you can get a longer lenses and shoot from farther away. Being farther away results in you having to point the lens up less to capture the same scene and, therefore, you get less perspective "distortion".
 
As long as we are talking about geometric distortion in the sense that straight lines are rendered as curves, you could fix them in software. It seems that several manufacturers employ in-camera corrections for their own lenses anyway.

You can think of lens distortions much the same way as map projections. The lens maps a the 3D scene in front of it onto a 2D sensor surface in an arbitrary way chosen by its designer. We mostly use rectilinear projection but not always - notable exceptions are fisheyes and anamorphic motion picture lenses. But you can always reproject your image - in the early days of digital when good ultrawides were few and far between, it was quite common to use fisheye lenses and "straighten" the images in post.
 
+1 for this.

The third type of distortion Dwig mentioned is called volume anamorphosis distortion by DxO. At the edges of the frame circular objects appear oval and square objects appear as rectangles. This can happen at any focal length but is most often seen with shorter focal lengths. Subject distance is also a factor. As far as I know, this is a fundamental issue with optics when a three-dimensional object is projected onto a two-dimensional surface (lens quality is irrelevant). I think the same effects would happen with film. DxO can correct for this distortion, buy I have not tried this myself.

And, keystoning or converging verticals is yet another form of optical distortion (again lens independent) when the camera sensor or film is not square to the subject in both planes. Tilt-shift lenses can eliminate converging verticals.


There are several types of "distortion":

1. Perspective Alteration (I prefer this to "Perspective Distortion") - This is not a real distortion as it is actually a faithful reproduction of the scene. What is "odd" is that the perspective (relationship of forground object to background object) is not what the viewer expects. It is causes only by the choice of working distance and is not a property of the lens. The lens' only influence is the framing.

2. Rectilinear Distortion - this is the primary real distortion seen in camera lenses. It is usually either barrel distortion, pincushion distortion, or some odd mix of the two. It generally manifests itself as straight lines not being straight.

3. (name=??) - there is another distortion that comes into play when the film/sensor is flat (rather than spherical) and the exit pupil of the lens is not an infinite distance from the film/sensor, that is "image side telecentric". This is most commonly seens when a large group portrait is taken using a wide angle lens. The faces nearest the corners of the image will seem stretched radially from the center of the image. Modern lenses made specifically for digital camras (e.g. not legacy film lenses) are generally partially image side telecentric designs (exit pupil is very far from the sensor though not necessarily an infinite distance). This is done to reduce/eliminate the issues with light striking the sensor at an angle but has the secondary advantage in that it eliminates or reduces the distortion associated with wide angle film lenses.
 
I think the OP is describing perspective distortion, not optical distortion, caused by an inferior lens.

Don't tilt the camera up or down and most likely the distortion will disappear.

The difference in focal lengths, crop factors, etc. are really not relevant when considering excellent RF wide angle lenses.
 
I think the OP is describing perspective distortion, not optical distortion, caused by an inferior lens.

Don't tilt the camera up or down and most likely the distortion will disappear.

The difference in focal lengths, crop factors, etc. are really not relevant when considering excellent RF wide angle lenses.

Tilting the lens does not create perspective distortion. It creates converging verticals and, or horizontals. Perspective distortion is 100% dependent on the subject to lens distance.
 
Hi!

These lenses were designed for 24 x 36 and the 1.5 crop sensors are 18 x 24....so don´t mind about distortion...

The edges of these lenses might show "flaws" but your tiny sensor will eliminate them while eliminating half of the frame!!!

Go shooting without any concern 21-25 mm lenses have huge dof so donñt even mind about focusing!
 
I have the 25mm Skopar, and can agree it is a low distortion lens. I'll just add a comment to the info already given on perspective distortion. I shoot a lot with wide angle lenses, and have often noted that they stretch the image progressively more the more closely the image extends to the edge. When the image extends toward the corners, the effect is further accentuated. I have a shot of a bridge that has a series of stone arches that appear progressively wider the closer they are to the edge (yet they are really all the same width). In another shot taken in the Denver civic center, a spherical stone globe on a column in the upper corner appears stretched into an oval. In both cases, the camera was the Hasselblad Superwide C with 38mm Biogon.

It is not distortion, it is the consequence of viewing things at such a wide angle. There are two reasons why things don't look that way when we view them with our eyes instead of a camera: for one thing, we don't see at such a wide angle. And for another thing, our brains automatically correct the info it gets from the eyes, to conform the image to what it knows the thing is supposed to look like.

There is a way to take a wide picture without this stretching effect. You have to use a swing-lens camera, like a Widelux. But then, instead of stretching, you get curvilinear distortion, or barrel distortion. Aim the widelux, perfectly centered and leveled, at a long, low building, and the roof line curves down at the left and right, while the floor line curves up. It happens for the same reason that the top and bottom rails of a fence appear to come together to a vanishing point in the distance.

So, there's no way out of perspective alteration with a wide angle of view. It's not even alteration: it's how things really are when viewed wide. Or, as a favorite professor used to say, "TANSTAAFL!" (There Ain't No Such Thing As A Free Lunch.)
 
Back
Top Bottom