Mamiya 7 Distortion & Winding Issues

I'm pretty recent to the Mam 7 lineup, but have been really enjoying it thus far. However, I have noticed a clear bit of fall-off and barrel distortion in many of my compositions. From what I've heard the system is supposed to be very good with both of these things, but most of my images have to be perspective corrected in post and many still contain irritating dark corners. Has anyone else notice/struggled with this? Could there be an issue with my setup, or lens?

 

At this point I've probably run 25+ rolls through it, and until yesterday have had no problem with the winding mechanism, or loading. I loaded a roll with no indication of an issue or looseness, yet when it came to the last two shots, the winding became stiff, and I heard the dreaded sounds of a roll getting caught. In the end it had come loose to the point that when I took it out it was beginning to separate from the spool and would definitely have let light in had I not stuffed it into a black bag I had. Wondering if anyone has had issues with this on the Mam 7 before?

 

I've included some images to illustrate the first point.

 

Thanks so much!

 

 

Thanks for posting. So we are looking at the same thing, what in the images are examples of barrel distortion and falloff? What aperture were the photos taken with? Handheld, or tripod?

What lens and exposure?

 

I don't see barrel distortion in those pictures, but the straight lines are fairly central, so it is hard to tell. I had the 43mm, 80mm and 150mm, and they were all pretty good. (OK, "pretty" is probably an understatement).

 

The Mamiya 6 & 7 cameras used non-retro-focal lens designs, which helps reduce distortion, but typically results in sin^4 vignetting. Due to the wide latitude of film, vignetting is less of an issue than distortion... unless you like shooting chrome. Then vignetting can be corrected with a center filter. Distortion cannot.

All lenses exhibit some degree of vignetting. Some is due to the angle of incidence at the film. The greater the angle, the same light is spread over a larger area. Mostly it's due to the narrowing of the aperture when viewed off-axis. The latter improves as you stop the lens down. It is very easy to correct when the scanned images are processed in Lightroom or Photoshop. In the Golden Age of Film, it was common practice to "burn" vignetting in practically every print, to pull your eye toward the subject, or to avoid bleed to white at the edges.

 

Distortion, if it exists, is only visible in straight lines, parallel to the film plane, near the edges of the frame. Try paralleling the camera and shooting at a brick wall. Most subjects will tolerate up to 3% distortion (brick walls excepted), and it's practically irrelevant for landscapes.

Thanks for the responses guys!

@Wilmarco Imaging They were shot at f11 hand held.

 

@tom_chow The 65mm. The second image is a particularly good example, where the main content looks fish-bowled. You can see the lines at the bottom of the fence and side rail look very curved, when in reality they were a straight line. You can also see it a bit less clearly in the first image where despite being a considerable distance from the main subject, the image still has bowed or convex feeling as if being pushed from the centre.

 

@Ed_Ingold I'm trying to avoid a vignette effect and rather aiming for evenly exposed skies. I've been doing a combo of vignette correction and burning in post to fix the appearance as well as distortion/perspective correction. Stopping down isn't an option unfortunately since I have been shooting urban landscapes that require as much sharpness in fore, middle and backgrounds as possible.

 

Perhaps I'm just accustomed to the limited distortion and clean lines of a bellows camera, but I've been finding the vignette and distortion very disappointing, especially in camera many consider to be the holy grail of medium format. Maybe, it just doesn't lend itself well to architectural style shooting. In portraits these issues are no big deal and can even be desirable.

Are you using a filter or lens hood, which might contribute to vignetting? I think f/11 would be stopped down enough to minimize circularity of the aperture. There is always the cosine effect, regardless of aperture. Stopping down will reveal even minor vignetting due to a filter or hood.

 

Frankly, I don't see any distortion, other than inherent in aging, shoddy construction. The camera is not leveled nor parallel to some surfaces, which causes visible convergence. Perspective is not distortion.

 

Not all view camera lenses are free of distortion or vignetting. If use a 65 mm lens on a 4x5" camera, you might look more closely.

A winding issue might be a loading issue. Make sure you are threading the film correctly, and that the insert matches the back. MF film is pulled by the take-up spool, and should remain tight unless there is another problem. A bent or mis-aligned roller might have a similar effect. I have an Hasselblad, and unfortunate early experiences with loading (a manual is only used as a last resort).

I'm not seeing barrel distortion. Bring the image into Ps or similar, and drag some guides into the image. It appears that subject and camera were not plumb, level and normal relative to each other. Additionally, a characteristic of wider lenses is that what may appear to be distortion, is in fact the true effect of perspective at a short camera to subject distance. The effect is more pronounced with wider lenses and shorter camera to subject distances.

 

In summary, I don't see any equipment flaws in the images.

Additionally, a characteristic of wider lenses is that what may appear to be distortion, is in fact the true effect of perspective at a short camera to subject distance.

 

- Very true!

Try standing close in front of a long wall or fence. Then turn to look sideways one way or the other down the fence/wall. It shrinks into the distance, no? Your eyes aren't distorting, it's natural perspective.

 

If you could take in the whole 180 degree view at once, you'd see a fence/wall that started small at one edge of your field of vision, curved and grew much bigger in the middle and diminished in size to the other edge of your view.

 

Just because we don't like the look of it in a 2 dimensional representation doesn't mean it's not 'natural'.

Thanks for the responses guys!

@Wilmarco Imaging They were shot at f11 hand held.

 

@tom_chow The 65mm. The second image is a particularly good example, where the main content looks fish-bowled. You can see the lines at the bottom of the fence and side rail look very curved, when in reality they were a straight line. You can also see it a bit less clearly in the first image where despite being a considerable distance from the main subject, the image still has bowed or convex feeling as if being pushed from the centre.

 

@Ed_Ingold I'm trying to avoid a vignette effect and rather aiming for evenly exposed skies. I've been doing a combo of vignette correction and burning in post to fix the appearance as well as distortion/perspective correction. Stopping down isn't an option unfortunately since I have been shooting urban landscapes that require as much sharpness in fore, middle and backgrounds as possible.

 

Perhaps I'm just accustomed to the limited distortion and clean lines of a bellows camera, but I've been finding the vignette and distortion very disappointing, especially in camera many consider to be the holy grail of medium format. Maybe, it just doesn't lend itself well to architectural style shooting. In portraits these issues are no big deal and can even be desirable.

 

I'll take your word for it about the barrel distortion look, but I can't measure any on the 1st shot (but the web resolution is pretty low) and having not seen the live scene, I cannot tell if the curve you see is real or a perspective. But the lens, when new, was quoted as having 0.08% distortion, which is a good as most LF lenses. A proper rectangular flat subject test (ie: like those brick walls I don't like to do) would be the best way to determine if the lens is working properly.

 

Vignetting is always present, symmetrical derived lenses have sin^4 falloff regardless of aperture, but at larger apertures, you may also get the front or rear objective cutoff (which is the source of "cat-eye" bokeh). This is also true for LF lenses.

 

Film jamming is usually due to poor loading technique. The paper needs to be square and centered in the spool. If it is off to one side, or crooked, it causes the paper to bunch up on one side and the film to not wind on tight. Also, keeping the film tensioned or flat (ie: not loose) when the back is closed will mitigate jamming when the paper is not square and centered to the spool. When loading, after the paper starts winding on the spool, I check it's position, and give it a little nudge to get it centered, before winding on to the mark.

symmetrical derived lenses have sin^4 falloff

 

- It's Cos^4.

 

Second time recently it's been wrongly quoted. Sin^4 'falloff' would get brighter toward the corners and be dark in the centre of the frame.

 

Anyway, the theoretical angle of light to image-plane has little to do with vignetting, otherwise we couldn't have fisheye lenses. Vignetting is mainly caused by aperture obscuration or geometry-change, and can be designed out to a large extent.

- It's Cos^4.

Yes, sorry, thanks! Writing stuff down late at night does that. Twice!

 

Fisheye lenses use distortion to avoid vignetting, the periphery which gets less light also gets compressed.

 

For rectilinear lenses, there are also various optical distortion techniques to reduce vignetting. Look into the front of a retrofocal lens and observe the shape and orientation of the aperture, you can see how it is distorted to present a larger opening to the incoming light at steeper angels. The rear corrects for this distortion, and extends the back focus to reduce the vignetting - but it can only get it to the cos^4 value of the back focus length. Which is pretty good. There are a few other optical techniques, but these are the basic limits of refractive optics.

 

The reduction of vignetting that you see on fast lenses is because the front and/or rear elements are undersized to reduce weight and complexity. Thus they cut off part of the aperture at large ray angles, resulting in larger values of vignette than cos^4. You can visually observe this in a lens but peering through the back at the angle of the raypath to the corner of your image plane, and see the edge of the front (or rear) lens occluding the aperture. Closing the aperture down till it is all clear tells you the aperture value that that particular lens will then only exhibit cos^4 vignetting (base on the back focus angle).

 

But for the OP, LF camera images at their optimum aperture should exhibit ~ the same amount of vignette as the Mamiya-7 lenses at their optimum apertures - provided we are talking about non-retrofocal lenses (symmetrical designs) at the same FOV for the format. Most LF lenses (for film) are symmetrical designs, and all the Mamiya-7 lenses are.