Scotch tape : the alpha and omega of photography

[Lumens Pixel]

Scotch tape is the first item a photograph should have in his camera bag.

I bought a number of lenses, mainly old ones and most were not satisfying in terms of homogeneity. To different extents, you always end up with a blurrier corner or border and hesitate to throw the lens in the bin (good for the general level of photography, bad for the environment; or resell the lens, worse for general photography, better for the environment).

But there is a different path: scotch tape.

I have red long time ago, that when Minolta released the famous MC 58 f/1,2, the lens was “coupled” with a body, often a SRT 101, and that such lens was “shimmed” for that body.

In other words, and especially for wide angles or wide aperture lenses, the register tolerance is quite narrow and, even for analogue photography, slight register differences might impact the quality of imagery.

Nowadays, 24 M pixel is for the people and I could not imagine the impact of manufacturing tolerances on 62 M Pix sensors.

Back to the point: scotch tape is 20 microns thick and, fortunately, bad quality tape is a little less, which provides for interesting options.

Grab your lens and focus precisely on each corner and note the differences.

Here is an example of correction: If on a given corner, focusing distance is less than on other corners (i.e. 5 meters rather than infinity), that means that adequate sharpness is achieved if, for this corner, the glass is farther from the focal plane. To achieve same focusing distance you should then locally add some thickness to the lens mount so all corners would focus at same distance.

Such shimming should be performed on the opposite corner since the lens flips the image projected onto the sensor.

Through trial and error, in a number of cases, you will be able to obtain a better lens. Lenses that suffered from bad manufacturing or have lived traumatising events won’t be corrected through this process.

Zooms can also be corrected the same way provided that the imbalance does not affect a group rotating in the body of the lens in which case zooming or focusing would rotate the shimming direction which would lead to unmanageable results.

Shorter focal length is more sensitive to correction than longer one and, if quite often you only need one layer of tape for a 28 mm, you might need 4 or 5 at 200mm. Sometimes miracles occur and no shimming is needed at all.

Pursuing this path, a pattern might emerge where you would note that a majority of lenses require to be shimmed on same side. Legitimately you would consider that the problem does not lie in the lenses but in an external element: the camera cage, the internal positioning of the sensor, or the adapter if you are using a mirrorless camera.

You gain the right to mitigate the number of the layers of tape, reporting the correction on the adapter or the camera mount and minimising the layers of tape on all the lenses.

I realised statistically that the mount of my A7II was crooked by the value of a layer of tape and restarted the whole process.

Eternal gratitude to Richard Drew, creator of masking tape, and benefactor of photographers around the world.

[Anthony]

I assume this is some sort of a joke.

[Lumens Pixel]

I assume this is some sort of a joke.

Not at all.

[Les Olson]

In other words, and especially for wide angles or wide aperture lenses, the register tolerance is quite narrow and, even for analogue photography, slight register differences might impact the quality of imagery.

Here is an example of correction: If on a given corner, focusing distance is less than on other corners (i.e. 5 meters rather than infinity), that means that adequate sharpness is achieved if, for this corner, the glass is farther from the focal plane. To achieve same focusing distance you should then locally add some thickness to the lens mount so all corners would focus at same distance.

But increasing the distance of the lens from the film/sensor, as with extension rings, means the lens focuses closer, not further away, and it takes quite a lot of extension to have any effect. To move the focus to 5 m instead of infinity you need 5 mm of extension, not 20 microns.

The effect of lens tilt is also well-known, and it is not to make one side of the image blurry.

[pluton]

My previous experience with scotch tape + lenses is limited to the AiS-era Nikkors that use scotch tape to set the focus calibration. The AiS 135/2 was one such lens. Combined with live view, it makes setting infinity focus a relatively quick and painless operation.
One could wish for a non-elastic material to do shimming of mounts and such.  Metal foil has been suggested for this purpose elsewhere.
Recently, the use of a manual focus lens with exact focus markings (Thypoch Simera 28mm f/1.4) allowed me to discover that my Z7II flange distance was incorrect.  I let Nikon USA fix it.

[Lumens Pixel]

But increasing the distance of the lens from the film/sensor, as with extension rings, means the lens focuses closer, not further away, and it takes quite a lot of extension to have any effect. To move the focus to 5 m instead of infinity you need 5 mm of extension, not 20 microns.

The effect of lens tilt is also well-known, and it is not to make one side of the image blurry.

Yes if you add tape the lens focuses closer. In this example that was the aim.

Let me walk you through.


- On a given lens three corners focuses exactly at infinity hard stop and one is blurry.
- In that case, the blurry corner turns sharp when the focus mark on the lens is at 5 meters (select any other value you are comfortable with, but obviously close from the infinity mark, we are discussing lenses in good shape).
- In that case you achieve sharpness at infinity on the "weak corner" with glass farther from the focal plane.
- So if you keep turning the focusing ring and hit the infinity stop you would have focused on that same corner farther than infinity, (from a photographic point of view, I am not an astrophysicist).
- You need then to bring back such corner from "past infinity" to infinity, hence adding some distance to the focal plane.
- Correction is applied at the opposite corner to what you see in the viewfinder since your lens flips the projected image.

We are not discussing tilt shift lenses although the idea is the same. We are correcting minor manufacturing issues to obtain a field of view parallel to the focal plane.

The defect discussed would surely be better corrected if one could locate inside the lens the precise glass that would need adjustment rather than "tilting" the whole lens. But this is far beyond my skills and my equipment and the results obtained are good enough to turn an acceptable lens into a very good one.

Note that your perfect lens would appear not so good if lent to a friend that would mount it on his camera which mount is crooked. This is why I have stressed the fact that it is important to understand statistically if the problem lies more with the lens or with the camera.