Can lens defects, such as scratches or fungus, improve image quality?

Question

There are cases where significant damage to lenses, such as major front-element scratches, have been shown to imperceptibly affect image quality. This has led me to wonder: Are there any known cases where lens damage, such as (but not limited to) scratches or fungus, have led to an improvement in image quality?

• The usual metrics for image quality may be considered: Sharpness, Resolution, Contrast, Chromatic Abberation, Field Curvature, Color Accuracy, etc. Since improvements with respect to one metric often involve trade-offs in others, it would be reasonable for answers to describe what aspect of image quality is improved.

• Image quality is not about how pleasing an image appears, which I would refer to as image character. For instance, a low contrast lens has low image quality in terms of contrast, but may produce an image with pleasing "character". Many film-era lenses have this "problem.

Some examples:

• Non-damage alterations: The addition of air-glass interfaces to a lens has the potential to reduce sharpness, contrast, resolution, color accuracy, etc. However, doing so does not necessarily negatively affect those metrics. Rather, additional lens elements are necessary to increase image-quality as measured by various metrics.

• Non-photographic scenario: A guitar a friend of mine used to own. It had a large 3cm hole in it that was unintentionally produced when someone dropped or hit it. Despite, or perhaps ¿because of?, that hole, he said it was the best sounding guitar he had owned. He was referring both to the character of the sound as well as more objective sound qualities such as the ability of the guitar to project sound.

  Suppose we measured the amount of resonance produced by the guitar before and after punching the hole. Like scratches to lenses, most holes in guitars probably don't affect resonance much. Too many holes, or holes in the wrong place (like scratches on a rear element), and resonance would worsen. But happen to punch a hole in the right place, and resonance would increase.

• Hypothetical scenario: Suppose I were to tinker with a lens and happened to damage the aperture blades so that the aperture were to be irregularly shaped. I might later notice that the lens had become sharper (as measured by photographing resolution charts). However, I might also be unhappy with its effect on bokeh (character).

Answer 1

Here is a fault-ridden image of mine showcasing image defects of a wide converter lens (corner softness, chromatic aberration, lens glare in connection with the main lens). I actually like that image. It draws the eye to the good center. Granted, the blown highlights aren't a problem of the lens but they add to the overall character of imperfection.

And arguably a limited depth of focus is an imaging defect that has shaped photography and framing with defocusing to a significant degree. If you look at photography instructions from the middle of last century, a limited depth of focus is usually portrayed as a necessary evil you have to weigh against motion blur and film sensitivity. And now we have a bokeh cult around it.

Answer 2

Cracked glass won’t improve image quality. Think of a lens like a pair of glasses on your face. If your glasses are cracked, scratched, smudged etc. you can’t see as well. I wear glasses and I’m rather rough on them and have to clean them to continue seeing clearly and have multiple scratches which like the answer provided already demonstrated gives a lens flare or bokeh effect. I get some “bokeh” effect looking through my glasses at street lights at night. But rambling aside, a complex functioning machine with any broken element necessary to perform properly will not produce higher quality stuff. To get more technical your lens reflects light from the outside world and puts that reflection on to your cameras sensor or film or what have you. Damage to your glass will damage image resolution.

Answer 3

I once allowed a camera with an EF 70-200mm f/2.8 L IS II that was attached to a 70" monopod to fall from completely vertical to the ground. The result of the fall was that the lens demonstrated significant tilt, along with being less sharp that it was before.¹ With the camera held in landscape orientation, objects in the left part of the frame that were closer to the camera were in focus while objects in the right part of the frame that were much further away were in focus. Held in portrait orientation using the vertical controls of a battery grip, the focus point was much closer for objects in the lower part of the frame than objects in the upper part of the frame.

_{Canon EOS 5D Mark III + EF 70-200mm f/2.8 L IS II, 70mm, ISO 100, f/2.8, 1/5000. Although it's harder to see at web size, the sharpest point of focus near the bottom of the frame is the grass just behind the closest music flip-folder. At the lower platform, the plane of focus is at the middle of the scaffolding. At the upper platform the far part of the rearmost part of the scaffolding is in focus. By the top of the frame, the utility pole behind the tower is in focus! Also, compare the top of the large smokestack in the background to the lower parts that are visible. Below are 50% crops of areas near the top, middle, and bottom.}

If you've ever desired to use an EF 70-200mm f/2.8 L IS with about 8-10° of tilt, I can show you just which angle to let it hit the ground to get there.

_{¹ After sending the lens to Canon service, it came back at least as sharp as it had ever been before, perhaps just a bit sharper.}

Answer 4

I will add examples as I encounter them.

• Anti-reflective coatings were discovered when an unintended effect was examined more closely. According to Wikipedia:

  The simplest form of anti-reflective coating was discovered by Lord Rayleigh in 1886. The optical glass available at the time tended to develop a tarnish on its surface with age, due to chemical reactions with the environment. Rayleigh tested some old, slightly tarnished pieces of glass, and found to his surprise that they transmitted more light than new, clean pieces. The tarnish replaces the air-glass interface with two interfaces: an air-tarnish interface and a tarnish-glass interface. Because the tarnish has a refractive index between those of glass and air, each of these interfaces exhibits less reflection than the air-glass interface did. In fact, the total of the two reflections is less than that of the "naked" air-glass interface, as can be calculated from the Fresnel equations.

• Peripheral haziness of a lens element may smooth out bokeh, similarly to an apodization filter, but at the expense of increasing susceptibility to veiling glare.