Loss of Contrast in A55 due to Mirror

Started Dec 7, 2010 | Discussions thread
Shop cameras & lenses ▾
ForumParentFirstPreviousNextNext unread
Flat view
Senior MemberPosts: 1,471
Loss of Contrast in A55 due to Mirror
Dec 7, 2010

Early on I stated that I felt there would be loss of IQ in SLT cameras due to the mirror sitting at an angle along the optical axis, unlike lens elements and filters which, modulo some small tolerance error, sit perfectly perpendicular to the optical axis. I probably would have never concerned myself with this if not for statements coming from Sony that they feel that this technology will eventually be the one and only way forward, once they think they can improve on the 850/900 OVF with an EVF. With those words I decided that I wanted to understand that future well enough to decide if I want to be a part of it.

At the root of this concern is a basic understanding of refraction.

When a light ray hits the surface of a piece of glass perpendicular to that surface(angle of incidence = 0) as in the case of a conventional filter the direction of that light ray is not changed because refraction does not occur. In reality light does pass through a filter at various angles, though light hitting the filter at angles greater than approx. A/2 (where A is the diagonal AOV of the lens in use) do not form the primary image. With an add-on filter, the refraction of all light rays with angle of incidence greater than zero is inevitable, with the angle of refraction depending on the nature of the materials used and the angle of incidence. The result is image degradation ranging from insignificant to severe -- study any comprehensive review of various grade filters for examples at various points along this spectrum. This is one of the reasons that photographers looking to optimize their optics are often advised to not use filters solely for protection, but to use them only when their benefits outweigh their potential downsides e.g. polarizers, GND, and so on. I do not intend to spark a discussion of such advice, only to point out that there is a school of thought that says do not put unnecessary items in the optical path, and explain some of the reasoning.

In the case of the lens elements, they almost always have a convex surface on the outside of the front element and then various combinations of concave and convex surfaces on various numbers of elements within the lenses. These surfaces are all carefully designed by the lens engineers to guide the light to a point of focus at the film or sensor, and as mentioned above are carefully aligned perpendicular to the optical axis. We don't mind the number of elements put in a lens because they are carefully designed to contribute to high quality imaging. How well this is achieved in various lenses is another discussion entirely.

The above is just background to explain that best practices in optics are to place any optical element in the optical path perpendicular to the optical axis. The following diagram has been posted here before by user Ehrik and again by myself. To preclude any misunderstandings as happened the last time around, the diagram is not meant to be a proof, but illustrative only. What it illustrates is how unintended light paths are created by internal reflections.

The black lines are the air/glass interfaces on the front(left) and rear of the mirror. A ray of light coming in from left to right will reflect to some degree upward into the AF array(not depicted) and some light will be refracted into the mirror material. When the ray encounters the rear surface of the mirror, most of the light will refract toward the sensor but some will reflect forward toward the front surface of the mirror, and a portion will be reflected back in the direction of the sensor. I would expect this portion to be a greater proportion(reflected/refracted) than the light reflection off of the back surface, because the front surface is silvered to some degree. It is this internal reflection that leads to the occasionally-observed ghosting.

I long suspected that even in cases where ghosting is not in evidence, these internal reflections could lead to loss of local contrast, also sometimes called micro-contrast. The internal reflections would cause light to fall on parts of the image plane which should have been directed to its own proper place on the sensor but instead falls on some other part of the image.

As it turns out, this does seem to be the case. There is in fact contrast loss due to the mirror. I do think it's at a level where some will be perfectly OK with it and others won't. I'm in the latter camp; I would not want this technology on a high-dollar camera and I certainly don't see it as a viable technology at the highest end, but for a cheap, fun, goof-around camera I could live with it. Details of the experiment and images in a second post since they are large. I am still working through results of other tests I ran and shots that I took in evaluating other areas of A55 performance, those will follow in a few days in the form of a blog post.

ForumParentFirstPreviousNextNext unread
Flat view
Post (hide subjects)Posted by
ForumParentFirstPreviousNextNext unread
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow