Mirror lens test
On another page, I argued that mirror lenses have a poorer performance than refractive ones. This has been verified independently by several photographers, as a web search can easily prove. However, to satisfy my curiosity and see the results for myself, I decided to carry out tests with the lenses I have at my disposal.
As a mirror lens, I used a Sigma R 600 mm f/8 (above). As a term of comparison I already knew to be of good quality, I took a Tamron SP AF LD 300 mm f/2.8. Both lenses are several years old. In order to perform a meaningful test, I used the Tamron 300 mm with a 2x Kenko teleconverter. This gives the same effective focal length to both lenses, so the results can be compared directly.
Several years after I wrote this page, I started using an Olympus OM-System Zuiko Reflex 500 mm f/8 mirror lens. This lens is one of the exceptions to the generally poor performance of mirror lenses. The results described on this page are still valid for the Sigma R 600 mm f/8, but do not apply to all mirror lenses.
The question I wanted to answer can be stated simply as: given a choice between a mirror lens which is said to be among the best ones (or at least, not among the worst ones), and a refractor lens of good quality with 2x teleconverter, which performs better? The alternative between these two lenses is very real for me, because I happen to own both. In terms of price, the comparison is not fair, because a mirror lens costs about one-fifth to one-tenth the price of a good refractor lens. However, many photographers (especially beginners and/or photographers with a limited budget) do consider purchasing a mirror lens at one time or another. This test can be directly useful to them.
I carried out the test by photographing a piece of furniture with a seashell placed on one of its shelves, from a distance of about 7 m in my living room. The Sigma 600 was focused manually, as accurately as possible (I relied on the electronic focus indicator in the viewfinder in addition to judging sharpness by eye, but both my eyes and the indicator had some trouble), while I let the Tamron 300 autofocus with the central sensor, which it did without problems because of the larger lens aperture. Illumination was provided by an incandescent bulb for focusing, and an SB-800 flash remotely controlled by the built-in flash of the D200 for the actual exposure. This eliminates vibration as a source of unsharpness. To make the test useful in real terms, I stopped down the 300 mm to f/4 (i.e., f/8 effective, because of the teleconverter), thus using the same effective aperture as the Sigma 600 mm. I already knew from a previous test that the 300 mm is considerably sharper when the diaphragm is closed one stop, but for a real-life test I intended to use both lenses in the same real-world conditions.
The following pictures are 1:1 portions of a D200 frame.
The above results speak for themselves. In spite of using a 2x teleconverter, the amount of detail provided by the Tamron 300 is far superior. Even the region from the corner of the frame contains detail around 1-2 pixels in size, i.e., at the limit of the resolution of the 10.2 megapixel D200. This confirms other tests I discussed here. The Sigma 600 clearly fails to resolve detail at this and a larger scale.
In addition to their lower resolution, mirror lenses are known also for providing low contrast, and being susceptible to flare. These faults do not emerge in the present test, probably because illumination was restricted to the test frame. Perhaps, one reason why flare is common in mirror lenses, besides their intrinsic optical properties, is that their lens shades are much shorter than those used on refractor lenses of comparable focal lengths. Photographers who choose mirror lenses do so partly because of their small weight and size, and typically do not have experience with professional-level refractive telephoto lenses, so they may not think about using a lens shade longer than the mirror lens itself. However, doing so might reduce the amount of flare and loss of contrast. I would suggest the owners of mirror lenses to try a rolled-up sheet of black velvet paper as a lens shade longer than the standard one, wrapped around the original shade and held in place with a rubber band.
While the quality of the image provided by the Sigma 600 is visibly much worse than that provided by a good 300 mm refractor lens, the results with the Sigma 600 are not that bad, if compared with a cheap consumer zoom at 300 mm used with a 2x teleconverter. In fact, based on previous tests, I should expect roughly comparable results. Typically, consumer zooms provide a maximum aperture of f/5.6 to f/6.3 at 300 mm, so their effective aperture becomes f/11 to f/13. In this respect, the Sigma 600 is a 1 to 1.5 stops faster lens. Its main drawbacks are the lack of autofocus (however, also with a consumer zoom you are very likely to be forced to use manual focusing in these conditions), and the rendering of out-of-focus points of light as doughnut-shaped halos (which is an unavoidable consequence of their optical construction).
Thus, a comparison between a cheap consumer zoom with 2x teleconverter and a mirror lens leaves no clear winner. A second-hand mirror lens of comparatively good quality does make a cheap and viable introduction to super-telephotography, unless you need better quality and/or already have a 300 mm telephoto lens that can accept a teleconverter without significant losses of performance. However, to avoid disappointments, you should avoid buying a decidedly bad mirror lens. Stay clear of the very cheap ones currently sold on eBay (Opteka, Samyang and the like). They are not worth even their low prices. Probably you should pass up also on Soviet and Russian models, both old and new (e.g., Rubinar, MTO, Zenitar) and Chinese ones (Seagull/Haiou, which might be interesting only as a collector's item). Do not buy a Vivitar Solid Cat (also a collector's item, not a lens to use). Do not use a small astronomical telescope (unless you already own one for different purposes). Instead, try to get a second-hand Reflex Nikkor 500 f/8 (especially the most recent model, with a red line on its barrel) or the Sigma 500 mm f/8 and 600 mm f/8. The Tamron 500 mm f/8 is probably a re-branded Sigma (or vice versa), and a comparable performer. There might be a few more Japanese mirror lenses of acceptable quality, but I have no first-hand experience with them (e.g., the Kenko 500 mm f/5.6, of which I know nothing, and the Sigma 500 mm f/4, perhaps one of the fastest mirror telephoto ever made).