photo sharing and upload picture albums photo forums search pictures popular photos photography help login
Token | profile | all galleries >> Noise test and other camera stuff. >> Lens Comparison tree view | thumbnails | slideshow

Lens Comparison

This is a lens comparison to see how a mirror lens, and a good mirror lens, stacks up against a refractive lens.

For this test two mirror lenses were used. One is a run of the mill Vivitar 500mm f8 Mirror Macro, currently selling for well under $200. The other is a Vivitar 800mm f11 Series 1 Solid Catadioptric lens. This lens sold for over $600 in the early 80's, and today sells in the $350 range used.

The other lenses used are a mixed bag of Canon made lenses. The 100-400 L IS, the 100-300 f5.6 L, the 70-200 f4 L, and the 28-135 IS. The 100-400 L IS and the 70-200 f4 L were also used with the Canon TC 1.4 L. All of these combinations were shot at their widest aperture to even the field, since the mirror lenses are not adjustable. All of these lenses were shot in manual focus mode only, since the mirror lenses are not auto focus capable.

Since the lenses are all of different focal length, an attempt was made to keep the capture areas about the same. This was done by moving the camera in relationship to the target. If the lens was 800mm the camera was placed 80 feet from the target. If the lens was 500mm the camera was placed 50 feet from the target. If the lens was 200mm the camera was placed 20 feet from the target.

The workflow for all comparisons are identical. The images were shot using the D60. The images were shot in fine jpg with no in camera adjustments. No saturation boost, not sharpening. Five shots were taken in each combination, the image was manually re-focused for each shot, and the best shot of the group selected. The camera was focused on the "2" in the "Test 1, 2, 3" text. Mirror lock-up was used. The camera was tripod mounted and the Canon TC-80N3 was used to activate the camera.

For each combination the image shows a 1 for 1, 100%, 1000 x 800 crop of the center of the image. The out of focus tack was cut as a 200 x 200 and inserted into the larger crop. The text was then inserted. The images were then saved in PhotoShop as jpg with a quality of 12 selected.
The basic setup
The basic setup		 The tack
The tack		 800mm f11 Vivitar Series 1 Solid Cat
800mm f11 Vivitar Series 1 Solid Cat		 500mm f8 Mirror Micro lens.
500mm f8 Mirror Micro lens.
Canon 100-400 L IS plus TC 1.4
Canon 100-400 L IS plus TC 1.4		 Canon 100-400 L IS
Canon 100-400 L IS		 Canon 70-200 f4 L plus TC 1.4
Canon 70-200 f4 L plus TC 1.4		 Canon 70-200 f4 L
Canon 70-200 f4 L
Canon 28-135 IS
Canon 28-135 IS		 Canon 100-300 f5.6 L
Canon 100-300 f5.6 L

comment | share
Token09-Feb-2011 16:26
Anonymous “Guest”, since you posted the same comment 3 times I have deleted the first 2 and only left 1 copy.

This test was done in 2002 or 2003 in response to a thread on a now defunct forum concerning the raw resolution and picture quality of reflective lenses with a center obscuration vs refractive lenses, and if there was any value to purchasing a more expensive lens. To that end, the original purpose of this comparison, the test is not flawed, in fact it was the methodology agreed to by all parties involved, both pro and con. The attempt was to remove focal length from the equation. The question was, indeed, how sharp is the image across the frame, regardless of how many pixels the subject covered. Specifically the claim was that reflective lenses (regardless of focal length, but the specific lenses most quoted were 500mm) are “good enough” that the image can be fixed in post processing to yield the same results and that buying more expensive lenses was not needed, indeed was a waste of money. The contention of several respondents in the thread was that once the data is lost or corrupted you cannot retrieve it in PP. Others contended that if you start with a better original the end result will always be better for a given amount of PP effort. Not shown in this series of images were the results of the PP, as each person in the thread applied their own to the images in an attempt to prove or disprove their positions.

Why remove focal length from the problem? Because no one in the discussion could agree on how many pixels needed to be on the target to make a satisfactory test. It would be possible to structure a test to make one or the other fail based on focal length. If the target fills the frame then almost any lens looks good, if the subject is 20 pixels tall then the more expensive lens will show the most detail. In the real World you will have all of these conditions at various times, and as such there are times when low end lenses are acceptable and other times when they yield no usable results. As such the decision was made to look at raw resolution only.

However, in response to your specific statements and question, how good is the resolution on the subject, perhaps you missed the test of the Vivitar 500mm f8 at 50 feet and the Canon 100-400 L + 1.4 TC (560mm) at 56 feet? Certainly that meets the criteria of your question of how good is the resolution on the subject?

There is no question that if you have the option to zoom with your feet you will get better results than if you do not. Closer is always better. But, zooming with your feet and using better optics will yield the best results.

Since you state I have done such an injustice to a debate I was not addressing, perhaps you could perform the test you desire, and provide the results for the betterment of the photographic community? In the past 9 or so years there have been many comments in forums or in emails about this series, and the overwhelming number have been positive. Perhaps most of those people took the test for what it was meant to be, and did not take offense because it did not do it the way they wanted it.
Pat Casey 11-Feb-2009 02:22
I have a Vivitar 800mm f/11 Solid Cat Series 1 lens. I am trying to find the "filter set" that was designed to be used with my lens. The filter set has 5 filters in it, and I "believe" the filer size is 35.5 I certainly would appreciate any assistance that you can provide me with in my trying to locate/purchase this filter set for my Vivitar 800mm f/11 Solid Cat Series 1 lens. Thank you.
Pat Casey
patc@everestkc.net
SetterSniper 18-Jul-2008 17:18
I came here researching the image quality of the Vivitar 800mm catadioptric lens. I suppose the answer to the stated intention of this page is that hardly any mirror lens can stand to be compared against the crummiest refractive lens. The Canon 100-400mm "dust sucker" is infamously known as a mediocre lens, especially at 400 mm, and yet even with the 1.4X TC it's visibly much sharper than either of the mirror lenses.

Although the most logical comparison would have been between the Vivitar 800mm and the Canon 100-400mm with a 2X TC, that would be superfluous given the comparison between the 500mm mirror and the 100-400mm with 1.4X TC.

BTW: My research suggests that the Vivitar/Perkins-Elmer solid glass lenses were not particularly good even among the sub-population of catadioptric lenses. The Nikon 1000mm f/11 mirror I use is much better than either of the lenses tested here, and even sharper than my Tamron 500mm and Sigma 600mm. In turn, the Canon 300mm f/4 L with Canon 2X TC is noticeably better than either of these latter two mirror lenses.
Gene Ellis (smeepe) 01-Mar-2006 02:07
I really like the way you set up the comparisons. The 70-200 really looks sharp. Just wondering if you backed down on the 100-400 to about 200mm would that eliminate the small differance in sharpness? This page convinces me to stay away from the mirror lenses.
click on thumbnails for full image