Mirror Lenses - a rehabilitation in the digital age?

[david_harris15]

These are links to two images taken with different mirror lenses The first one was taken with a Vivitar 500mm.

 

http://img223.imageshack.us/my.php?image=vivitar500mmmirrorfe6.jpg

 

The second was taken with a Tamron 500m SP

 

http://img100.imageshack.us/my.php?image=tamron500mmspmirrordy7.jpg

 

The difference is quite noticible. Both were taken with the same camera, with the same film, using a tripod & cable release within a minute of each other.

 

Also, although both lenses are stated as being F8, the Vivitar was always a stop slower that the Tamron. The darker viewfinder that the Vivitar gave made it a lot harder to focus.

 

Would suggest steering clear of the Vivitar, or any of the cheap 3rd party mirrors on sale on ebay unless you have seen what they are capable of beforehand.

[douglas_greenberg]

Shun and I have gone back and forth about mirror lenses for several years now, and the debate will probably continue for years to come. He must have truly had a TERRIBLE experience twenty years ago :-)).

[philaret]

Thank you, David, for giving and example of two different mirror lenses. Running on a mediocre mirror lens is not that uncommon and you are not safe with a DSLR here. I have been thinking for a day whether I should comment on Shun's latest post about '20 year old frustration' and a 'warning'. I believe 20 years ago the first autofocus SLR was just out, the Minolta Maxxum 7000, and autofocus was still more of a curiosity. And I am not sure if there was any decent color film with ASA 800+ at that time. I got my Tamron 500mm between '93 and '95 and it took me a couple of years to master this lens and understand its shortcomings. For a while I mostly shot black and white, because I had full control over film speed and contrast. But then came the films with flat crystal technology and life became somewhat easier. At this point of the discussion I feel compelled to present another image which I got today walking around a nearby provincial park. Cloudy day, grey doves with not a lot of contrasty detail to focus on, sitting in the shade amidst a lot of branches...I think everybody would agree that this is a tough situation to focus both manually and automatically. I took several shots in succession using that 'focus float' technique, I also bracketed exposure. One is here for you to judge. With film, you would stand more than a 50% chance ending up with a weak image. No postprocessing except downsizing, but camera image controls (contrast, sharpness and saturation) were set to max.<div></div>

[mark_chappell]

Alexander, since you asked us to judge (I think)....

 

On a small web image like your mourning dove picture, it's not possible to tell much about

sharpness (unless the lens is awful). With respect to other qualities, it's certainly a matter

of personal opinion, but I'm very definitely in Shun's camp here: I simply loath the way

mirror lenses render contrasty, out-of-focus areas. IMO, FWIW, apologies for being blunt,

etc. etc.: the dove image is ruined by those distracting, obnoxious double-edged blurred

branches. It doesn't matter how sharp and contrasty the in-focus subject is if the image

as a whole is marred by an ugly background.

[philaret]

Here is the larger version of the doves:

 

http://img71.imageshack.us/img71/184/img0061ap6.jpg

 

However, opinions indeed differ whether out of focus shapes ruin the image or not, as I said in my original posting. Besides, obstructing foreground shapes are rather difficult to remove by postprocessing, while with the background it is fairly easy using 'Gaussian blur'in Photoshop. I was tempted to do this, but I think it is more informative to show the image as it was taken.

[mark_chappell]

<I>Besides, obstructing foreground shapes are rather difficult to remove by

postprocessing, while with the background it is fairly easy using 'Gaussian blur'in

Photoshop.</i><P>

 

In my experience it's only "easy" to remove or smooth out annoying background

components if the shapes you're trying to remove are not close to the main subject of the

image. Otherwise, it's a royal pain and often impossible to do cleanly -- and even when it

<B><I>is</i></b> possible, it's extra work.<P>

 

I don't always agree with Shun but I'm with him 100% here. As he says, by all means use a

mirror lens if it suits your purposes. I'm sure it's possible to make good images with one,

although I never succeeded. Personally, I've no wish to waste time with a lens that is

effectively crippled for a big fraction of the images I'm likely to take, and imposes a large

amount of additional postprocessing work to boot.

[carbon_dragon]

Here is a shot from the Minolta Maxxum 500/8 on the 5D (1.5x crop) so effectively 750mm.

The picture and a piece from the top edge. Not too bad.<div></div>

[carbon_dragon]

And here's a shot of a house way up on the hillside with the same lens. On the digital the

contrast seems to be lower but it's not bad.<div></div>

[ShunCheung]

A couple more comments: (1) If you want to show sharpness, a small web image is completely meaningless. To demonstrate it, crop a samll but meaningful portion from your image to 511 pixels across and show us the actual pixels, e.g. the eyes of an animal. (2) Photography is not about showing one successful shot. There are plenty of people who argue that AF is not necessary for sports photography, and they promptly show 1 good sports image shot with manual focus, 1 image.

 

You can get 1, or a few, lucky shots occasionally. What we haven't seen are all the rejects. Without showing all the rejects, you are providing a misleading argument. Another thing with the Minolta is that it has an AF mirror lens; I don't know how well that AF works, but as a friend of mine puts it, AF would be God send for mirror lenses, because accurate focuing is the primary problem. That is a problem you will encounter for every shot you take. And unless you are getting an AF mirror, any Minolta result doesn't apply.

 

It is exactly right that 20 years ago, AF had just started. That was why there was still a market for mirror lenses back then. Today, most major brands don't bother to make them any more because people are accostumed to the higher standards from AF now.

 

The bottomline is that I think Alexander is completely convinced that a mirror lens is the way to go. The best way to find out whether that approach works for you is to try it yourself. I did 20 years ago.

[douglas_greenberg]

To continue the argument: :))

 

Shun alludes to a significant issue here when he rejects the validity of a web image to demonstrate "sharpness." I've long maintained that too many photographers make a near-fetish of "absolute sharpness" at the expense of handling characteristics, convenience, and yes, price, which is a real-world consideration, particularly for amateur photographers. Sure, I would like all of my lenses to be absolutely maximally sharp, but until I win the lottery, I often have to make some compromises.

 

The truth is, most of us are NOT using our cameras to make gigantic blowups for public presentation. And even if we were, the truth is that at the distance from which the average person views a poster-sized photo, slight differences in sharpness don't really make all that much difference.

 

The point is to "get the photo in the first place." I think that those of us who have posted images taken with mirror lenses have shown that you can indeed get some pretty good shots using these lenses. At the same time, for every one that is a keeper there are probably twenty or thirty that are tossed (this is similar to what happens with "digiscoping," another imperfect but surprisingly popular way that people on a budget can get into long-lens photography). But so what? Unlike film photography, digital photography makes it possible to delete those twenty or more crap shots at absolutely no cost!

 

Bottom line: for those of us who mostly use our photos to make fairly small enlargements and yes, for web presentation, a mirror lens can be a very useful tool that is well worth the modest cost. I say this as someone who owns a couple of "big glass" teles and knows that ultimately, the quality thus obtained is superior. But frankly, I often go out to take photos under circumstances where committing to a giganto-telephoto rig complete w/tripod just isn't practical. Sometimes the mirror lens is just the thing (or else my 80-400mm. VR zoom, but even that was pretty costly), and I know that if I hadn't brought along the smaller lens I might not have brought the camera out at all and I would have ended up with no photos at all.<div></div>

[philaret]

Well, first off, I am indeed completely in Douglas camp. And, not to sound offensive, Shun, I now know what kind of photographer you are. You made your choice 20 years ago. ;-) With the risk of being trivial, sharpness is not only something to be demonstrated 'on the pixel level'. I am happy you requested that. That identifies you with a certain crowd. Another point of view is that sharpness is not only a characteristic of a lens, it is also an illusion and sharpening algorithms use this feature. Much like special effects in movies use other illusions of the eye. We will go off topic if we start arguing that sharpening of a scanned film image does not do the same job as sharpening of a digital image. What I really wanted to demonstrate with the latest image is that the illusion of acceptable sharpness can be achieved with a mirror lens image. Probably because its rays hit the digital sensor at the right angle. Which makes it more usable than before. You cannot do this with an old expensive wide angle lens. For digital sensors these will have to be designed from scratch. And finally, I believe photography is about creating an image that is appealing to yourself, to other people and ultimately to a market. That is associated with a lot of rejects! All the time. Fortunately, as Douglas mentioned, we do not have to pay a huge price for them anymore.

[mark_chappell]

<I> And, not to sound offensive, Shun, I now know what kind of photographer you are. You

made your choice 20 years ago. ;-) With the risk of being trivial, sharpness is not only

something to be demonstrated 'on the pixel level'. I am happy you requested that. That

identifies you with a certain crowd. </i><P>

 

Hmm. So, did you check out Shun's portfolio before you offered up these condescending

statments about <B><I>"what kind of photographer"</i></b> he is? IMO, he's quite a

good one, and if he's in <B><I>a certain crowd</B></I> that gets fine results by avoiding

equipment with obvious optical warts, then I'm gonna apply for membership.

[philaret]

There is nothing in my post, Mark, that identifies Shun as a bad photographer. My statement is not 'condescending', but it is 'classifying' to the same extent he is trying to say I am convinced mirror lens is a way to go. Now, I recall Shun's portrait is obscured by a Contax 645. What about Zeiss Mirotars? "Carl Zeiss Mirotar T* 500mm f4.5, one of the fastest ultra-telephoto lenses in the world today. This lens design has totally eliminated chromatic aberrations so prevalent in conventional high-speed lenses of this class..." Granted, this lens is neither cheap nor small, but do you have doubts it is a quality optics? The last of them, the 500mm, I believe, was re-designed only 2 years ago. Any 'condescending' comments about Zeiss?

[ShunCheung]

Alexander, you are getting off topic. What does my portrait and Zeiss have to do with this discussion and debate? If you disagree with my opinions, that is fine. When you start making personal comments, it become a problem.

 

An image on the web may be 600 pixels across. At 300 DPI, that means a 2-inch or 5cm tiny print. That is why a typical web image cannot show the true sharpness and can easily hide a lot of problems. Cropping out the actual pixels is a simple way to demonistrate true sharpness at a level that a lot of us can easily understand.

 

If all you care about are tiny prints or web images, you indeed don't need a whole lot of quality. In that case I would say just use a 300mm/f4 or 400mm/f5.6 AF lens, and make the crop in PhotoShop afterwards. Obviously with that level of extreme cropping, the final result will be mediocre, but at least you don't need to struggle with focusing with a dim viewfinder in every picture you shoot.

[mark_chappell]

<I> Any 'condescending' comments about Zeiss?</i><P>

 

Sure. Happy to. If the Zeiss Mirotar is a mirror lens, it will produce out of focus highlights

that are, in my opinion, butt-ugly. And it doesn't have a diaphragm, so there is no control

over depth of field. That's just the nature of the reflex-optics beast.<P>

 

I think Shun is offering good advice, based on considerable experience. You have a different

opinion and are welcome to it -- but that says nothing about the value and pertinence of

Shun's viewpoint.

[jeff h.]

"...feisty bunch of guys wondering if a $150 500/8 lens will take as good a photo as a $1500

500/4 lens."

[mark_chappell]

<I>"...feisty bunch of guys wondering if a $150 500/8 lens will take as good a photo as a

$1500 500/4 lens."</i><P>

 

I sure wish the last part ($1500 500/4 lens) were true.

[philaret]

Shun, I mentioned Contax because this looks like your favorite system and because you said: "Today, most major brands don't bother to make them [mirror lenses] any more because people are accostumed to the higher standards from AF now." A hint that you may not be entirely accurate even for your favorite system. Tamron still makes an umptienth version of their mirror lens, which received only positive marks in this discussion. If you can find their 350/5.6, it is even better. My purpose was to collect opinions from people who actually shoot with mirror lenses, so that we could identify the 'lemons' and pinpoint the good ones. It is you who imply that I am trying to push my point of view by deceiving the public with appropriately scaled images, or worse, that my standards of quality are too low. When I scaled a 8 megapixel image to 1024 pixel on the long side, it is not enough for you to judge the sharpness! If you have time, go to http://www.photographyreview.com/, find reviews on Zeiss Mirotar 500/8.0 and wonder how different they are. Unfortunately, they are not about digital. -My- purpose was to find it about the performance of specific lenses with digital sensors. I think we succeeded in collecting a few benchmarks that people can use in the future. More would be very welcome. BTW, their prices, Jeff H, spread all the way from $50 to your 500/4 refractor price and do not exactly correlate with the quality. Finally, I do not find a 20D viewfinder image dim, even at F/9. But it is small and with a certain kind of vision problem this can be difficult. I sure miss the autofocus with the Tamron, but I am waiting for my chipped EOS to M42 adapter (thanks to Bob Atkins who pointed us to it) and hope to improve on what I get right now.

[pvp]

<B>Mirror Lenses are Crap</B><P>

<img src="http://home.comcast.net/~w7apd/public/heronbokeh.jpg"><P>

Any questions?

[jeff_drew4]

This has been interesting. I gave up trying to use my Meade 102 mirror lens for serious wildlife shots many years ago. It still works fine for visual spotting and occasionally as an astronomical telescope. If you want to try something new, make it an APO short-tube refractor. Yes they are fixed aperture, a bit longer and heavier than a mirror, but sharper and priced in the same range as CATS. Checkout your telescope retailers for options. In the long run, it always seems to be a compromise tho'. The really crispy-sharp APOs are expensive as are the OEM long lenses for cameras!

[jean_jacques_lemaire]

I have a Rubinar 500m F/5.6 mirror lens. Great on some aspects, bad on others.<BR>

Pros:<BR>

- compact, light<BR>

- very short minimum focus distance<BR>

- f/5.6<BR>

- quite nice focus ring (very precise)<BR>

- no chromatic abberations, no vignetting<BR>

- quality/price ratio<BR>

<BR>

Bads:<BR>

- sharpness not of a high level, but correct though. Mine was second-hand and in a bad shape when I got it. I had to realign the lens (<I>very</I> difficult job) to get good results. I don't know if brand new ones get this optimization from the beginning.<BR>

- f/5.6 but because of the partial reflexion of mirrors, effective aperture was around f/8.0<BR>

- my main complain: no aperture setting ring. When you shoot 3 meters away, depth of field is very very shallow (around 2-3mm). The background is completely blurred, *but* you can't have a bird completely in focus.<BR>

- for Canonists, I recommand to get a AF-Confirm chip to help manual focus: depth of field is so shallow (with this 500 f/5.6) that you need to be very careful at focus.<BR>

<BR>

If you are on budget, this lens is great. BUT not usable on any subject. It is a nice toy for bird photography IF you can be close to them and have a not-so-close background.<BR>

The donut-like bokeh is often a problem, but if you care, you can play with it and get good results. Not on all pictures, but you can.<BR>

[philaret]

It is not clear, Jean-Jacques, if your experience refers to using a Rubinar on a DSLR or on film, but obviously certain point are the same. And, so that Shun, Mark and others do not think I am an unconditional adept of mirror lenses, I should add that using a mirror lens at 'macro' distances is THE biggest problem IMO (3m is 'macro' for a 500mm lens!). Since stalking some insects is difficult with a typical 100mm macro (+1.4X TC), on a full-frame DSLR is conserned (or film), I once got a 300mm Rubinar and tried it at short distances. Well, the donuts are the worst obstruction in this range. While you can shoot a dragonfly at a considerable distance, even in those places where you cannot physically get closer, you get a really weird looking image wherein tiny dohnuts line up every bright area, while there is little or no background problem. However, getting small animals completely in focus at close distances can be a challenge for any lens, since it can be tough to choose a spot over which to align an autofocus sensor and choose the right depth of field. I did not use the other Rubinars except the 300mm, but they are better built than some of the plastic Japanese mirror lenses and sharper/contrastier than they are. However, Rubinars are a by-product of the Lytkarino optical factory making large telescopes and there is little hope they will be optimized for digital sensors, while pricier mirrors may well be optimized in the future. Finally, IMHO Alan's image would not look good with any lens stopped to the same F-stop. The background is bright and really distracting unless you have one of those 500 F/4s. But then at F/4 the whole bird may not be in focus ;-). A powerful flash may be in order to isolate the bird from the background. With the risk of being slammed again for low quality, here a 2-minute fix for Alan's image.

[mark_chappell]

<I>The background is bright and really distracting unless you have one of those 500 F/4s.

But then at F/4 the whole bird may not be in focus ;-)</i><P>

 

In which case, if you are using a standard telephoto, you can stop down.

[philaret]

This is an undisputable point, Mark. Slamming down mirror lenses should begin with "...but they do not have an iris", not with anything else previously mentioned. Some of them did, but as a general rule they do not have one. However, stopping down in Alan's shot, as I said, may not really salvage the situation. And, what if you got a lens with a 6-blade iris?

[pvp]

Alexander, I did make a version of the heron photo with the background selected and blurred. While that's a possibility with images intended for prints, it seems like a lot of extra effort when you could start with a lens that doesn't mangle the image before it even reaches the film.

 

Also, not very feasible when shooting transparencies that are intended for projection. (Yes, some of us still do.)

 

Mirror lenses are crap. They are a liability much more often than an asset.