Canon FD to EOS Infinity Adapters: Are They Really That Bad?

[cooltouch]

You know, everybody says these FD-EOS adapters with the corrective lens are no good, and that they degrade an image. When I bought mine, I was aware of these assertions, but I figured it was worth the gamble. And on balance, I'm glad I made the gamble because it has paid off positively in some ways -- although not in others.

Just for edification, in case it might make a difference, the adapter I bought was one with glass made by Hoya. I was counting on the problems being less severe and maybe even non-existent because of the Hoya-made element. But as things turned out, only half of my hopes were met.

The main thing to remember with one of these lensed adapters is that they will produce flare if a lens is opened up past about f/4. So, if you were hoping to use your FD 50mm f/1.4 SSC wide open on your EOS, the image you'll get will be totally unusable. And PP will get rid of a lot of the blooming flare, but it won't get rid of all of it.

So, how does it work if you use a slower lens, or stop a faster lens's aperture to f/4 and beyond. Well, this is where it gets interesting. I have been able to detect no image degradation. None. And I've got some images to prove it now.

This first set of images I took with my Tamron 90mm macro with a Nikon Adaptall-2 mount on the lens. I then used a Nikon-EOS adapter to mount the lens onto my DSLR. So there was no glass between the lens and the camera's sensor. The second set of photos was taken with a Canon FD mount on the Tamron and the FD-EOS adapter was installed with the corrective element.

The following images were resized only. No post-raw production was done on any of them. I did do some brightness and contrast adjustment to the raw images before they were translated over.

Tamron 90mm with Nikon to EOS adapter:


The same image, but full-size:
http://michaelmcbroom.com/images/lenstests/rose_tamron90_nik-eos.jpg

Tamron 90mm with FD to EOS adapter:


The same image, but full size:
http://michaelmcbroom.com/images/lenstests/rose_tamron90_fd-eos.jpg

While I was taking photos with the FD-EOS adapter, an ant walked through the scene. Captured a few of it as well. I thought these would be good to show too because they do show what sort of sharpness the FD-EOS adapter is capable of. These are 100% crops of the original images.





So, some of you may argue that the above photo of the flower with the Nikon to EOS adapter is sharper, but I would recommend that you don't compare the same areas, but different ones with the FD-EOS adapter image to see equivalent sharpness. It was windy today and I was focusing where I could, when I could, plus the shutter speeds I was working with were not conducive to freezing the action all that well. Also, by the time I took the photos using the FD-EOS adapter, I was losing light and I had to use the camera's flash -- so the color balance and contrast are somewhat different.

I've got to go run an errand now, but when I get back I want to post some images I took with my (new to me) Canon FD 400mm f/4.5. With and without the corrective lens in place, and to see if there are any significant differences.

[ChromaticAberration]

This is indeed an interesting test since there are so many FD/FL mount lenses around. I am interested in reading more experienced photographers opinion but to me the examples above show a potentially good lens being transformed in an OK lens.

[koji]

The fast lens does not work well with these FD->EF with correction lens
adapter, more precisely the lens with a large exit pupil wont work well.

But if your lens is slow, it may work well beyond an ordinary lens
if the original one is superb of course. FD 35/2 (the first version) with
a concave front element + rare earth element + FD->EF adapter with a lens
(cheap adapter around $35).

mostly at f8, camera was EOS 5D, C1(Capture One), hand-held

[koji]

I did experiment with FD 55/1.2, the result was a disaster. The central
part of the image has whitish circle. With 135/2.5 (rather fast tele),
the image was also not so good. So I stopped doing it.

[Attila]

Closer distance in my experience ok with correction glass , try it at near infinity or longer distance which is harder for most lenses. Result will be dramatically worst with cheap correction glasses.

[koji]

Probably so, but I took the correction lens away and threw it.
My FD->EF adapter became a ext. ring, so no more testing.

[cooltouch]

Thanks for contributing some more examples, Koji. It was after corresponding with you that I decided to go ahead and gamble on getting one of these adapters. Like you, I've been disappointed also, but I'm learning how to put mine to worthwhile use, I hope.

The tests with my Canon 400mm I posted yesterday of that water kiosk at 400 meters -- those were done with the FD-EOS adapter, and, while I think the contrast was a bit soft, resolution was good. Maybe my adapter is one of the better ones because of the Hoya element, but it still isn't as good as I would like.

As for the other comments from Koji and CA -- I don't know from exit pupils. But doesn't the size of the aperture opening determine the exit pupil size? Anyway, your coments got me to thinking, so I grabbed a couple of fast lenses, and in the waning light, went outside and took a few more shots.

The lenses were an FD 50mm f/1.4 SSC and an 85mm f/1.2 SSC AL.

I was getting tired of shooting flowers. Besides, flower shots are always iffy unless you can dial in a lot of depth of field. So I looked around and settled on the BMW roundelle on my motorcycle. I took four shots with each lens hooked up to the FD-EOS adapter, and the corrective element was in place. Each shot was taken at wide open, then f/4, f/8, and f/16. I used a flash, which helped knock down some of the flare.

The 50mm f/1.4 at a 100% crop:

Wide Open


At f/4


At f/8


At f/16


Now for the 85mm f/1.2 SSC AL, reduced to maybe 70% or so:

Wide Open


At f/4


At f/8


At f/16


I reduced the size of the 85mm's images somewhat so they would fit within the 1024 pixel limit here at the forum.

If you look closely at the 50mm's wide open shot, you can see the flare along the edges of the circle and the letters, etc. By f/4, things had sharpened up a lot. I don't detect a lot of difference between f/4, f/8, or f/16.

With the 85mm, the wide open shot is really blurry because the contrast was so low. I just couldn't tell at what point focus was the sharpest. But by f/4, things had improved remarkably. And, same as the 50mm, I don't see a lot of difference between f//4, f/8, or f/16.

Okay, so these shots may not be macro quality, but I do believe they indicate pretty clearly that once the lens is stopped down by a certain minimal level, it's possible to record good detail. Even with lenses that have large rear elements, such as the FD 50/1.4 and 85/1.2.

If I woudn't have run out of daylight, I would have tried some distance shots, especially after reading Attila's comments. Maybe I'll add some more to this tomorrow.

I was gonna add some more shots I took with my 400mm f/4.5, but now I don't know which are which -- shot with and without the glass, that is. Because it focuses well past infinity, I found that its focus would reach all the way out to 20 meters. Wow, that's a big difference, compared to my 50s and 85 and others that won't focus much past about 20 or 30 centimeters without the corrective element. So, really, this ends up being quite useful for a variety of situations, such as small animal photography, more flower pics, candids, etc.

So anyway, I'll add some pics from my 400mm tomorrow. I suspect its "exit pupil" is small enough where it might not be very easy to tell the difference between shots with and without the glass.

[cooltouch]

Okay, I wanted to do another test with that adapter and a long lens -- again the Canon FD 400mm f/4.5. It doesn't have a large exit pupil, I'm presuming.

This is what I did. I'm tired of shooting flowers -- besides, there are only a couple in our garden that are worth photographing right now, and I'm tired of looking at them. So I chose my bike again. This time I focused the lens on a side plate on the carburetor. It says "Bing" and has a numeral stamped into the plate just beneath Bing. I shot photos at lens settings of wide-open, f/8, f/16, and f/22.

I've combined the photos taken with and without the corrective lens in place so you can see side-by-side comparisons. You'll note that the images on the left in these combinations show slightly more magnification than the images on the right. Both are 100% crops, but the ones on the left are showing the slight teleconverter effect the corrective lens in this adapter has.

To shoot the photos, I mounted the lens/camera combination onto a sturdy tripod and used Live View for focusing accuracy and to pre-release the mirror. ISO was set to 100 to eliminate as much noise as possible. There was no post processing of the images, other than cropping or reduction for display here.

Here's a full frame image of the subject. The bike was about 10 meters away from the camera.


The following 100% crop comparisons are self-explanatory.

@ f/4.5


@ f8


@ f/16


@ f/22


You see that reddish area that begins to become evident in the image with the corrective lens at f/16? It's more pronounced at f/22, and in fact it reveals an arch shaped perimeter. Well, it's actually a circle that is occurring at the middle of the image, and it's a rather pronounced bit of circular CA caused by that corrective element. I'm glad it doesn't begin to manifest itself until f/16.

I believe these four sets of images prove my assertion that the corrective element does not degrade the image in any appreciable way with at least some lens designs. I cannot detect any significant reduction of resolution, except maybe at f/22 -- but I'll probably never use this lens beyond f/16 anyway. There's quite a bit of sharpness falloff that is becoming evident by f/16, and is very evident by f/22. There is also a very minor softening in contrast that begins to become evident by f/8, but a light touch of PP will take care of this.

I've taken note about what has been said about lenses with large exit pupils not working well with an adapter like this, and I've also done a bit of reading on the subject. The exit pupil size of a lens is determined by the size of the lens's aperture opening. So this may explain why a fast lens exhibits so much flare until it's been stopped down to about f/4. By that point, the exit pupil is small enough such that the smaller size of the corrective lens is no longer having a deleterious effect.