35mm L lenses on APS-C sensor cameras?

[Daniel Browning]

My testing method was to focus on an object at a reasonable distance (50 feet or so) and take a picture with the 18-200mm at 200mm and the 70-200mm L at 200mm, then overlap them in Photoshop and estimate the relative cropping. It was significant. Bryan notes the same phenomenon in his review of the 18-200mm, and the pics of the starfish seem no different at 170mm vs. 200mm.

It's probably just two additional factors: 1) the difference between x feet and infinity, and 2) the difference in the center of perspective. If you setup a tripod and just change lenses, then you may not realize that you're also changing the location where light enters the lens-- the entrance pupil. For example, imagine you have a 50mm lens that is 5 feet long, but through the magic of optics gives you an angle of view that is truly 50mm. Then you have a 50mm lens that is only 5 inches long. If you setup the tripod and compare them by just swapping lenses, the 5-foot-long 50mm lens will give a different field of view. But if you instead align them based on the center of perspective, which is the entrance pupil (or virtual image of the lens' aperture), for both lenses, then compare them, they would yield the exact same field of view (and angle of view).

[ChadS]

I tested one a while back, and it was more like ~160mm at the long end when compared to the same shot taken with the 70-200mm f/2.8L IS II.

I'm curious to the testing method as well. Stars are the only target I can think of that one could use as an infinite-distance source. I've used systems in the lab that are functionally 25% less magnification than their quoted values at "real" distances but if we test them with a collimated source they generally are right where the manufacturer says they are. Though I would not be shocked if these lenses (or at least the one tested) did not turn out to perform as spec'd. That happens all the time in the lab as well...

[HDNitehawk]

Hi!


If I attach for example Canon EF 70-200mm f/4.0 L IS USM lens to my Canon 600D, do I get at maximum zoom the same results as wit Canon EF-S 18-200mm f/3.5-5.6 IS Lens? Or I get with EF 70-200 L lens 200x1.6=320mm which is more zoom than EF-S 18-200 lens?

Thanks in advance!

No, they will be the same. All of the 1.6 hype is talking about camera bodies and sensors. Unless you plan on buying a full frame camera all of the Crop to FF comparisons are really useless information for you.

There is one thing you will get out of the 70-200mm L. Better IQ....

[Daniel Browning]

I found several sites with images and lots of information, but did not get a clear answer.

It may help to compare camera format with digital zoom. Let's say you start with a 5D2 and 200mm lens, and want to compare various methods of "zooming in":

1. Optical zoom: switch to a 320mm lens.
2. Digital zoom: use the 200mm lens, but crop the 21 MP image down to 8MP.
3. Digital zoom: use the 200mm lens on a Canon 30D (8 MP).
4. Digital zoom with high resolution: use the 200mm lens on a Canon 60D (18MP).

Options 2 and 3 yield nearly identical resolution, even though they are different cameras, because they have the same pixel size, pixel count, and focal length. But option 4 is going to have superior resolution as long as diminishing returns don't cancel out the benefit of smaller pixels.

[neuroanatomist]

Maybe Canon measures focal length in 'man-millimeters'?

[ChadS]

Maybe Canon measures focal length in 'man-millimeters'?

Can you shoot some stars tonight with this lens and a 70-200 (and a 200 prime if you have one) all at max focal length? I suspect the problem Bryan saw was in the combination of zoom and focus. Up close the optical changes required to focus the light may have masked the change in focal length. And, unfortunately, Canon's definition of "close" might be half a mile.

Remember, the effective focal length of a lens actually only applies to paraxial (near the optical axis) rays of an infinite source. That's not to say that's how a lens will behave at finite conjugates nor anywhere but on the optical center (just read your profile so I'm sure you're well aware of this fact - microscopy is littered with marketing half-truths - same as many diagnostics fields).

[neuroanatomist]

Can you shoot some stars tonight with this lens and a 70-200 (and a 200 prime if you have one) all at max focal length?

Sorry...I don't own an EF-S 18-200mm, I just tried one out in a store out of curiosity (I brought my 7D and 70-200mm f/2.8L IS II with me).

[Rubar]

Hi again!
To be honest, I did not expect so many replies. Thank you all for your input. As I see there are many different opinions and excitement, but still no clear answer. At least I still don’t get it. I shoot different objects with Canon EOS 600D at about 100-300 meters (328-980 feet). With EF-S 18-200mm f/3.5-5.6 IS lens I get quite satisfying results for my purpose. Only thing I’m missing is a little extra zoom. While planning to get a new lens I consider getting a higher quality one.

I think to give a clear answer to my question a real-life test is needed. Unfortunately I don’t have an opportunity to borrow such lenses from anyone and our stores are not going to allow me to do this either. Maybe someone has those lenses or can borrow them and test them in a distance.

So the test would have a camera with APS-C sized camera. For this test I believe there won’t be a big difference which model it is. Take 3 photos at a distance of approximately 300 meters (980 feet), but definitely all there at the same distance with the same camera:
1) one photo with EF-S 18-200mm f/3.5-5.6 IS Lens (or similar EF-S lens) at 200mm
2) one photo with EF 70-200mm f/4.0 L IS USM Lens (or similar EF lens) at 200mm
3) and one with EF 200mm f/2.8 L II USM Lens (or similar EF lens) which has fixed 200mm
4) unfortunately there is no EF-S lens with fixed 200mm, that I would know of.

Then if those photos are uploaded to this forum, we all can compare them and see the difference. Of course if such a test will take place then it could be expanded to more different focal lengths, object distances and different lenses. What you think? Would it make few things clear?

Anxiously waiting for results, new thoughts and ideas

[thekingb]

Hi again!


So the test would have a camera with APS-C sized camera. For this test I believe there won’t be a big difference which model it is. Take 3 photos at a distance of approximately 300 meters (980 feet), but definitely all there at the same distance with the same camera:
1) one photo with EF-S 18-200mm f/3.5-5.6 IS Lens (or similar EF-S lens) at 200mm
2) one photo with EF 70-200mm f/4.0 L IS USM Lens (or similar EF lens) at 200mm
3) and one with EF 200mm f/2.8 L II USM Lens (or similar EF lens) which has fixed 200mm
4) unfortunately there is no EF-S lens with fixed 200mm, that I would know of.

If someone did this test, I suspect the results would be as follows. Each of the 3 shots would have a similar (or per Neuro's testing, maybe not so similar) framing of the subject at 200mm. Take 3 different lenses on the same camera at 200mm, and you will get 3 (at least) slightly different framings because of manufacturing variations. The focal length stamped on the camera is an approximation. But the variation shouldn't be major. And if you are looking for more reach, you need a lens with a longer focal length. It's not easy to capture subjects 300 meters away unless the subject is quite large.

Neither the 70-200 or the 200 prime will get you significantly more reach. For that you need a 70-300, 100-400 or a super tele prime lens.

For what it's worth, the hypothetical test would also yield quality differences....at 200mm, I suspect the 18-200 would have the "worst" IQ of the bunch, with the 70-200 and the 200 prime duking it out for the win, with the prime edging the zoom. Then, the f-stop selected for the test plays a role too -- not in how the photo is framed, but in terms of depth of field. The 18-200 at 200mm will have a max aperture of f5.6 (more depth of field); the zoom at f4 (less depth of field); and the prime at f/2.8 (even less depth of field). Depending on the focal point of the shot, the results from the three lenses could look quite different due the aperture setting. If all shots in the test were taken at f5.6 (the least common f-stop), the depth of field should be the essentially same in all three shots -- and I would suspect the prime would be an even easier pick for the IQ winner in that case.

[HDNitehawk]

I agree with thekingb, if what you want is more reach get a longer lens than 200mm. If you really get a tighter framing from the L lenses, it isn't going to be enough that you really notice in real world situations.

You are going to have a hard time finding someone that has all three of those lenses. Most people with those two L lenses in their kit have no use for the 18-200mm, unless they just upgraded.