Side By Side Lens Comparisons

R

robotsound

Well-known member
There is a lot of discussion here about the relative merits of various lenses. There are plenty of opinions, but not enough opportunities for direct comparison.

My hope is that this thread could be a place where people would post relatively scientific side by side comparisons of lenses, so that everyone can see exactly how big (or small) the differences are. I feel it's better to see it for yourself, rather than try to get a consensus of opinions. The problem with relying on opinions is that different people have wildly different needs and budgets, and it's good to see the differences first hand.

In an attempt to get things started, I have posted some screen captures from some A/B tests that I performed myself. I will offer them here for direct observation, and let you draw your own conclusions.

The contenders:

- Canon EF-S 18-55mm f/3.5-5.6 IS (aka the T2i kit lens)
- Canon EF 50mm f/2.5 Compact Macro Lens
- Canon EF 17-40mm f/4.0 L USM
- Canon EF 35-80mm f/4.0-5.6 (kit lens from an old 35mm film camera)

As much as is possible, I have tried to shoot the same angle, focal length, and aperture on multiple lenses here as much as possible. Obviously some of them have additional pros and cons, such as being faster, having image stabilization, etc. What I'm interested in here is comparing the optical quality differences between lenses when operating them under (nearly) identical conditions, to allow direct comparisons.

These are screen captures (command shift 3 on my 20" Cinema Display). You can't see the entire picture, but you can see most of it. These are screen captures of video files, because I am most interested in how these lenses perform in video mode. There are already a lot of good comparisons online of various lenses taking stills. If there is sufficient demand, I may re-export full sized stills from the original video files.

http://media.ratherlargerobot.com/dvxuser/lens-comparison-01/

Comments welcome. And if anyone else has performed similar comparisons, please post them here!
 
Last edited:
These are video centric tests, so all of them are at 1/50 of a second, which approximates a 180 degree shutter angle at 24 frames per second. When shooting video, you effectively can't change the shutter speed.

The general idea here, as I envision it, is that you pick two sample images that used the same ISO and aperture settings, and compare those two side by side. For example, you could compare the following two lenses and settings:

17-40mm L @ wide f4.0 iso800
vs.
18-55mm IS @ wide f4.0 iso800

And so on.

If you want just raw lens comparisons, there are numerous places on the internet that have done so for still images. The ISO settings were adjusted to keep the exposure more constant.
 
Last edited:
wow! that kit lens is sharp! even more than the 17-40L!! what a pity, that awful focus ring that it has...

(thanks a lot, nice comparison)
 
There's no movement in the frame so constant shutter speed is not required for your test.

The ISO will change the image quality, though.
 
Please refer to my previous post. The intended use of these tests is to compare two images from different lenses with the same ISO values, under real world conditions... e.g recording through the h.264 video codec at a shutter speed of 1/50. Thanks to these images, you can see how these lenses stack up against each other in a typical house using only the available light.

I'm not interested in lab results and MTF readings here so much as I am in real world differences.

If anyone else has similar lenses (e.g. an old manual Nikon and a newer Canon) that have overlapping apertures and focal lengths, it would be great to see more side by side comparisons!
 
Your test results are kind of confusing, though. Is your intent to compare picture quality in regards to video?

I don't really understand your test results organization, either. E.g., you say 35-80mm @ 50 f5 iso800, then 35-80mm @80mm f5.6 1600, and then just 35-80mm @ f4 iso800. What focal length was that last test?

Is the variable across these lenses supposed to be how they handle changes in focal length that force the aperture to stop down, thereby demanding higher ISO?

A suggestion might be to take your raw videos of these lenses in action and cut them together as a video, where every 5 seconds is another lens in action. The screen grabs you posted don't seem to be that different, or the differences are hard to find because we have to tab back and forth between the images.

robotsound said:
35-80mm @ 50mm f5.0 iso800

35-80mm @ 80mm f5.6 iso1600

35-80mm @ f4.0 iso800
Click to expand...
 
A suggestion might be to take your raw videos of these lenses in action and cut them together as a video, where every 5 seconds is another lens in action. The screen grabs you posted don't seem to be that different, or the differences are hard to find because we have to tab back and forth between the images.
Click to expand...

I agree. A quick cut or side by side will point out variations much more efficiently.

I do wish your tests had more of a pattern/consistency to it.
 
Wow.

OK, these were just tests I did myself one evening to compare the lenses that I had available to me. I thought it might be helpful to post them here for others, but now I'm sort of regretting it, as it seems to have just confused everyone.
 
don't regret it, it helped some people, or at least they helped me: I'm considering kit lens vs 17-40L as general-purpose and stills lens for my next body, to be bought in a couple of months, and now I know that, if I can somehow manage to stick a follow-focus gear to the kit lens (even if it has to be with glue) that one will serve me well; so thanks :)

and, about your "it would be great to see more side by side comparisons!", here are my One Dollar Resolution Chart results:
http://www.similaar.com/foto/lenstests/lenstestsa.html
(those are stills, not videos; I'll make a video-centric test when I get the new body)
 
NormanBates said:
don't regret it, it helped some people, or at least they helped me: I'm considering kit lens vs 17-40L as general-purpose and stills lens for my next body, to be bought in a couple of months, and now I know that, if I can somehow manage to stick a follow-focus gear to the kit lens (even if it has to be with glue) that one will serve me well; so thanks :)
Click to expand...

Don't be fooled by the video resolution. The 17-40l is a much better lens when it comes to stills or anything shooting higher then uprezed SD, which these DSLR's video more or less shoot. If you are just shooting video with the T2i, your right there really isn't much of a difference, but if you plan stills or video on a better camera you are going to see a change. :smile:
 
my plan is to use a not-very-expensive body (60D or 550D, maybe even 2000D if video is kind of ok) while I save for an eventual full-frame body that doesn't do line-skipping, hopefully 5Dmk3 in march 2011 or so

I could buy the kit lens for now, or I could skip that and go for the 17-40L; if I knew the 5Dmk3 would come out in march 2011 and be good enough for what I want, I'd get the 17-40L, but who knows, maybe the next generation won't be the good one yet, and even worse, maybe it's nikon that gets their act together and solves all these issues, in which case I'd regret having spent 6x the money on a not-as-future-proof-as-expected lens

I'll have to go to a shop and hold it in my hands, but right now kit lens + glued gear looks like the best option
 
Norman,

I bought the 17-40 thinking along the same lines as you (getting ready for the 5diii). This strategy has it's pros and cons.

Pro:
-Nice L series handling: makes it feel like a real camera
-Great image quality for still and video
-Fast,accurate, quiet AF for stills
-lens profiles for peripheral illumination correction and distortion in stills

Cons
-F4
-No IS
-Limited range (really need at least a 50 too)
-Limited use on full frame (great for certain shoots, but still need a 24-105 F4L IS, or better yet a 24-70 2.8L and 70-200 2.8L IS

So weigh carefully because for $300 more you can get a better lens now and get good use for up to a year before the 5diii. I am only saying this because I keep waffling back and forth between selling the 17-40 and waiting it out.
 
You must log in or register to reply here.
Back
Top