fredmiranda.com
Login

Moderated by: Fred Miranda
  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

       2       end
  

Archive 2004 · TESTED: 70-300DO vs 70-200/f4&f2.8

  
 
amirm
Offline
• • •
Upload & Sell: Off
p.1 #1 · TESTED: 70-300DO vs 70-200/f4&f2.8


Here are the results of my second run at testing the new Canon 70-300 DO zoom against the standards: 70-200f4 and 70-200f2.8. The first attempt was marred by the poor autofocus accuracy on these lenses.

This time, the testing was much more rigorous and defensible and as such, I have included sample images for you all to see. My data now generally matches the theoretical MTF charts from Canon, adding more confidence. Original RAW files are available if anyone wants to host them.

Test conditions: Resolution chart taped to the wall was used as a target. The camera was mounted on Gitzo 1348 CF tripod, raised waist high. Remote capture was used with mirror lock up. IS was off.

Source image: Composite of resolution patterns from Norman Koren and PDI color test image (which has a nice set of resolution bars in it). I used two different sizes of the PDI image to make it easier to see what is going on. The smaller was placed close to the center and the focus was optimized for that.

This set of tests is for 110mm focal length to cover the wide end. I suspect the restuls for 200mm will be similar. If there is sufficient interest, and I have another weekend to kill, I will test that too .

Exposure settings: ISO 200, F5.6, 1/80sec. Camera 1Ds. Raw capture with conversion in Photoshop CS and Canon SW.

Focus control: As I mentioned, using AF with center point was totally unreliable with focus errors in excess of 1.5%, which basically invalidates the true rankings. This time, I used multi-point AF as a starting point. This worked better since the camera would find the high contrast parts of the sample image and focus on those. Since the target is flat, the error here is minimal. I then used manual adjustments by using a focusing rail, and slightly moving the camera forward and back. The images were captured and analyzed at 200% on a 21” monitor, making it easier to figure out if they were in focus. The work as you can imagine, was quite tedious. I spent about 8 to 10 hours generating the results that you see here!

Now for the Results. I explain them since the crops below may be hard to see.

Center Resolution: Tie between F2.8 and F4. Both resolved 4.5 vertical bars in the smaller PDI image. The DO finished clear step behind resolving about 3.5 lines in the center.

Center Contrast: F2.8 barely out the F4. DO is very poor in comparison. It is mostly a gray mush compared to the other two.

Corner Resolution: F4 finished first, resolving 6.5 (larger) bars in the corner. F2.8 finished second with a score of 6 bars. DO resolved nearly 6 bars so a close second to F2.8.

Corner Contrast: No contest. F4 wins by a landslide. The bars are nearly solid black. F2.8 is next in line but a step down. The DO is again poor compared to these.

Chromatic Aberrations: As reported before, the DO does superbly here. There is just a hint of blue bleeding. Both L lenses suffer from both red and blue CA which is visible in the high frequency sections of the resolution chart. These are mistakenly assumed to be moiré distortion caused by the Camera having too weak of an anti-aliasing filter. It is not. The additional blue-red lines beat against each other and black lines in the image causing the patterns you see here.

Color Balance: The added contrast on the F4 makes the colors “pop” better than the other two. This matches my first experience with this lens where I found a level of color saturation I had not seen with the F2.8. The DO again does poorly here, finishing behind F2.8. The difference between the F4 and DO is fairly evident (to my eyes anyway, looking at full size RAW conversions). Sorry, I don’t have samples for you to look at (have to go back to my day job ).

Summary: Despite the good showing on CA, the DO lens as expected, is not in the same class as the L lenses. The lightweight F4 gives strong competition to the DO lens, at half the price and nearly the same weight. It even outperforms its bigger brother in the corners. If you want to travel light or highest image quality without IS, the F4 is the ticket. If you want to save some space at the same time, then you have to put up with the deficiencies of the DO.

The F2.8 has the best balance of all very good image quality and IS with a wider aperture. If you can put up with the weight, you can’t go wrong with this lens.

[UPDATE] I tested the 280mm setting also. Please see later in the thread for the results.


Edited by amirm on Jun 05, 2004 at 08:07 PM GMT



Jun 01, 2004 at 07:10 PM
amirm
Offline
• • •
Upload & Sell: Off
p.1 #2 · TESTED: 70-300DO vs 70-200/f4&f2.8


And here is the test image as captured by the 70-200f4 and resized for the web (i.e. you can't tell anything about resolution/contrast from it). Note that the lighting was uneven (three seperate lights). So please don't make any judgement on that poor aspect of the setup. It did not affect the kind of testing that I did (which was all relative to each lens).





Jun 01, 2004 at 07:11 PM
amirm
Offline
• • •
Upload & Sell: Off
p.1 #3 · TESTED: 70-300DO vs 70-200/f4&f2.8


Here is the 70-200f2.8 output. Again, not very useful for anything but color...



Jun 01, 2004 at 07:16 PM
amirm
Offline
• • •
Upload & Sell: Off
p.1 #4 · TESTED: 70-300DO vs 70-200/f4&f2.8


Last but not least, the 70-300Do.


Jun 01, 2004 at 07:17 PM
DaShiv
Offline
• • •
Upload & Sell: Off
p.1 #5 · TESTED: 70-300DO vs 70-200/f4&f2.8


Looks very thorough! If you have time to repeat the tests at different apertures and focal lengths, I'd be interested to see it.

I just got my copy today. First impressions: build seems solid enough and it has the L "splattered" finish, but the zoom extension is definitely more flimsy than that of the 24-70 (the only extending L-lens I've shot with). I've found the zoom to be stiff only when zooming "against" the lens's weight--if I point the camera down slightly when zooming out and up slightly when zooming in (i.e. using zoom creep to help), it's smooth enough. It's very stiff when fighting the lens weight though--might not be a good lens for shooting birds in flight, since that requires zooming out with the lens pointed up. It'll take some getting used to, that's for sure. For a telephoto lens, the DO is wonderfully small when the zoom is locked at 70mm with the hood reversed, and fits very nicely into my backpack--but it's still less discreet than I'd like when shooting fully extended with the hood attached. Still, for a 300mm lens it's not bad at all. It's still very obviously a telephoto lens while shooting though, to any casual observer.

The IS is every bit as good as the one on my 70-200 IS, which means it's great. Got some 1/60 shots at 300mm no problem, and I'm sure I can go even slower if needed (with some luck).

Image quality: the DO doesn't quite have the "pop" of my 70-200/2.8L IS, as I'd expected. Corners seemed okay to me, but then again I'm using a 1.6x camera--which camera did you shoot your test with, Amir? I shot a blank wall wide-open and there seemed to be a hint of vignetting wide open at the long end, but I'm not sure. The DO lens degrades a bit wide-open and doesn't give me the same confidence shooting that way like my L lenses do. There's a small amount of softness full-tele at 100% pixel size--this is definitely not a substitute for those L tele-primes that one would be happy to post 100% crops from. Overall, image quality is acceptable (so far) but won't be knocking people's socks off, I'd imagine. I think anyone buying the lens solely for its performance might be disappointed.

I'm planning to shoot some candids this weekend--I'm curious to see whether this DO lens will give me donut bokeh on spectacular highlights, how it holds up under some real world shooting (backlit, high contrast, etc), and to get a better idea of how sharp this thing is.

I'm already glad I that I still have the 70-200 IS in my bag for more "serious" shooting. As it is right now though, the three lenses tested above each offers something a little different: the 2.8 IS offers maximum versatility, the 4.0 has the best price/performance, and the DO for maxium compactness. After shooting a bit with the DO though, I agree with those who say that the DO is overpriced for its performance--compared to the price, it's nowhere close to 2x the quality of the 70-200/4 (although it does have IS, which is very helpful at 300mm). I wouldn't recommend this lens for anyone who's very price conscious. But it's usable, and it's small. The jury is still out for me as to whether it's "good enough" to justify keeping it. The 70-200 zooms are simply very, very good.

Because of zoom weight/stiffness and possible sharpness issues, I think this is more of a "people lens" than a "wildlife lens."

Did I mention how nicely the DO lens fits into my backpack though? Okay, I'll mention it again.



Jun 01, 2004 at 10:17 PM
amirm
Offline
• • •
Upload & Sell: Off
p.1 #6 · TESTED: 70-300DO vs 70-200/f4&f2.8


I used the 1Ds for testing because it has the highest resolution and tests the entire lens just in case someone tries to use it on that body or the 1DMKII.

I will probably test the other focal lengths next weekend. You may also want to look at the older thread that I started on this topic. While those results are not as reliable as these here, they are good data points. If the 1.4X results repeat, they really put the F4 in good light. In that thread, I also documented the same issues you have with the zoom ring. I also pointed out that IS works well but it is a bit noisier than my other L zooms (so is AF).

Like you, I think everyone of these "duplicate" lenses has a place. I ordered this one before I got the F4. But then I got impatient while on a trip and bought the F4 too and haven enjoyed that lens immensely. I will do some "normal" shooting this weekend too with the DO and see if the performance is good enough for that. I supsect for candids and such, it should be nice. And coupled with a 10D, it knocks the socks of any of the 8 megapixel P&S cameras, even if it doesn't come close to the L performance.




Jun 01, 2004 at 10:34 PM
AJSJones
Offline
• • • •
Upload & Sell: Off
p.1 #7 · TESTED: 70-300DO vs 70-200/f4&f2.8


Belated thanks, Amir, on the testing. MR's review is now up at Luminous Landscape and could well be saying the same thing. Your pictures help me understand (a bit better) what BxU's are in the DxO system. I did a "blur more" in PS (i.e. added 1 BxU) on your resolution image and switched back and forth to compare that effect with the comparison between the lenses and it's actually quite modest (didn't try 2 and 3 BxU's). I also tried some local contrast enhancement (USM 20/40/0) but couldn't bring the DO lens close to the 70-200's. Having just borrowed a friend's 70-200 f/4, the weight is nice and light and I can live with the length. Your test seems to confirm my initial casual impressions (and all the glowing reports about the f/4 here and elsewhere) that there is a non-trivial hit in quality in exchange for the extra 100mm, compactness and IS and $700ish delta in cost. I'm leaning heavily to the f/4 and will have to test out my handholding skills to see how important the IS is for this focal length range.

Andy



Jun 03, 2004 at 06:05 PM
amirm
Offline
• • •
Upload & Sell: Off
p.1 #8 · TESTED: 70-300DO vs 70-200/f4&f2.8


Thanks Andy. I was relieved to see MR's review comments match mine. It is a shame he did not do the identical test agains the L zooms and publish that. We would have had more precise figures to go by than my testing.

Alas, I have not seen much interest in this thread. I was going to do more objective testing with other "duplicate" lenses (and with TCs on these lenses) but given how much work it is, I probably put that project aside and go shoot some pictures .



Jun 05, 2004 at 11:38 AM
skimmel
Offline
• •
Upload & Sell: Off
p.1 #9 · TESTED: 70-300DO vs 70-200/f4&f2.8



Well, I'm only one person, but I am *very* interested in this thread. I suspect that if you took a poll of who'd like to see the 70-300 tested as rigorously as you're doing, you'd get a lot of interest.



Jun 05, 2004 at 01:15 PM
MikeBinOKlahom
Offline
• • •
Upload & Sell: Off
p.1 #10 · TESTED: 70-300DO vs 70-200/f4&f2.8


Amir, I for one am very interested in your results, and (though I'm not absolutely positive at this point) they may have been important for me. With all the chatter about the lens, I'd been considering getting the DO lens to use travelling light with my 1Ds (despite a nagging feeling I'd not be happy with it). Thanks to what I see here, and from MR, I am probably going to go back to my original plan of getting a 70-200 f/4.

Sorry you aren't getting more applause on this, but thanks for your efforts! I'd still be happy to see results at 300mm if you have the time.



Jun 05, 2004 at 02:07 PM
MikeBinOKlahom
Offline
• • •
Upload & Sell: Off
p.1 #11 · TESTED: 70-300DO vs 70-200/f4&f2.8


One more thing--If you look at the number of views on this thread, I'd say interest is pretty high. Apparently enough people considered your work comprehensive that they didn't feel a need to ask additional questions, or even to make snide remarks!

So please don't feel like your painstaking effort was unappreciated.



Jun 05, 2004 at 03:00 PM
eric21
Offline

Upload & Sell: Off
p.1 #12 · TESTED: 70-300DO vs 70-200/f4&f2.8


It's a great test.
I've read every test on the 70-300 so far, and when it's available
here in the Netherlands i'll buy one. So i guess i'm one of those guy's that forgot to say "thank you".
Hope it's not to late : THANK YOU!!!!!!!!!!!!!!



Jun 05, 2004 at 05:51 PM
amirm
Offline
• • •
Upload & Sell: Off
p.1 #13 · TESTED: 70-300DO vs 70-200/f4&f2.8


Thanks again everyone. I really appreciate it. Didn't mean to ask for "thank you" notes but just couldn't tell how folks felt about the results.

Here are some more data at the far end, namely 280mm. I used my Canon 1.4X TC on the F2.8 and F4 at maximum zoom. Strangely, the DO at its 300mm setting still had slightly smaller image. So you get a bit more reach with the L setup. The exposure settings was the same as the wide end.

Results:

Center Resolution: Tie between F2.8 and F4. DO behind the two. The number of bars visible were, F4 (6.5), F2.8 (6.5), DO (6)

Center Contrast: F2.8 clearly the best. F4 one step behind and DO dead last.

Corner Resolution: F4 slight better than F2.8, DO half a notch behind. The number of bars visible were: F4 (7.5), F2.8 (7.1), and DO (7).

Note that the corner bars are larger in the chart to begin with so these numbers are not comparable to the numbers for the center (the corner PDI image was an 8x10 and the center was 4X6 so I suppose you could multiply these numbers by .6 to get the equiv results).

Corner Contrast: F2.8 & F4 tie. DO far behind.

CA: DO first by far, followed by F2.8 and F4. The L lenses suffered from a lot of red bleed. The DO had a slight blue bleed (which probably made its image look even more cold).

Overal Image (color fidelity/"pop"): F2.8 first driven by sharper images with better contrast/color, followed by F4 and then DO. The images from the DO are muted, soft and colder.

So the results here between DO and L lenses is pretty much the same as the wide end despite the huge distadvantage of the 1.4X on these zooms. I should note that CA was quite high on the Ls, especially in the red channel. While this may not lead me to use the DO, it does give me pause to use the TC on these zooms. They do not make good long telephots.

Here is a 100% crop of the winner, the F2.8. It is from the center region. Converted in PS CS, with a 1.0 exposure increase but no other processing.




Jun 05, 2004 at 07:02 PM
amirm
Offline
• • •
Upload & Sell: Off
p.1 #14 · TESTED: 70-300DO vs 70-200/f4&f2.8


Here is the second place finisher, F4. Again, +1 exposure compensation, center 100% crop. JPEG-10 compression (same as others).



Jun 05, 2004 at 07:04 PM
amirm
Offline
• • •
Upload & Sell: Off
p.1 #15 · TESTED: 70-300DO vs 70-200/f4&f2.8


Last but maybe least , is the DO center 100% crop with +1 exposure comp.



Jun 05, 2004 at 07:04 PM
Kyle Yates
Offline
• • • • •
Upload & Sell: Off
p.1 #16 · TESTED: 70-300DO vs 70-200/f4&f2.8


Thanks for the tests -- these are always interesting --especially when you don't have the time / inclination / equipment to do the tests at home yourself.

I certainly am finding the results with the DO interesting since at one time I was considering a DO prime -- guess for now I'll keep the 300 2.8 --still a magnificent lens by any stretch.

I have never had any complaints whatsoever with my 70-200 2.8 and really don't find hand holding this a problem especially if you leave the tripod collar mounted on it.

If you don't need the extra stop (which I do use quite a lot) or don't want to shoot with the narrower DOF @ 2.8 then the F4 looks indeed a winner.

Anyway still shows how good the 70-200 in either version is.

I'm a bit disappointed -- and puzzled by the lack of performance so far in the DO technology -- I really was hoping that this would deliver lighter and smaller lenses at traditional 'L' quality -- especially at the price of DO.

Well I'm still dreaming on.



Jun 06, 2004 at 02:26 AM
subsen
Offline
• • •
Upload & Sell: Off
p.1 #17 · TESTED: 70-300DO vs 70-200/f4&f2.8


Precisely ! I don't have anything to add, but I appreciate the work and was very interested.
Sub



Jun 06, 2004 at 03:03 AM
FredericB
Offline
• •
Upload & Sell: Off
p.1 #18 · TESTED: 70-300DO vs 70-200/f4&f2.8


Thank you very much for the test - it finished convincing me that the 70-300DO will not meet my needs (even though I would love a lighter lens than my 100-400 since I do most of my shooting while hiking in the rockies).

I do not know if you've had a chance to do similar tests on the 400 DO vs 500 F:4 but I'd love to see it if you have for I am in the hunt for a long telephoto (not for the hikes this time!)



Jun 06, 2004 at 08:01 AM
amirm
Offline
• • •
Upload & Sell: Off
p.1 #19 · TESTED: 70-300DO vs 70-200/f4&f2.8


I own the 400DO and did similar tests last year which led me to buying the 500F4 (the 300f2.8, 400f2.8, 500f4 and 600f4 beat it). Since then, I have not used the 400. However, those tests were not as riggerous as these. I will repeat them and report. The only problem though is that it won't be as good a comparison since the two are not the same focal length. I will give it a shot anyway and see where it takes us.

Just in case you are wondering why I didn't learn my lesson about the DO the first time , it is what SubSen said. I really needed something lighter with more reach on business trips. For example, when I was in Japan on business, it was the Cherry blossom festival (amazing experience if you haven't been there). Due to weight, I left the 70-200f2.8 behind and took the 135 with the 1.4X. Then I got there and saw a million beautiful birds in those trees and the 135 combo did not even come close to having the reach. So I ordered the 70-300 and while waiting, I went to Japan again without the F2.8. Frustrated, I bought the 70-200f4 while there (and paid more than I would here!). Was super happy with that choice and then the DO showed up a week later.

I am still going to take the DO out for a real-life ride and see what that does. I suspect it will still take nice pictures, but I will always know there is a better choice!



Jun 06, 2004 at 10:04 AM
amirm
Offline
• • •
Upload & Sell: Off
p.1 #20 · TESTED: 70-300DO vs 70-200/f4&f2.8


One more note regarding the 70-200f4. It seems to have what we call "High-Q" in engineering circles. Its contrast shoots up rapidly over a narror focus region and drops outside of that. Imagine a tall narrow bellshape curve (with vertical being image quality and horizontal focus position) compared to wider one for 70-200f2.8 and even wider with DO. This means that focus accuracy is key with the F4. This is what invalidated my first set of tests.

Given the gross auto focusing errors that I see at least with these low light situations, you may not see the kind of performance that I report. Outdoor though, it should do superbly.



Jun 06, 2004 at 10:16 AM
       2       end




FM Forums | Canon Forum | Join Upload & Sell

       2       end
    
 

Welcome back
Log in to your account