Can EF135/2L be better than this?

[aleksandar_kujucev1]

Hi,

<p>I just went back from a New York shopping trip.<br>

I bought EF 135/2 L, EF 85/1.8, TS-E 90/2.8 lenses and a EF 100/2.8

macro for

a friend.</p>

<p>After returning from B&H to my hotel, I wanted to check my new

lenses for

problems, since I've read a lot about Canon QC issues lately.</p>

<p>I've put my 10D on a tripod and started taking pictures through

hotel window.

<br>

I placed the same object in the center and in all four corners of

the frame.<br>

Since the objects I've chosen were near infinity focus, I believe

that recomposing

was not an issue.<br>

I tested all lenses at full aperture for focusing issues, lens

elements misalignment,

etc...</p>

<p>The 85, 90 and 100 lenses performed as expected, but on the 135 I

noticed a

pronounced sharpness loss at left side of the frame, especially

lower left corner.</p>

<p>I took it to the B&H for a replacement. I tested the second

sample right

in front of the store and judging from magnified LCD view it seemed

OK.<br>

I'm apologizing for the technical quality of the pictures. The

weather was really

bad for taking pictures, dark cloudy rainy day, so I had to raise

the ISO to

400 since I only had a kind of flimsy tripod that I borrowed for

testing. But

I think that you can see the difference between the lenses easy

enough...</p>

<p>The shots are all JPEGs - I was judging the performance from the

LCD at maximum

magnification, so no need for RAW. I know that viewing on a computer

would be

better, but I had none at the time. I also thought that if I notice

a flaw on

the LCD, it must be real!</p>

<p>I went back home the next day.</p>

<p>Now when I view the pictures from the second 135/2L on my computer,

they are

better indeed, but not perfect. Left side of the frame is a little

bit softer

again.</p>

<p>So the question is should I bother to return it again? <br>

This time from across the globe it wouldn't be so easy and I have no

guarantee

that the third lens would perform any better (it may even be worse).

</p>

<p>Or maybe I'm expecting too much? I was pixel peeping at 100%.<br>

I mean 135mm lens @ f2 must be of complicated construction and not

so easy to

manufacture. <br>

I was completely blown off by the sharpness at the right side of the

frame at

full aperture!<br>

Many other lenses I tried are much worse than the less sharp left

side of the

frame on my second 135/2L<br>

By f2.8 everything looks much better...</p>

<p>I just don't know if this is normal and what should I expect from

this lens!</p>

<p>What are the experiences with your samples of the same lens?!?!</p>

<p>I posted the samples on this address:</p>

<p><a href="http://www.kujucev.com/135/135.htm">

http://www.kujucev.com/135/135.htm</a></p>

<p><br>

The page is loaded with photos so it may be opening a bit slowly.

</p>

<p>There is also a quick comparison with my new EF85/1.8 too...</p>

<p>The first lens I bought is lens A and the replacement is marked as

lens B -

which looks better to you?<br>

I know the photos are not that good - I just didn't have time for a

proper shoot

since I got really busy right after returning home.<br>

Samples from lens B are especially of low contrast, but I think the

resolution

differences can be judged or at least guessed :-)</p>

<p>Thanks,</p>

<p>Aleksandar</p>

[aleksandar_kujucev1]

I forgot to mention that I ordered 1Ds mkII.

Because of the bigger sensor, it will be less forgiving on the edge unsharpness...

[PuppyDigs]

I had an EF 70-200 4L USM that rendered the left side blurry. It suffered from misaligned elements and needed to be repaired by Canon at a cost of $125. It was fine when I first bought it but the elements drifted off after about 2 years of light use (no dropping or knocks).

[digitmstr]

One Q I would have is this: how did you get the same framing on the opposite corners? Do you know what I mean? You didn't have the same pictures showing L&R side of the lense. You used two pictures and obviously, the lens was moved to frame the same subject on the opposite corners. That would mean that you either rotated the camera and re-focused or moved the tripod and re-focused.

 

I'd like to see the opposite corners detailsof the SAME frame.

[aleksandar_kujucev1]

Giampiero,

 

I forgot to explain focusing and recomposing part...

 

I focused only once using central AF point at the objects that you can see on the samples. Then by moving the camera on the tripod I recomposed and placed them for each shot in four different corners of the frame.

 

My camera is set to central AF point only with custom function 4 set to 1. This way I use the * button to focus and pressing the shutter button is not causing refocusing. I used the RS-80N3 remote switch to fire the shutter.

 

So I recomposed for each shot and put the same object in different corners. I already mentioned that the objects I've chosen were near infinity focus and that recomposing wouldn't affect their focus.

 

I tested this way, since DOF on this lens is extremely shallow and I wasn't sure that I had any objects that were absolutely plan parallel to the sensor plane.

 

The structures were also very different, so comparing the same object would bring out the differences more easily.

[digitmstr]

>>I focused only once using central AF point at the objects that you can see on the samples. Then by moving the camera on the tripod I recomposed and placed them for each shot in four different corners of the frame.<<

 

Aleksandar - is it possible then, that the difference be explained by the changed axis? Since ALL lenses showed the L side as the 'worse' side that could explain it: by moving the camera to the RIGHT of the orginal AF point you'd changed the axis and distance of the lens to subject therfore making the L corner softer.

[aleksandar_kujucev1]

Dear Giampiero thanks for showing interest in this thread :-)

 

 

>>is it possible then, that the difference be explained by the changed axis? Since ALL lenses showed the L side as the 'worse' side that could explain it: by moving the camera to the RIGHT of the orginal AF point you'd changed the axis and distance of the lens to subject therfore making the L corner softer.

 

 

I don't think so... I moved camera to the left of the orginal AF point too and the right side didn't become soft at all. Since I focused on the center of the screen, if moving the camera around would cause any softness, it would be the same in all corners. Anyway the objects in focus were too far away, so their focus could not be affected by changed axis (in my opinion at least)... Not ALL lenses were affected only two copies of 135s. Other three lenses 90, 85 and 100, were just fine - corners were almost identical, so it seems that my test procedure was working after all.

[digitmstr]

YOu are welcome :)

 

I would still suggest to try a test without moving the tripod/camera, taking a picture of a suitable subject (maybe a brick wall, building, etc...) and then examine the L & R corners just to see if the results are close to those you have now.

 

IMO, it would be best to eliminate all possible sources of variation/error, just to be sure.

[peter_foiles2]

I am trying to remember who it was, maybe Norman Koren, who did a detailed analysis of the focus and recompose method and found it to be a source of frequent focus errors. His findings (who ever it was) were that it was better to use one of the off center focus points rather than recompose. I know at the distance you were shooting it seems unlikely that the reduced dof of the 135 would be a problem but don't be too sure.

[frdchang]

well, the 85mm looks fine...

 

which gives weight to the 135mm being a poor performer.

 

(the 85mm 'suffered' the same flaw of recomposition, but it still looks fine)

 

if both lens sufferend a left blur, then i would suspect technique.

[olivier_de_lame]

Hi,

 

1) I have 135mm 2L and I use film (that is full frame). Such a problem would be huge in full frame. I do not have such a problem.

 

2) Using recomposition for testing the lens is problematic in itself. The problem is that the distance from the film (sensor) plane to the object is changed when you use recomposition. At 135mm @ F2, such a small difference is enough to create a bad focussing.

 

Regards,

 

Olivier