Am I just unlucky, or is this how 24-105 f/4 L lens is supposed to be?

After reading all the rave reviews, and after falling in love with my 70-200 f/4

L lens, I caught a bad case of L craving and bought a 24-105 f/4 L. However,

testing it against my 24-85 non-L lens, I discovered that in most circumstances,

the 24-85 was sharper. By "most circumstances," I mean in tripod-mounted shots

on my 20D of the same scene, at the same aperture. This was with the image

stabilization turned off. I exchanged that lens for a second one, but my second

one is no better. On the theory that the problem might be the UV filter, I tried

shots without it, but that wasn't the problem.

 

Yes, the L lens is built like a brick, but what's the point if my flimsy little

24-85 outperforms it? Am I missing something here? Could there be some sort of

user error? Or am I just unlucky?

My friend, the answer is obvious. You are missing the Nikon experience. Just kidding. Merry Christmas, Barry, and a wonderful new year.

Here are two 100% crops. Both are taken at the same aperture (4.5) and roughly the same shutter speed (1/800 and 1/1000). Camera shake shouldn't be a problem. The blurry one is the L lens.

Can you show us proofs or samples of the images of that lens?

 

Are you sure its not user error, for example, did you wait 2 seconds halfpressing then firing? The IS needs atleast 2 seconds to stabilize the image and itself, if you know what I mean.

Attached

Hi, Mars. IS was off; my understanding is that it can actually decrease sharpness if it's used on a tripod.

My 24-105 is one of the sharpest lenses I have and compares well against primes.

 

For the record I use Hoya SHMC Pro1 UV protective filters with a 20D.

 

I found no difference on tripod with mirror lockup if IS was on or off.

You said "most circumstances" That means it includes hand held shots,

 

But lets get to another concern, What is the serial number of the L lens? does it belong to the recalled batch?

looks like a focusing issue to me...

did you try manual focusing on the tree with both lenses?

The 24-85 seems to be a little bit sharper in general according to the reviews on photozone, but i don't think it's that much as seen in your sample pictures.

 

http://photozone.de/8Reviews/lenses/canon_2485_3545/index.htm

 

http://photozone.de/8Reviews/lenses/canon_24105_4_is/index.htm

Barry,

It looks to me like both images are out of focus, with the 24-105 being the worst. Try retesting using manual focus. If you see a difference than you have an auto-focus issue (are you half/pressing the shutter release to establish focus?). I separate the focus and exposure settings using the C.Fn-04. I have a 24-105 which is dead sharp. I have not used it on a tripod, just hand held though.

Nothing wrong with your 24-85mm. Send the 24-105mm back and be happy. Use what works for you. I have both. My 24-105mm is better than the 24-85mm in sharpness, contrast and color, probably the best zoom I have used. But the 24-85mm, specially on the 1.6 crop cameras, is not bad. I remember one pro who called it his money maker lens. I've tried to talk myself into selling the 24-85mm but haven't so far. Good luck. Use what you like and what makes good pictures for you.<div></div>

Are the images posted here cropped from the corner of your test shots?<br><br>

I am really not sure that this is a good test of the lens sharpness, so much as the camera's focus ability and decision making. The 24-105 shot doesn't look like the branches in the foreground are in focus. May I make a suggestion, that you take the camera inside and shoot at something within your house? That will account for some variables that might be different between the two shots.<br><br>

IS on the 24-105 will work on a tripod, but for something like this I would turn it off.<br><br>

You may have gotten a bum L, or perhaps you have a particularly excellent 24-85...<br><br>

Good Luck.

Thanks for all the answers so far . . .

 

Mars: By "most circumstances," I meant in controlled, tripod mounted tests. I tried each lens at different apertures and focal lengths and tried to make sure I was comparing apples to apples. In most of these comparisons, the L lens seemed less sharp.

 

As to both images being out of focus, note that I haven't applied any sharpening to either. What you're seeing is just what came out of the camera.

 

The crops I posted are from a little below the top edge of the frame, and in the middle third left-to-right.

 

The results, however, seem pretty consistent across the board, though I notice on this particular pair of images, that the L lens is sharper than the 24-85 at the extreme lower left of the image, but very oddly, not at the extreme lower right (where they are about the same) or at the extreme upper right, where the 24-85 is noticeably sharper.

 

Thus, it appears that in this test, there are areas of the image where the L did better than the 24-85, but for most of it, the opposite is true. Does that suggest that the camera is somehow twisting during the exposure? Or that something is misaligned in the L lens?

If you want to do contrlled tests I would really recommend using a flat target parrallel to the sensor and doing in doors (newsprint is the old standard).

 

That way you elliminate a lot of varaibles like small shifts in focus, DOF, field curvature and subject movement due to wind.

 

Even than you should repeat the test say three times with a new AF operation each time. This is because the AF will have some statistical variation. You then use the series of fstops from the sharpest of the wide open shots.

I took your suggestion. Very interesting results. I taped a sheet of newspaper to the wall and set up my tripod. I shot with each lens at 24, 50, and 70 mm, all at f/5.6. Here, generally, the L out-performed the 24-85. But what I thought was really interesting, and it fits with your comment, is that there seemed to be variations in the auto-focus. Thus, with the L, though auto-focus seemed to work well most of the time, it was way off in the shot I did at 70 mm. My manual focus was tack sharp, way better than the auto-focus, which resulted in unreadable print. More experimentation appears to be in order.

Ok, I've done further testing. With each lens I took four tripod-mounted shots of the newspaper on auto-focus, then four on manual focus. All were at 50 mm, same shutter speed, same aperture. This time around, the 24-85 auto-focus results were awful, plainly out of focus; the manual results were good; the L auto-focus results and manual focus results were both a little better than the 24-85 manual focus results, and were about the same as each other. I checked the middle and the corners on all images. So, it's looking to me, as someone suggested, that it's a focus issue, not an optical quality issue. Or is it even a focus issue? Is there just naturally a certain amount of variation in auto-focus from frame to frame?

I have both the 24-85 and 24-105 L and can say the L out performs the 24-85 by a wide

margin a F4-5.6--sharper and more contrasty. It's also much sharper on the edges of the

frame, even with FF. Stopped down to F8 there ain't much diff.

With L zooms you're mainly paying for wide open perfomance and beefy build. Most lenses

are excellent at F8 or F11.

 

I also have a 20D and 5D and find the 24-105 L does much better on the 5D. Why? AF is

more accurate. The 20D misses more often. There may be AF issues with your 20D.

You might want to get the lens and camera calibrated together. Canon will do it free of

charge and it will probably improve your hit rate. I had my 70-200 4L calibrated and it

improved greatly (it sucked bigtime).

How do I go about doing that? Drop the camera and lens off at my local Canon dealer? How long does it take?

I tested the 24-85 lens, and @ 50mm-f/8 it was 'almost as sharp' as one of my 50mm primes (The problem with that lens is distortion - one of the worst at 24mm...). So @ f/8, you might have a difficulty with judging sharpness.

 

Most likely, as others suggested, the problem is one focusing. I think nowadays people get their lens-camera combos calibrated (I do not know how long that takes, or how much it costs). I would think twice about issues that might arise with other lenses you use after you calibrate your camera with the 24-105.

 

Good luck.

Mine wasn't consistent when I got it new so I sent it in for calibration and now you couldn't rip it from my cold, dead fingers. Best lens I have ever had.

I have owned both lens and 24-105 is pretty much sharper especially under a 23 inch

display. I think you should return your 24-105. There must be some pb on it. BTW, forget

about the lens test on photozone. Their testing result on 70-200 f2.8L IS is so bad but in

reality is not. In fact, I doubt using 350D to test lens is a good idea.

It sounds like you may have an AF calibration issue then as you suggest. When I test my lenses I normally do 3 AF operations just to ensure that I don't use a 'bad' one.

 

Normally however they are fairly close, often almost exactly the same.

 

Just to confirm, you are doing the tests in reasonable lighting? This can affect AF accuracy.

 

I test in doors with the target in a bay window, so it is reasonably bright but not very. Typical exposure was 1/8 f4 ISO 100 for your info.

 

Good Luck!

I've had mine just over a year, and it's a favorite of mine. Here's a link to a small gallery of pics my wife asked me to take in her garden a while ago.

 

http://deanwcains.smugmug.com/gallery/2075924

Given that the lens, under proper conditions (particularly, manual focus) can be sharper than anything I get with the 24-85, I'm beginning to think that it's an autofocus issue, rather than something inherently wrong with the lens. This is particularly so, given that I have now had two copies of the lens, and they have each performed the same. There probably wasn't any problem with the first one, either. Now I'll have to check into calibrating it, the camera, or both.