24-70mm Blurry Images Horror Stories

Well, it's all hand-held,...I am telling you man...I am no PRO, just really getting into this business, but my little, tiny visual experience tells me that THERE IS NO WAY ON EARTH I am paying 1300 dollars to get that image when my plastic, cheap kit lens get that other one...HAND HELD!!!!!!

This baby is brand new and it just came back from the Factory Service Center... If this is not a Horror Story I wonder what else could be...

I'd try it again with a tripod. The difference in the weight btw. the two lenses is considerable & could have an effect on IQ

Persephone wrote in post #9502408
The 17-85mm shots are of course going to be at f/5.6 because they were taken at the long end of the lens. The OP, I am sure, bought the 24-70mm to shoot at f/2.8. Besides that, since the lens is a f/3.5-5.6 lens, wouldn't the shot be considered to be "wide open" and the "sweet spot" for that lens being f/8 to f/11?

Fully understood and this is why I will only use a constant aperture zoom.
The test is still not apples to apples.

Interesting how people say f/2.8 softness is over blown/hyped on the internet. I've found my 24-70L seems to have better sharpness wideopen when I went 5D classic and more so with my 5dmkII. In my case the XTI and 50D was a tad soft wideopen with the same lens.

Juan,

ditch the lens. Thats unacceptable.

Hand held at 1/20 of a second at 70mm? Of course it's going to be blurry. The cheaper lens has IS helping it out. Plus there is something amiss with the photo, the DOF is not the same which leads me to believe it's front focussing. It's not a fair test, use a tripod, MLU, manually focus, go somewhere with more light and up the shutter speed. That second photo from the brick shows camera shake.

Juan Huerta wrote in post #9502385
APPLES TO APPLES...sharp to crappy.
Full manual test

Juan, at 70mm, a 1/20s just isn't going to be sharp with no IS. If you need IS (i.e. lower shutter speeds), then the 24-70mm is certainly not a good choice.

But those were obvious comparison shots. Don't you have any real-world shots that show your original issue with the lens? If they're all at similar shutter speeds, then camera motion blur is likely the biggest (and perhaps only) issue.

Juan Huerta wrote in post #9502385
APPLES TO APPLES...sharp to crappy.
Full manual test

Not denying that it may be a bad lens: but this test right here is seriously flawed.

A good rule of thumb to eliminate camera shake is to use a shutter speed that is the reciprocal of the focal length. In this case, you would want to be faster than 1/70th of second to be on the safe side. Since you're on an APS-C sized body (the 40D) the field of view is decreased, making the effective focal length equal to 1.6x the actual focal length. So, in this scenario, to be on the safe side from camera shake you need to use a shutter speed higher than 112th of a second. 1/20th is far too low and you're going to definitely get some camera shake in the shot.

Your 17-85 on the other hand has an image stabilizing mechanism built in, so it's naturally going to be much sharper at that shutter speed.

This is what you need at the very least to do to prove the lens is absolutely as bad as you think:
Tripod
Mirror Lockup
Manual Focus

okay I have some questions. Why would you say to MF the lens? If he's not going to manual focus it IRL, then a true test of whether it's front focusing or back focusing would be how it actually AF?

Tripod yes. Especially at those shutter speeds.

Mirror lock up? Why? Isn't the point of these tests to see if the lens will work in real life circumstances? If he uses mirror lock up all the time then yeah, test it with that. Otherwise, set up the tripod, use a shutter release cable to eliminate user shake and test it with AF.

We can't forget that these lenses are used for more than cardboard tests. They have to work in real life and more often than not, that does not include manually focusing, mirror lockup shots.

Permagrin wrote in post #9502462
I'd try it again with a tripod. The difference in the weight btw. the two lenses is considerable & could have an effect on IQ

Yes, that's way too low a shutter speed to be hand held for testing. Why so low???

-otherwise-

Send the camera and the lense to Canon for calibration.
This is a very sharp zoom, and it's not voodoo that makes it work; if something is misaligned or out of spec, Canon can detect it.
It's either the lense, the cam, or both.

blackhawk wrote in post #9502592
Yes, that's way too low a shutter speed to be hand held for testing. Why so low???

-otherwise-

Send the camera and the lense to Canon for calibration.
This is a very sharp zoom, and it's not voodoo that makes it work; if something is misaligned or out of spec, Canon can detect it.
It's either the lense, the cam, or both.

lol Great response!

Permagrin wrote in post #9502567
okay I have some questions. Why would you say to MF the lens? If he's not going to manual focus it IRL, then a true test of whether it's front focusing or back focusing would be how it actually AF?

If he's testing the AF, yes. If it's the elements simply being mal-aligned so that it's horrifically soft, that's another issue. He should use MF just to narrow the diagnosis.

Permagrin wrote in post #9502567
Mirror lock up? Why? Isn't the point of these tests to see if the lens will work in real life circumstances? If he uses mirror lock up all the time then yeah, test it with that. Otherwise, set up the tripod, use a shutter release cable to eliminate user shake and test it with AF.

I thought the point of these tests was to figure out whether or not the lens was unacceptably soft, even in the best conditions possible. Otherwise he would probably be out shooting right now, not worrying about the sharpness.

Permagrin wrote in post #9502567
We can't forget that these lenses are used for more than cardboard tests. They have to work in real life and more often than not, that does not include manually focusing, mirror lockup shots.

Disagree entirely on the manual focusing issue, as that narrows to issue to simply autofocus or a generally soft lens. But if you're going to put it on a tripod and use a remote release why not use mirror lockup just to be absolutely sure? I don't see the harm in it.

Permagrin wrote in post #9502602
lol Great response!

Thanks, not sure why some people rip into this lense, I think it maybe the $1G twitch switch that does it

TheBurningCrown wrote in post #9502625
I thought the point of these tests was to figure out whether or not the lens was unacceptably soft, even in the best conditions possible. Otherwise he would probably be out shooting right now, not worrying about the sharpness.

Canon can measure the lense to see what it's MTF profile actually is, among other tests.
If the AF calibration is off, it's either calibrated and/or defective parts are replaced. Optical elements can be checked for alignment as well as the housing and barrel.

When I'm shooting without a tripod at 1/20th sec, I don't expect every shot to be a keeper... and Canon repair is a lot better set up to do these "tests" anyway.

Point is go out and shoot! Otherwise send it to Canon so you can!

No complaints with mine and my friend has the same lens on a 40d. We love it.

Juan Huerta wrote in post #9502247
Gee, I looked into the stuff I have posted here and don't remember saying that I wanted the L-series to improve my skills. I am just concerned about the lack of sharpness on such expensive lens, right out of the box, that's all.

And for the records, I was right. The Factory Repair Service report states that after running their tests they DID find out the lens was in really bad shape, optically and electrically speaking.

Yes, I understand, and I'm not denying that the lens could indeed be a very bad copy. My point is that you have this expectation that simply because the lens costs X amount, basic methods and techniques for hand held exposure go right out the window. Depending on the situation, yes, you may have to brace yourself against a wall or use a tripod to get a sharp image, even if the lens was $1300.

Again, that is not to say that your 24-70mm is or may be really terrible. But like others pointed out, shooting at 1/20 hand held at 70mm is bound to create some blur.