Canon focus mystery

<p class="MsoNormal"]Here is an odd one:<span>
Why do I get perfect focus results on a 7D, and varied results a 5D MKII
and 50D requiring micro adjustment with the exact same lenses?
<p class="MsoNormal"]I own a 5D MKII and three lenses, a 70-200mm 2.8 IS II, a 20mm
2.8, and the 24-105mm kit lens <span>and two of
these three lenses do not focus properly and require micro adjustment.<span> The 70-200 was dead spot on out of the box,
the 20mm was just a tad out but not really noticeable in pictures, and the
24-105 was so bad, it was just plain blurry.<span>
I sent the 24-105 to Canon where they adjusted it somewhat but it still took
blurry pictures.<span> Of course, the micro
adjustment in the camera makes everything perfect for these two errant
lenses.<span> So, except for some worry about
Canon quality, I am very happy now with great focus performance on all three
lenses.
<p class="MsoNormal"]Now, my buddy purchased a 50D and when I put my lenses on
his camera, I got the same focus results requiring micro adjustment (and yes,
the 24-105 was just plain blurry).<span> OK,
so far so good, the issue appears to be the lenses with consistent results
across bodies. <span>Unfortunately, he sent
the camera back because he was getting Error 80 seize ups but instead of
another 50D, he sprung for the 7D with a 28-135 kit lens.
<p class="MsoNormal"]Now here is where it gets weird:<span> All four lenses focus dead accurate on his 7D
with no adjustments needed.<span> The thing simply
focuses just plain spot on.<span> Huh?!?
<p class="MsoNormal"]Ok, what’s up with this?
<p class="MsoNormal"]
<p class="MsoNormal"]Some notes:<span> We never
bothered to micro adjust on his 50D as so technically, I can’t say the results
were exactly the same, but, given our eyeballs, they were the same and the
24-105 was obviously incorrectly focusing on both cameras to a large
degree.<span> We did not try his 28-135 on my
5D MKII.<span> Looking back on it, I probably
should have exchanged the whole 5D and kit lens for such poor focus performance
but I didn’t as I could adjust it in camera (1 click less than the max).<span> I decided to send it back after a few months
because it just bothered me.<span> <span>For focus testing, I used to use a print out
laid on the floor with the camera at a 45 degree angle – what a pain in the alignment
– but now the best I found is a focusing target you display on your PC screen
and then your live view with the resulting moiré will show perfect focus and
very quickly. <span>Read about this target
here:<span> [url="http://www.komar.org/faq/camera/]http://www.komar.org/faq/camera/[/url] <span>(Alek is a great guy!<span> Thank him if you use his targets.)
<p class="MsoNormal"]

Re: Canon focus mystery

I have found the same lens acting differently on different cameras, while others are all the same. its a bit of a mystry, but not a problem, since AF micro adjust takes care of it. its the reason i will not buy a 60d.

Re: Canon focus mystery

Now that you mention the topic, I

Re: Canon focus mystery

Carlos: Yes, well this is part of the mystery isn

Re: Canon focus mystery

Quote:

Originally Posted by Carlos Lindado

If the autofocus sensor detects proper focus, it's because the lens has focused correctly on the subject and is "transmiting" a correct image to the autofocus sensor, so I really don't understand why the image would be unfocused when taking the picture.

First, lets catch everyone up on how phase detection autofocus works:http://www.lensrentals.com/news/2010.07.05/how-autofocus-often-works. Scroll down to the section about phase detection for info related to this discussion, though the rest of the article is very interesting as well.

The problem lies in the fact that the autofocus sensors and the image sensor aren't one in the same. These two components need to be aligned properly so that light focusing on the autofocus sensors when the mirror is down is also in focus on the image sensor. This might be why some lenses work wonderfully on one camera body and not so great on another.

Toward the end of the article is a section on disadvantages of phase detection autofocus. Here is a quote from the article:

Quote:

The light path to the imaging sensor must be calibrated with the light path to the autofocus sensor, so what is in focus for the autofocus sensor is also exactly in focus for the imaging sensor. Each lens contains chips that provide feedback to the camera, telling it exactly what position the focusing element is in and how far it moves for a given input to the lens motor. This must agree completely, so that the lens actually moves exactly where the camera told it to move, and so the camera knows exactly what position it’s in. If any of these systems aren’t calibrated perfectly, autofocus becomes inaccurate. Even if they are calibrated perfectly at the factory, if they expand and contract slightly with heat or cold they may become inaccurate, at least temporarily.

I would image the purpose of AF microadjustment is to tell the camera to modify the distances it tells the lens to move.

Hopefully this helped. However, I may be wrong about some of this, since my knowledge comes only from this and a few other articles I've read. If someone knows more than I do about this topic, please share your knowledge as I'm always happy to learn.

Re: Canon focus mystery

hey bruce i also have a 50d and have noticed some focus problems to. i have a sigma 17-70 canon 70-300 III and canon 100-400 L the 100-400 works best for me but still get some bluring but my sigma seems to allmost allways need correction.i want to get a 7d or mabe 5dII so ive been a littlle concerned about this but gald to hear good reviews on 7d thats the one i like most.....rod

Re: Canon focus mystery

I think that Trowski hit the nail on the head (after the lensrentals.com article author). Misalignment between imaging sensor and AF sensor is the culprit.

Re: Canon focus mystery

Consider also that there is a third path: from the main mirror to the focussing screen. If the screen is not in the right position then the viewfinder can appear out too. I believe Canon use thin shims to align that. I had an original EOS1 that had VF disagree with AF.

Paul

Re: Canon focus mystery

Hey Trowski, thanks! That article definitively unveils many mysteries about phase-detection autofocus, it was the explanation I was looking for!

I think some bodies could just "tell" the lens to move its focusing group a certain distance, but don

Re: Canon focus mystery

Well, the article does explain why AFMA is necessary, but doesn't really address Bruce's fundamental mystery as I interpret it. I've seen something similar myself, as has scalesusa it seems.

For example, in my case:

The 16-35mm f/2.8L II requires an adjustment on the 5DII that is 4 units negative relative to the adjustment on the 7D.
The 70-200mm f/2.8L IS II requiresan adjustment on the 5DII that is 4 units negative relative to the adjustment on the 7D.
The 85mm f/1.2L II requiresan adjustment on the 5DII that is 2 units positive relative to the adjustment on the 7D.
The 100mm f/2.8L Macro IS requiresan adjustment on the 5DII that is 6 units positive relative to the adjustment on the 7D.
The 24-105mm f/4L IS requires the same adjustment value on both my 7D and my 5DII.

<div>I had thought there would likely be systematic differences - a particular camera was x units off in one of 'true', and a particular lens was y units off of 'true', in theory if you knew those values for a given camera and lens you could predict what the AMFA should be for the combination of them. But in practice, that's not the case - each combination (at least with the bodies and lenses at my disposal) seems unique.</div>
<div></div>
<div>So, from a theoretical standpoint is remains a mystery. From a pragmatic standpoint, it's irrelevant - I will just continue to do an AFMA for each body+lens combo I have...</div>