Lens back focusing

[singlespeed]

I have just taken some test shots to confirm the correct AF of my Sigma 18-200 f3.5-6.3. I was refering to, and using the test chart here.

18mm f3.5

18mm f11

18mm f22

200mm f6.3

200mm f11

200mm f22

Being as the AF couldn't lock onto the gray text, it was definately locking on the central block. I would say this is back focusing But i'm not sure if its at an acceptable level or worth sending back as it's only 2 weeks old.

Is this method flawed with the change in relative angles from the top/bottom of the page to the lens/sensor? Any help or info appreciated

 

[psenior1]

Doesn't this sort of test need to be at exactly a 45 deg angle for it to be accurate? To my highly untrained eye the first pics looks like it could be backfocussing a little - but generally looks OK to me....

 

[Canon Bob]

I think your method of testing is flawed in several ways.
The focus line on the test chart is insufficient for the AF sensors in the camera body to achieve a reliable decision. The chart looks to be far too close (for the 200mm end) and should be around 50x the focal length away from the lens...this reduces the chance of lens aberattions influencing the AF. A 45 degree chart is a reasonable way of measuring how far off the lens calibration is but it's not a good method to determine whether a probpem exists.
Mount a chart perpendicular to the lens at least 50x the focal length away. Use maximum aperture (the AF is always performed at max aperture) and initiate AF. Then manually turn the focus ring to see if sharpness can be improved. Alternatively, initiate AF and then move the target slightly forwards or backwards to see if sharpness is improved.
If this test shows that AF is off (front or back focussing) then use the test chart at 45 degrees to determine approximately how much....this will require several shots for repeatability purposes....and initiate the AF after setting the focus ring to infinity each time.

HTH

Bob

 

[HoppyUK]

Sorry to be blunt, but that test is notorious. While it is okay in theory, as Bob points out it is so far removed from the real picture-taking situations the camera/lens was designed for it is sometimes possible to detect slight shifts which lead people to believe there's a general problem in everyday photography.

On the evidence you've presented, I'd say there is nothing 'wrong' to be seen at all. Why did you do the test in the first place? Did you have reason to think that your focus was out? Stick to the real world with this lens or, if you like to photograph A4 sheets of paper at 45 degs, get a macro

Best regards,

Richard.

 

[Richard Peters]

Sorry to be blunt, but that test is notorious. While it is okay in theory, as Bob points out it is so far removed from the real picture-taking situations the camera/lens was designed for it is sometimes possible to detect slight shifts which lead people to believe there's a general problem in everyday photography.

I agree, I setup my 200-400 and 600 using this test, got it all looking nice with the fine tune then went out to take some real shots and found out I had introduced a major backfocus!

 

[singlespeed]

Thanks for the objective replies.

I did a couple of tests due to the AF giving several duff results yesterday when togging some swans. The first test was square on to photograph some text in a book and the results seemed to be acceptable sharpness and repeatable over a few shots.

Then I did the test above and it showed up the possible back focus problem. However I wasn't convinced this is actualy a problem or a flaw in the test which on the face of it seems be a good idea :bonk:

Right, I've been having a think which is usualy dangerous. The thing which i was uncertain about is the relationship between the paper and the lens. If you look at the pic below, the center line of the lens and AF is targetted on the center of the paper.

The cone is the same angle, yet the distance between the AF point is shorter at the front than the rear. I think this is the cause of the apparent back focus, when the lens is probably spot on

 

[Canon Bob]

The focal point will usually "appear" shifted backwards and it becomes increasing so as the focal length increases. The point of sharp focus (assuming calibration is spot on) will be the centre line but your brain will see a little more depth beyond that line and tell you that the true point is moved backwards....because you're conditioned to think that it should be centred.

Bob