Questions about autofocus

[RalphLambrecht]

Hi everyone,

I've got a few questions, I'm hoping someone can set me straight.

I bought a Canon EOS 1000f body so I could use my 85mm f1.8. I was out shooting and I wanted to give the AF a try, but I remember the camera focusing just behind my subject, whereas I intended on focusing on the face of my subject. Anyway, I turned the AF and focused manually.

But now I've just gotten the roll back, and I can see that while it appeared that I was focused on the subjects, I was actually focused just behind the subject!

Does anyone have any idea of why this might be?
Is it an issue with using a newer lens on an older camera?

I've included an image where you can see how it's focused on the plants just behind the subject.


Thanks, and any help is very much appreciated!
there apparently is a misfit between the camera lens mount and the lens mount itself.You may be able to shim this with a piece of paper yourself or you need to have the camera CLAed.
Ryan.
View attachment 191554

 

[shutterfinger]

Yes, I googled for exactly "Canon EOS 1000f" but nevertheless got as result at top an excerpt of the Wikipedia article for the d-model..

Google use to be a great search engine but has increasingly become less reliable as the current generation of users can only look at pictures or watch videos. The programing for the search function is now what the programmer thinks you want whether it matches the search term or not. I sometimes find what I'm looking for on page 2 to 4, if it hasn't shown up by page 5 I try a different search term. The problem is worse with technical items as the coders have no clue beyond the latest chip designs.

 

[Theo Sulphate]

Establishing focus error can be done easily:

1. Go outside in bright sun or in a very bright room.

2. Stretch out a tape measure its full length, across a table or along a fence. If you don't have one, then use playing cards or make little test targets out of paper.

3. If it's an AF lens, focus and make photos at several marked points on the measure, e.g. 2', 4', 6'; if it's a manual focus lens, do the same - however first ensure your diopter setting is correct (if you have one) by focusing at infinity and aiming at something about 200' away or more. Use the widest aperture on the lens for your test photos.

4. Examine photos to determine if focused point was really in focus.

I generally take lens distance markings as a very rough approximation - even on my very best Hasselblad lenses they disagree with each other.

 

[cooltouch]

You can take Theo's recommended test a bit further if you wish to find out the exact amount of discrepancy. Warning! It involves math. The Pythagorean Theorem, to be exact. First you need to measure the exact distance from the floor or ground to your eye. You can use the same measuring tape to find this number. Then, lay the tape measure on the ground and mark out the point or points that you want to take your pictures at along the tape measure. To make this easier to see, you can lay out pieces of paper next to the tape measure with bold, easy to see gradations, for example. You want to record these values of the distances you focused at -- that is, where you saw the image to be in focus. Your eye height and these horizontal distances form the legs of the triangles we'll be using for our computations. Then, after you've gotten your film back, closely examine the photo(s) to determine where along the tape the focus point actually occurred. For each point where you focused and took a picture, you'll have two different triangles -- the one with a horizontal leg that has points you saw to be in focus and the one with a horizontal leg that contains the actual point of focus as taken by the camera. And then you'll want to solve for their hypotenuses. The square of the hypotenuse is equal to the sum of the squares of the other two sides, yadayada. Do the math. The difference between the two hypotenuses is the discrepancy that is occurring.

Doing this test will also tell you whether you're having a front focus or back focus error. Judging from your photo, though, it looks like a back focus error. This is not at all uncommon. Shortly after I bought my first DSLR, I had to return it to Canon because it was back focusing by a noticeable amount. And I'm sure I wasn't alone.

 

[shutterfinger]

While marked distance scales on lens are not critically accurate they will show if a lens is off a foot at 15 feet, about the distance Ryanse was from the subject in the thread example and is much faster than waiting for film to be processed.
Now if he had focused on her eyes and her nose was sharp and the eyes soft at f2 then I would have suggested/recommended a more critical test.
If the focus was off less than 6 inches at the subject distance then the problem would be less pronounced than it is.
Quick and simple test to determine if a problem exists, refined test to determine the exact amount if needed.

 

[shutterfinger]

Please update us when you find the culprit.

 

[E. von Hoegh]

Google use to be a great search engine but has increasingly become less reliable as the current generation of users can only look at pictures or watch videos. The programing for the search function is now what the programmer thinks you want whether it matches the search term or not. I sometimes find what I'm looking for on page 2 to 4, if it hasn't shown up by page 5 I try a different search term. The problem is worse with technical items as the coders have no clue beyond the latest chip designs.

Phew!
Thanks, I thought it was me forgetting google search syntax!

 

[naaldvoerder]

The problem probably stems from a misaligned focusing screen in the camera. My (digital) 5D body showed the same problem when manual focusing with a 1.4/85 mm mounted. In that particular model Canon never bothered to seat the focusing screen properly, because they assumed everybody would autofocus anyway. A set of shims was/is available from Canon to fit the screen properly yourself.

 

[AgX]

Quite weird behaviour of Canon.