Canon lens possible back focus issue

Scotty Pro

Scotty Pro

Messages
841
Name
Paul
Edit My Images
Yes
Just messing around at the moment since the weather up here is really miserable and I came across this shot of an old lens I took.

Upon closer inspection it seems that the sharpest part of the image is around the outer lens barrel and not in the centre of the Canon
(my focusing point was the centre n)

Does this come across as a slight backfocus issue, or am I nitpicking ?

 
This wasn't shot on a tripod, it was hand held, the exif should be intact
f5.0
1/60
flash fired
 
Looks like my tripods coming out, will try the Duracell battery test, will that show if there is a problem and what setting do I choose f1.4, for an accurate test.
 
Battery test

the one on the right is exactly half a battery thickness behind the one on the left, looks as if the right one is a bit more focussed than the left which is what the centre focus point was aimed at. Not very sharp either, me thinks its going on a journey.

[URL="
 
Hi mate,

Not sure if the 40D can do micro adjustments? I had a play with it on my 50D the other week & had some good results.

Here http://www.dgrin.com/showthread.php?t=116078 is a good article on the process. I definately improved my Canon 70-300 USM IS.

Mick
 
Mickwreay said:
Hi mate,

Not sure if the 40D can do micro adjustments? I had a play with it on my 50D the other week & had some good results.

Here http://www.dgrin.com/showthread.php?t=116078 is a good article on the process. I definately improved my Canon 70-300 USM IS.

Mick
Click to expand...

no micro adjust on the 40D :bang::bang:

going to take a couple of more shots with a test chart I found, if it still shows a problem then I might just take it back tomorrow and ask for an exchange.
 
Ideally, Paul you need to allow a decent sized perpendicular area for the lens to focus upon. This means you may need to adapt the focus chart you've found.

Generally the charts are designed to be held @ 45 deg but that can lead to errors in focus!

If you do a search on here you'll find some designs for slightly more reliable targets. I believe it's also recommended to shoot at 50x focal length.
 
Don't use a chart at 45 degrees and don't shoot anything at an artificially close distance just so the depth of field is very shallow.

It is not unusual for some lenses to be a fraction out at very close range, and if you correct that, it will then be out for normal distance shooting. Canon recommends 50x focal length, but 25x is okay I think.

There are quite a few threads here on this.
 
I'm going to throw a spanner into the works....

It doesn't look bad enough me to warrant any major concern.
I've seen far worse where front / back focussing issues are far more evident.
I'm not having a dig at you but I think that sometimes there is too much emphasis on pixel peeping that photography itself.
If you showed the photo to someone who is not into photography, I would lay odds on that they wouldn't fault it.

Does a small stone chip spoil the look of a nice car?
 
Last edited:
Looking at that image of the batteries, to me it looks like your focal point is on the engraved numbers going up the side of the left hand battery, maybe the AF struggles with the shiny surface of the battery and found the etching to be the only thing to grab ?
 
BigRuss said:
Looking at that image of the batteries, to me it looks like your focal point is on the engraved numbers going up the side of the left hand battery, maybe the AF struggles with the shiny surface of the battery and found the etching to be the only thing to grab ?
Click to expand...

The problem with a test like that is that it's at such close range that DoF is measured in millimeters, and with a non-flat target you don't know where the AF sensor is actually locking on.

The area of the sensor is quite a bit larger than the little square in the viewfinder and within the active area it will tend to lock on to the best bit of contrast, which may not be quite in the middle.

The only way to be certain is to use a perfectly flat target. And a much larger one too, to avoid the potential problem of very close issues affecting the result at more general shooting distances. This is the target I use - a cerial box and rule.

Apart from that, Hashcake makes a good point - too much pixel peeping is bad for your health. No AF system is absolutely perfect, at all focusing distances and all focal lengths. Personally speaking, when it comes to working with very shallow DoF, the problems of user error through shifting the camera, or the subject moving, or focus-recompose errors, are much greater than anything introduced by the camera. Within reason of course ;) It's good to get our equipment working as well as it can, but without getting too obsessed about it.

IMG_3381-1.jpg
 
Richard, thanks for the explanation and the photo of your test subject. I have found the focus chart and printed it out on my laser, it suggests lying it flat and mounting your camera on a tripod at 45 degrees, then focussing on a certain point. Will try that and also try the setup above, everyday things are easy to source.

In the image above where is your actual focus point ? is it on the mark on the rule ?

I'm not really pixel peeping (or am I), it was just when I looked at the lens in the first post I thought it was quite obvious that the focus was aimed at the letters on the outer rim when in fact it should be the centre point.
 
Scotty Pro said:
Richard, thanks for the explanation and the photo of your test subject. I have found the focus chart and printed it out on my laser, it suggests lying it flat and mounting your camera on a tripod at 45 degrees, then focussing on a certain point. Will try that and also try the setup above, everyday things are easy to source.

In the image above where is your actual focus point ? is it on the mark on the rule ?

I'm not really pixel peeping (or am I), it was just when I looked at the lens in the first post I thought it was quite obvious that the focus was aimed at the letters on the outer rim when in fact it should be the centre point.
Click to expand...

No, focus on the front of the box, close-ish to the rule. If you focus on an angled target, you don't know where you are. The rule is just there to show you exactly where the lens has actually focused, in front or behind the face of the box.

The focus chart that I think you have downloaded is often criticised for being shot at an angle, but the way the target is presented on it the AF sensor can't really get that wrong. The main problem with that is the artifically close distance that it's used at. With a longer lens, even with the chart on the floor and a tripod at full stretch you're just going to be unrealistically close where there might well be fractional errors that don't exist at normal range. And the further away you are, while it might be more difficult to see what's going on because depth of field increases (which also tells you something about real world tolerances) the less critical the setting up becomes.

The set up with the cerial box is essentially identical to something like the LensAlign tool which costs $80 here :eek: http://www.whibalhost.com/lensalign/ That guy seems to make a decent living out of selling ridiculously priced and unnecessay widgets. I just do it in the kitchen. The worktop is at a convenient height and there's usually a pattern on the floor to help line the tripod up square. Just use a tape measure to get the lens at the same height and you're set. If you have a decent LCD, your can zoom in on that and see what you've got.
 
You must log in or register to reply here.
Back
Top