forward focussing and service advice

Deadlock

Hi all,

I recently discovered my 20D was forward focussing.
I was performing a test ziggy53 recommended - high contrast object on a fence 3 metres away - and found that my old canon 100mm macro to be forward focussing by about 5cm and the just purchased 24-70L to forward focus between 50cm and 1 metre. I returned the 24-70L and replaced it with another.

I also took the camera into the canon repair centre. They confirmed back focussing and 'fixed' it.

When collecting the body I took in my lenses to test and the new 24-70L matched the focal point of the 100mm macro. However, they were now both back focussing by approx 1-2cm for an object about 1 metre away.

So, after being serviced by canon to fix a back/forward focus issue, what do they think becomes an acceptable amount of error?

Surely I am not being too critical in wanting it to be accurate?

ziggy53

Front focus and back focus are confirmed by a flat target.

http://www.dgrin.com/showpost.php?p=1433603&postcount=10

Quoting myself, "2) A fence line or similar, shot at an angle to the subject. Put a singular strong-contrast target on the top of the middle post and use a single focus point in the camera to focus against the target. This shows focus accuracy and/or how easily distracting for/aft objects influence AF accuracy (compared to the above wall shot). It also shows bokeh tendencies at different aperture settings."

As long as the "wall" test (test #1 in the above link) shows accurate focus, test #2 will show tenancies of the AF system to be distracted by surrounding objects. People believe that AF systems allow "point accuracy" because the viewfinder shows points indicating the centers for the AF points. In reality, AF points are more nebulous. If test #2 shows consistent deviations, after test #1 shows focus accuracy, it mostly means that you need to change your AF strategy and technique.

BTW, the least nebulous, most defined, AF points exist in the advanced AF systems of the Canon 1D/1Ds AF systems.

Deadlock

Thanks,

Re, the first 24-70
- saturation was poor compared to the 100mm macro
- forward focussed on both wall and on fence tests.

Re the second 24-70 tested post 20d 'fixing' of the forward focus issue
- no wall shot this time just a variety of objects.
- saturation was as good as the 100mm macro
- for all real world examples of subjects taken (apple, mobile phone, fire extinguisher, donations box, etc) the focus point was approx 1-2 cm behind the AF point displayed in Digital photo professional. However, the observed focus point was exactly the same this time as the 100mm macro.

I'm now exploring portraiture (where I use the closest AF point to, and position on, the leading eye), as opposed to landscapes, where the dof is more critical. In macro work I'd always manually focussed. This is why I'm noticing it now.

I contacted canon uk after making the opening post and explained the situation to them. I was told to expect the AF point to focus 'exactly' on the subject.

ziggy53

Saturation qualities of a lens do not vary unless the lens is exposed to some sort of lens flare.

Please give links to full sized images with full EXIF if you want further help.

Deadlock

Thanks.
I've just received my camera back from the repair shop.
I shot this brick wall at 24mm & f2.8
I'd appreciate your impressions.

brick wall

ziggy53

Deadlock wrote: »

Thanks.
I've just received my camera back from the repair shop.
I shot this brick wall at 24mm & f2.8
I'd appreciate your impressions.

brick wall

Contrast is a little less than I would hope for, but sharpness is not bad for f2.8. Focus looks good. (Looking at fine detail like cracks in the mortar, for instance.)

Did you use a lens hood?

Was there a filter on the lens?

What was the distance to subject?

Deadlock

Did you use a lens hood?
- Yes (sun approx 45 degrees behind camera left)

Was there a filter on the lens?
- No

What was the distance to subject?
- Approx 1.5 - 2m

Can I ask you also what you think of the amount of barrel distortion?

ziggy53

Deadlock wrote: »

Did you use a lens hood?
- Yes (sun approx 45 degrees behind camera left)

Was there a filter on the lens?
- No

What was the distance to subject?
- Approx 1.5 - 2m

Can I ask you also what you think of the amount of barrel distortion?

I see no reason for lens flare so this appears to be what you can expect from this lens in this circumstance.

All zoom lenses which include a range of wide angle coverage tend to have some barrel distortion at the wide end of the zoom. The Canon EF 24-70mm, f2.8L USM is very well corrected but it does have some barrel distortion in this shot. For more normal "organic" subject matter it should be a non-issue but for architectural images you can treat the image to some post-processing correction, if you desire. This image sample looks pretty normal.

Deadlock

Thanks ziggy. I appreciate your input.
Normal is good!