AF Micro Adjustment

[dje]

This is one of those "just interested" questions as my 600D doesn't have AF micro adjustment capability. But the higher end cameras do and I was wondering if anyone has ever had the need to use this or whether it's seen as useful ?

I did a simple AF focus calibration test on my camera by comparing results of a test chart with shots taken using manual and Live view focus (both operate on the main sensor) and AF focus (which operates on a different light path to the AF detect sensors). I couldn't really see much difference in sharpness so I concluded that my AF focus is calibrated fairly well. I did however only do this for one lens at one focal length.

I would be interested in other people's views on this fairly fundamental issue which I suspect most "normal" people don't even think about ! Thanks

Dave

[herbert]

Hi Dave,

Focus calibration is more useful when the depth of field on your lens is very narrow. So this would be wide aperture lenses, f2.8 or wider. You will notice it when shooting wide open, but when stopped down the depth of field increases and so tends to cover any slight focussing error.

When I adjusted my wide aperture and long focal length lenses my hit rate when shooting wide open did improve. I think the feature is a must for any camera body where you will be using wide aperture lenses.

Initially I did try to adjust the lens myself using angled targets and also using the Moire interference pattern technique. Both gave inconsistent results probably due to operator error and uncontrolled set-up. When I switched to using Focal software to do it automatically I got consistent results with a lot less effort.

Alex

[Colin Southern]

I've had a play with it a few times - never found any benefit. For a zoom lens there's probably too many variables anyway (focal length and focal distance)

[dje]

Thanks Alex and Colin. It's interesting to hear of your experience.

Dave

[Clactonian]

Those of us using older cameras don't have the facility anyway, so there's not a lot of point in us worrying about it. I think the best advice I was given is to learn as much as possible about your regular equipment, its strengths, weaknesses, and compensate accordingly. Still trying in my case!!

[rubyb]

I spent about 2 hours one afternoon trying to super calibrate 3 lenses- 2 zoom and one prime. It MAY have been useful with the prime, but with the zooms any change in focal length seemed to undo any benefit, and I eventually discarded it. I did retain a very minor adjustment on my 85mm 1.4.

[Markvetnz]

I messed around with this for a while after I convinced myself my struggles with my new 1dx were the cameras fault and not mine. I ended up confusing myself even more and eventually I found out that the 1dx firmware has a bug which causes it to lose all your MF adjustments every time you toggle the af orientation linked option. It has been a popular point on forums. In the end I gave up with it.

[herbert]

I spent about 2 hours one afternoon trying to super calibrate 3 lenses- 2 zoom and one prime. It MAY have been useful with the prime, but with the zooms any change in focal length seemed to undo any benefit, and I eventually discarded it. I did retain a very minor adjustment on my 85mm 1.4.

Good point. It can happen that the focal distance and length effects the AFMA. In the case of distance you should calibrate at the focal distance you use most often. With my telephoto lenses that is usually 15-30m. If I get bad performance elsewhere then I am unlikely to notice since I shoot those distances less often.

As for focal length the new Canon cameras (5D3, 1Dx) can AFMA at two positions of a zoom range. The AFMA is then linearly interpolated for other focal lengths. It's not perfect since you ideally require a table for AFMA across the entire zoom range but it will be no worse than a single AFMA value. On older cameras the advice is to AFMA the zoom at the focal length you use the most.

Alex

[Colin Southern]

I messed around with this for a while after I convinced myself my struggles with my new 1dx were the cameras fault and not mine. I ended up confusing myself even more and eventually I found out that the 1dx firmware has a bug which causes it to lose all your MF adjustments every time you toggle the af orientation linked option. It has been a popular point on forums. In the end I gave up with it.

Hi Mark,

There is a similar bug that caused the camera to lose the settings if it was powered off (I think I was the one who first reported it to Canon ) - it was fixed with a firmware update they released, so may well have fixed the one you described too? Probably a moot point.

[dje]

Thanks everybody for your comments. Interesting "mixed bag" of opinion/experience.

It's also interesting to hear about firmware bugs on the 1DX. I guess it's a very competitive scene in the camera making business these days with a lot of pressure on the designers at individual companies to keep up with or ahead of the opposition. New models coming out so frequently must mean that there is not a lot of time to fully test and debug.

Dave

[Colin Southern]

Thanks everybody for your comments. Interesting "mixed bag" of opinion/experience.

It's also interesting to hear about firmware bugs on the 1DX. I guess it's a very competitive scene in the camera making business these days with a lot of pressure on the designers at individual companies to keep up with or ahead of the opposition. New models coming out so frequently must mean that there is not a lot of time to fully test and debug.

Dave

Hi Dave,

It's probably not so much a case of "not enough time to debug" (it was around a year from announcement to shipping) as it is the fact that testing is only done by a limited number of people and some just "slip through the cracks" (keeping in mind that most of the field testers are probably just using it as they normally would).

The micro adjustment bug is probably a good example -- I doubt that many testers would be using it -- I don't, but it was purely by co-incidence that I was having problems with a lens following repair that I gave micro adjustment a go to see if it would help (it didn't) and discovered the bug. I dropped an eMail to Chuck Westfall (Canon's tech guru) and the bug was news to him at that stage as well (and to make matters worse, it didn't show up on their 1Dx, so an even more obscure bug that may have slipped through the net).

Unfortunately the firmware for these things has got to the point where it's hellishly complicated.