Steve Crist , Jul 02, 2009; 07:40 p.m.
I want to do the micro-adjustment with all my lenses. If you have a zoom lens what focal length do you set the zoom for? Thanks.
Enrique Bocanegra , Jul 02, 2009; 08:55 p.m.
I was excited when I played around with this new feature on my 5d2, I learned to leave it alone because it makes things worst. I corrected my ef 50mm f1.2L to be dead on with shots in close range wide open, but when I took shots of objects much further away, it was very blurry. I had the correction to +15 and set back to disable. When I worked on my 24-105, same thing. At 105 it was corrected for close ups, and messed up everything else at long shots. At the widest it was all blurry. That lens had correction of +10. Set everything back to disable. Maybe I need to learn how to use properly. It took me about two hours for each lens of taking and retaking shots. Only to go back to disable. I will follow this one if someone has some good tips on how to test and set properly. Thanks
Michael Liczbanski 
, Jul 02, 2009; 09:18 p.m.
Start here (the author works for Canon USA.)
Enrique Bocanegra , Jul 02, 2009; 09:38 p.m.
thanks Michael, I'll give it a second chance.
Scott Ferris
, Jul 02, 2009; 10:53 p.m.
This is good too, (link) it has a downloadable screen target that relies on Moire effect to achieve accurate focus.
Steve Canon say you should do zooms at their longest focal length, so 200 for a 70-200.
Robin Sibson
, Jul 03, 2009; 02:43 a.m.
Enrique's problem shows exactly why in-camera micro-adjustment on a per-lens basis is treating the symptom not the cause: whatever correction you make, it is good only for the distance, and, in the case of a zoom, the focal length, at which it is done. Whilst it may also improve accuracy at other settings, there is no guarantee of this and it can make things worse.
My understanding is that when Canon calibrate a lens, they install a new lookup table into the firmware of the lens with enough "control points" at different distances (and focal lengths where relevant) to give suitably accurate performance under all circumstances. Does anyone know, is this REALLY what happens, or is it simply a description of how it might work in an ideal world?
Mark Kissel 
, Jul 03, 2009; 07:14 a.m.
I don't know the answer to that one, Robin. So you think the micro-adjust is shifting the lookup table values? Or just the value at that specific focal length?
I have a 24-70 that I adjusted and it is absolutely spot on now; and it appears to be across the focal range. It was always good, but it didn't perform quite like that before. I find fewer missed focus shots with that lens now at any focal length/distance.
I also wonder how much variability there is in the acceptable deviation. If you intentially adjust a lens to out-of-focus twenty times, and let it re-acquire autofocus, how much does it vary?
Jack Nordine , Jul 03, 2009; 01:36 p.m.
I've read far more people praising this feature than the other way around. I wish I had it on my XSI. I can understand why it may might not be a perfect fix for zooms, but I would think with primes it would work perfectly.
Alec Myers
, Jul 03, 2009; 02:43 p.m.
Does anyone know, is this REALLY what happens,
I believe that is the case. I do know (can't find a source) that factory (re-)calibration involves more than setting a single number.
Steve Crist , Jul 03, 2009; 10:13 p.m.
Thanks for the help everyone.