Upload & Sell: Off
So last weekend I played around with lens MA, partly to see if it even mattered for my camera/lenses...
1. Printed out test charts available free online (e.g. http://www.bealecorner.org/red/test-patterns/ISO_12233-reschart.pdf) and following Canon's recommendations (e.g. http://www.pixel-shooter.com/forum/showthread.php?t=21099)...
2. Experimented with DIY lens microadjustment suggestions, like http://regex.info/blog/photo-tech/focus-chart
3. Tried arash_hazeghi's trick using Canon's EOS utility, a tethered setup, and live view (http://www.birdphotographers.net/forums/showthread.php/58042-AF-microadjustment-tricks)
4. Variation of #3 above, but taking shots after each < or > microadjustment in EOS utility's live view window
Below are some observations based on 24-105L (on a 5D2)...
For an f/4 lens, doing microadjustment _without_ a "45 degree ruler" approach is very very difficult since there are at least 5-10 adjacent microadjustment settings where the sharpness is indistinguishable. This is especially true if you follow Canon's official recommendation of using a target distance of 50x the focal length.
For example, the smallest DOF for the 24-105L with the 50x recommendation is 1.75 ft at 105mm (and f/4) with a target distance of 16.4 ft. 2 ft of DOF means a _lot_ of identically sharp images you'll likely see with various MA's (even at 100% crop).
So approaches #3 and #4 simply didn't work for me. Maybe they'd work with a faster lens...
Then I tried #2, even though Canon recommends the target plane must be parallel to the sensor plane. And for now, I also resorted to using close to MFD to reduce the DOF as much as possible. Success finally! At f/4, 105mm, and a target distance of a little over 2 ft, the MA was +8 for the 24-105L.
- Has anyone had a "better" experience with an f/4 like the 24-105L using DIY approaches?
- Why does the target plane need to be parallel to the sensor plane (per Canon)? Is it just to get reproducible AF results?
- Why does Canon's instructions suggest using the longest focal length (of a zoom lens), then also recommend using the "most commonly used" focal length? (And if there's such a thing as a most commonly used FL, wouldn't everyone just use a prime at that FL instead? :-))
- Why does the target need to be 50x away? (25x away according to some non-Canon sources, e.g. LensAlign.) What if MFD is used? Shouldn't we pick the "most commonly used" target distance instead? :-)
- Even with "perfect" micro adjustment, if there are differences in the MA at different focal lengths, what setting does one pick?
I can see that MA matters most with a shallow DOF (fast and/or long lens), or if the lens is badly misaligned (for the camera).
I've decided to order LensAlign Mk II and wait for it before checking MA on my other lenses. (FoCal is not for me, since the more hands-on the better. :-))
My approach with LensAlign will be to try and measure the MA at different focal lengths (e.g. 24, 50, 70, 105mm for the 24-105L) and distances of MFD/25x/50x. If the MA is very different, I'll lean towards using the setting for the longest focal length (which I _will_ use) and the "smallest commonly used" :-) distance, since the DOF will be the shallowest there, and so will matter the most there.
What has been your experience with lens MA?
Edited on Jun 13, 2012 at 07:42 AM · View previous versions