So is a bug better or worse than a design flaw ?
I really can't see how Canon can release this lens onto the market. It's got such a serious design flaw that it's not fit for purpose. It should be the ideal lens for me to go with the 70-200 f4L Is but as it is I will stick with the slightly heavier but less sharp 24-105 f4L.
Thanks BC.
I have the 5DII, same as Klaus. I'll check the lens out with 6D users over at DPR. Also, I think DPR will be testing the lens soon on a 6D, which seems to be their current Canon FF body used for testing. No doubt they'll have read Klaus' review and will look into the 'problem'. They have also commented in the recent 24-70 f2.8L II review that the AF was fast and very accurate with the 6D and a bit hit-and-miss with the 650D. So maybe the AF accuracy is body dependent. I'd be very surprised if RSAs were body dependent though.
Just to mention once more - the RSA test was done with deactivated AF.
The AF is fine at max. aperture.
Are you saying that if one uses AF, then the focus shift will not be a problem? i.e. that the camera will correct it in some way?
Crystal ball stuff-
Do you think RSA issue likely in all copies, at least to some build date?
Is Canon likely to correct issue, or is it destine to be forever 'feature'?
(EOS newbie with no EOS glass between 22 and 70mm.)