« on: March 24, 2013, 05:10:38 PM »
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Sorry it's confusing but the camera model that shows up beneath our names has nothing to do with what we own. I own a 5DII and a micro 4/3 camera. Before that I had an old Rebel (XS).
The standard deviation and the "tolerance range" in number of AFMA units are not fixed quantities -- variability of the AF will depend on things like subject distance, and available lighting (either ambient or flash/ AF assist). Then the range for which an object appears to be in focus could be 2 or 3 or 5 standard deviations depending on depth of field.
Roger at Lens Rentals had some tests where he looked at the variability of autofocus of some different bodies (including the 5DIII) with some different lenses. My recollection was that with many tested lenses, phase detection on the 5DIII is more consistent than older bodies. Contrast detect (used in live view and on mirrorless bodies) is more accurate but slower. I'd suggest searching lensrentals for the article if you are intersted, it was a good read.
I don't think it's really necessary on Rebels since the focusing on them tends to be inconsistent anyway (at least that was the case with my old T2i). AFMA wouldn't fix that.
I wouldn't say pointless. I had a zoom that on my 7D needed +3 at the wide end and +7 at the long end. Leaving it at zero wouldn't be optimal, obviously. Generally, the DoF will be shallower at the long end. In that case, +6 was the compromise value.
Having the two settings is nice, though. The most recent zoom I tested on my 1D X needs 0 at the wide end and +5 at the long end. But I can imagine that sometimes two settings wouldn't be enough. The camera does a simple linear regression with focal length between the W and T values. For that lens, the two intermediate focal lengths I tested had AFMA values that fall right on that line. If they hadn't, the lens would have gone back.