skitron said:Craig Richardson said:EDIT 2: Lens is completely unpredictable, would not recommend at this point.
One thing I noticed about mine was if the camera is not steady (i.e. good handhold technique) I will often get both motion blur and OOF. It's as if the AF can't deal with much camera motion and still do its job. Only an issue at low aperture values.
Another thing I'm going to try even though I'm having good luck with this lens is shooting in AI servo mode when at low aperture values due to the very thin DOF. Several folks are recommending this in the Canon vs Sigma 85mm thread and it makes alot of sense. May be worth a try for you.
I can't get my head around how a lens can be so out of focus (like -11 needed to take a picture of the CN Tower because of back focus) but the phase detection can declare it in focus. What exactly did Sigma have to do to reverse engineer this focusing system? I also don't understand how this switches to front focus under 10 feet... or why when I use high quality B+H UV, CPOL, or ND filters the focus problems sometimes reverse themselves. Gentec has alot of work cut out for them when I bring this thing back again to them.
EDIT: I have done some light reading and came up with this gem:
Source: http://www.zen20934.zen.co.uk/photography/Canon%20AF%20System.htmWhen you half-press the shutter release (or the * button, if you've used the custom function to move focusing control there), the activated AF sensor "looks" at the image projected by the lens from two different directions (each line of pixels in the array looks from the opposite direction of the other) and identifies the phase difference of the light from each direction. In one "look," it calculates the distance and direction the lens must be moved to cancel the phase differences. It then commands the lens to move the appropriate distance and direction and stops. It does not "hunt" for a best focus, nor does it take a second look after the lens has moved (it is an "open loop" system).
If the starting point is so far out of focus that the sensor can't identify a phase difference, the camera racks the lens once forward and once backward to find a detectable difference. If it can't find a detectable difference during that motion, it stops.
So what basically happens is that when the camera gives the instruction to the lens to move, the Sigma lens understands that it is a move order but does not move precisely enough. Because there is no second check on focus, errors are left uncorrected. This is why AI-SERVO seems to work better because the camera is constantly looking at the focus and making corrections, it creates a closed loop and errors eventually reach zero, or so you would think!
This also explains the strange pulsing behavior of this lens in servo mode because the camera is expecting the lens to nail focus after the first move order, but it doesn't so therefore the camera thinks the subject must have moved so it calculates a delta for subject shift and comes up with a move solution and tells the lens to execute, which it does but the lens screws up again, and so on and so on.
Another update: AI Servo works really well.
Upvote
0