Re: Focus problems with the Canon 7DII? (yes, controlled example tests included)
Part 1
Found this, and similar, thread after having concluded my own 7D II is so unreliable in focus, I have to return it for a fix / swap. Tried ignoring it to start with, to maintain satisfaction of the new workhorse, but not possible anymore, as it sometimes lets me down big time, both for pro work and leisure. Frustrating to work with, and the chimping gets out of control when not relying on basic AF anymore.
Luckily I see some have had succes with a new copy of this body, so I cross my fingers, as everything else is nice about it, except that current copy is useless...
But first I wanted to document it here, and for the shop I return it to. A pain, but hopefully useful for others.
I have done some thousands of shots now, covering events with the newly aquired 7D II now included, and have noticed unacceptable focus inconsistencies unlike other used bodies like 5D III and 50D backup (to include another crop sensor body for tests with same EF-S lenses). Lots of it trivial AF work with one-shot, point focus (surrounding also tried, with no more luck).
Tried it a little locally also with birds (BIF) to stress test the claimed superb AI servo tracking. But focus keeper rates across usage has also here been lower than expected - although focus point tracking seems very responsive and promising! If fixed or new body fixes inconsistency it looks like a 100-400 II needs to be coupled with it too
More controlled testing has been done for documentation and verification that the body indeed has a temperamental AF. The very basics: Static scene, tripod, one shot, center focus, focus priority, RAW, tests that should of course be no issue to get consistently in sharp focus.
Primary lenses have been 17-55mm f2.8 IS and 70-200 f4 L IS. Both potentially capable of some nice sharp results if AF is accurate. AFMA adjusted as good as the focus inconsistency has allowed. FoCal and own double checks made. Adjustments seem valid enough. 17-55 has +6/+7, 70-200 is at +9 for tested 200mm IIRC.
1)
First example set is with 17-55mm, f2.8 at around 2 meters. Static scene, tripod, repeated "one shot"s, center focus, focus priority, no or infinity defocus after each shot. Head to head against the old and more trusted 50D (same setup, slightly lower exposure though).
7D II:
1/80, f2.8, ISO 640, 55mm
Optimal focus plane (live view / contrast detect):
_M9A0109_reference.jpg
Then phase detect:
7 out of 22 is in focus, example:
_M9A0141_sharp.jpg
15 out of 22 were not in focus(!), examples:
_M9A0140_unsharp.jpg
_M9A0142_unsharp.jpg
Focus plane all over the place. I can sometimes hear the AF readjust, and see the distance scaling jumping a little nervously, sometimes obviously ending in a new position, even though scene and camera are static. No consistency, and impossible to rely on.
50D:
1/125, f2.8, ISO 640, 55mm
Optimal focus plane (live view / contrast detect):
_MG_7401_50D_reference.jpg
Phase detect:
Around 26 of 28 are in focus, example:
_MG_7404_50D_sharp.jpg
The few debatable look like this, worst case:
_MG_7403_50D_less_sharp.jpg
_MG_7418_50D_less_sharp.jpg
That’s how I expect it to be for such a trivial job. Just nailing them (except the two almost there). Not a single one close to how unpredictable my 7D II can be at times. 50D has way less variance and nice accuracy. Not once did I notice a refocus/readjustment attempt as I observed the 7D II sometimes did at same setup.
I will note that the runs above for the 7D II were particularly bad. I have normally not seen _that_ many way out of focus, but often still way too many, and when it’s in a particularly bad mood it can go all crazy like shown here.
Ironically I bought the 7D II for it's claimed much improved AF, so having it fail that much compared to the old 50D is not good. Is the 50D always perfect? Not at all, that's why I got this newer crop body. But no luck on this one.
Part 2 coming up with FoCal consistency tests and comments (need split to more than 8 files, trying to avoid zips).
Anyway, there are so many possible condition out there that the camera AF will face, not surprise if there are certain condition that the testing procedure miss that will cause the AF system to be confuse.
