Just to be clear we're talking about complete focus failure (Weakness 1).
...
I have a couple of buts though ...
Sometimes the Small Spot is entirely on the bird and there's no more distant point to focus on.
Some of the completely failed focus shots' EXIF recorded no AF point but what's in the distance looks sharp or sharpish. So there may be an infinity default (when in doubt, rack it out).
...
There is another aspect of out-of-focus features that can help confuse things. When objects are out-of-focus they get larger. The blurred image is much bigger than the in-focus, sharp one. Obviously, the af points themselves have a finite physical size but they may not be the same size as the green box shown on the screen. So out-of-focus objects that are actually shown to be outside of the green box may actually be "seen" by the af point. So it could be that a background feature is used even when it slightly outside the af area because when the af cycle started, the blurred image of it actually was inside the af point boundary (wherever that is).
I've seen lots of past demostrations of cameras focusing outside of there apparant af point boundaries so I wouldn't be surprised to see it on Sonys too.
Let me know how you find the zoom out technique if you try it.
There is another aspect of out-of-focus features that can help confuse things. When objects are out-of-focus they get larger. The blurred image is much bigger than the in-focus, sharp one. Obviously, the af points themselves have a finite physical size but they may not be the same size as the green box shown on the screen. So out-of-focus objects that are actually shown to be outside of the green box may actually be "seen" by the af point. So it could be that a background feature is used even when it slightly outside the af area because when the af cycle started, the blurred image of it actually was inside the af point boundary (wherever that is).
I've seen lots of past demostrations of cameras focusing outside of there apparant af point boundaries so I wouldn't be surprised to see it on Sonys too.
Let me know how you find the zoom out technique if you try it.
I have seen many times where the Sony's "see" outside of the displayed AF Zone. I've had to switch to a smaller focus group in order to force focus to drive to a grossly OOF near target but with a wider group selected it kept focusing on background even though it was just a bit outside the Zone.
IMO the only real weakness I have seen on the A9 at least is that it does not handle backlighting very well. It will definitely miss a lot of shots if the subject is strongly backlit. It is much worse than my Nikon D500 in that regard. I have experienced it at airshows and with shore birds like sooty oystercatchers regularly. Overall it's got fantastic AF. For the A9III I will be disappointed if we don't get x-type AF sensors.
Pixel Perfect wrote:
IMO the only real weakness I have seen on the A9 at least is that it does not handle backlighting very well. It will definitely miss a lot of shots if the subject is strongly backlit. It is much worse than my Nikon D500 in that regard. I have experienced it at airshows and with shore birds like sooty oystercatchers regularly. Overall it's got fantastic AF. For the A9III I will be disappointed if we don't get x-type AF sensors.
I think that is subject dependent. All our couple shots and wedding stuff is shot against the light and there is NO issue at all. The A9/A9ii finds focus perfectly (the A7riv has bigger problems, but still works fine). That may be different with BIF, but for human beings I don't see any problems compared to all other cameras I used before.
"I haven't seen God so s/he does not exist" doesn't wash in empirical science unless I've looked in all corners of the cosmos. The looking and not finding is evidence. The not-looking can't produce evidence.
People here have said they haven't experienced Weakness 1, 2, 3 or 4. Fine. To be expected if they weren't looking or they were shooting with different subject matter or settings.
Mastering your tool is learning what it does well and what it doesn't.
It's dead easy if you want to master these tools to test for 1, 3 and 4. Can be done in minutes. The most revealing one is to set your camera to AF priority for some BIF bursts.
So far in all the discussions I've read about 1 before this thread a range of settings to try have been thrown up by users, and I've tried them and found little or no improvement. Having taken 50,000+ bird shots with 2 Sony bodies and 3 Sony tele lenses I came to the conclusion that there's a systemic problem.
somersettr has described what that might be and what his workaround is which is similar to mine.
I've experienced weaknesses 1, 2 and 3 with various cameras including the A9II but I suppose I have just put them down to the technological limits and physics of the systems. The A9II is the best camera I have used for af.
I get most problems with small, fast-moving birds, even against a clear sky. If the bird is too blurred when I first half-press the shutter button, and the AF racks the wrong way to try and find something, then you are lost because you can't track the bird as it has become too blurred to even see. Even if the AF racks the right way so that the pdaf can start to work properly, you still need to keep the bird under the af area for long enough to acheive a good lock.
I have noticed a peculiarity on the A9II (and probably all Sony bodies) with Wide AF area. That is, although the camera should find a subject anywhere in the frame, it will only acquire focus when the subject is within the centre ninth of the screen. (i.e. if you put a rule of thirds grid up, the subject has to be in the centre rectangle for the camera to acquire focus). Once it has acquired a subject initially, it will then use points anywhere in the whole frame to track it. You can test this by selecting AF-C Wide, point at a clear sky, half-press the shutter and then lower the camera to say a tree to see when the green boxes start to flash.
Not knowing that can cause a lot of missed shots and frustration. This centre rectangle is actually smaller than the Zone af area which means that sometimes it is easier to acquire focus on an erratic subject with Zone than it is with Wide simply because the "acquiring" af area is larger with Zone. But then it's harder to keep the subject within the Zone area than Wide.
I have concluded that, even with the A9II, a lot of practice is required to improve my chances. Pre-focusing, tracking the subject accurately and using the most appropriate af area make all the difference.
Holger wrote:
I think that is subject dependent. All our couple shots and wedding stuff is shot against the light and there is NO issue at all. The A9/A9ii finds focus perfectly (the A7riv has bigger problems, but still works fine). That may be different with BIF, but for human beings I don't see any problems compared to all other cameras I used before.
You may be right, but I don't shoot humans only more genteel species.
Pixel Perfect wrote:
IMO the only real weakness I have seen on the A9 at least is that it does not handle backlighting very well. It will definitely miss a lot of shots if the subject is strongly backlit. It is much worse than my Nikon D500 in that regard. I have experienced it at airshows and with shore birds like sooty oystercatchers regularly.
What are the settings when it fails?
Is it more likely to happen with monochrome subjects?
It appears from the data extracted by A7info.exe that SAF uses contrast detect only, while CAF uses both phase and contrast detect ('hybrid' AF).
That would explain the preferencing of the further lock (weakness #1), and poor performance under low light and low contrast conditions. But weakness #1 is produced by CAF on static shots too. CDAF is focus determined by software so the algorithm is critical.
I've also been testing turning OSS off for my fast shutter speeds and am not seeing any improvement in sharpness rates. This is with the A9, 400/2.8 and both TCs, with AF priority. In good light the A9 was skipping few shots.
I have experienced weaknesses 2 and 3 on A9ii and 200-600mm, it is real and what is amazes me it happens in a good light. I will also add it is not rare, at least with my camera/lens. I see it every time I shot BIF. I thought it is me but I see a lot of people have.seen it.
osv2 wrote:
a7info lists "PhaseAF Locations", per the button at the top left of the screen.
i don't see anything labeled as hybrid af.
af-s uses hybrid af, it's not cdaf-only, unless the light is bad or the lens is stopped way down.
This is how I read the data...
It shows where the focal plane was (the circle) and what, if any, PD points were used in setting it (the squares).
SAF shots show no squares.
Ziggy99 wrote:
Interesting observation by Ianiro... OSS 1 on the 200-600 & A7R IV is producing soft images. 2 isn't.
Yep, that is what I've been telling people for months now...heck maybe almost a year. OSS 1 is the main factor in the A7RIV/200-600 AF saga. I find OSS 3 or OFF are best.
tester_V wrote:
"OSS 1 is the main factor " I do not use OSS1 with A9ii and the problem is there...
I was talking about the specific issues with the A7RIV/200-600. If I use OSS 1 with that setup then I get back focus and very inconsistent focus. If I use OSS 3 or turn OSS OFF I get much better results.
With my A9II I use OSS 1 all the time. Occasionally turn OSS OFF for swallows IF at 1/3200 or faster. But I have no big issues with my A9 or A9II.
As to the original set of 5 issues. I do see Issue #1...that plagues all MILCs. I don't have issues 2 nor 3. I can replicate #5 so I agree with that.
