chippy_jersey wrote:
OK here's the 2nd test frorm today, nothing creative, just AF tests.
First set, the fella on the bike was single centre AF point with surrounding AF points assist. Looks like it went fractionally off as the pole crossed the frame, but could have been me as it was haldheld.
It may have went off focus as the pole crossed the frame because III-4 was set to 0.
lidesun wrote:
The last 1D Mark IV was in stock was Jan 28, and the last Nikon D3s was in stock was Jan 27, doest it mean Nikon D3s having any problem?
It just doesn't make any sense
True but I asked because Canon is shipping with new Firmware and not sure about Nikon, don't follow Nikon.
FWIW, I went out and photographed oncoming cars driving at about 60km/h with my 400 2.8 @ 2.8 and the results were better than I had with the 600 f/4 IS and the skiers on the weekend. I opted for the 400 today because I know it's sharp and I'm somewhat suspicious about my 600 wide open.
I made it a point to set the main AF related CFn. III settings to the defaults, then went through a series of sets where I changed just one setting, then later tried a few series were two settings were changed (primarily III-4-1 and III-8-x).
Cars are certainly different from birds and my feeling is the AF expansion points tempted the camera to wander off the manually selected point more easily because the cars are so large with more potential detail to attract the AF system. My initial impression is that single point without expansion was somewhat better. Based on the subject matter I don't think III-2 had much influence (I did change it around somewhat without noticing a difference). I tried a bunch with ring of fire and it generally worked well, but also caused the greatest OOF blips during sequences.
I still observed, just as with the skier session, that if the camera goes out, it's more likely to be front focused. And every burst had frames that were out. Sometimes it was one, sometimes a string of two or three. Towards the end of the session I tested micro focus adjustments at +5, +10 and +15 as compensation for the perceived front focus tendency. I need to shoot more at each setting to see if there is much of a difference, but +15 was, as one would expect, frequently back focused. However, it was fairly close between +5 and +10. And as I saw with the skier photos, the closer the subject was to the camera, the greater the likelihood that focus was good. The closer photos also reveal much less of a difference at various micro focus settings.
At the moment I'm leaning towards leaving C.Fn. III settings 2, 4 and 8 at the default, at least with larger subjects like people running or cars on the street, etc.
I still need to do quite a bit more testing with fast moving targets (since my most frequent use of the camera is motorsports photography), but I am relieved that today's testing went better than yesterday's.
I decided to restore the settings to the default center-based focus (had tried more of a ROF approach yesterday). Most of the shooting was still at f/2.8 or thereabouts, but I also varied the aperture to test the focusing in a variety of ways.
I used both the 300mm f/2.8 IS, and the 70-200 f/2.8 IS, both hand-held. Yesterday I had trouble even getting non-moving targets to be in clear focus. Today was much better, though the moving targets still need some work (I think I was expecting too much shooting at 2.8 with the dog running toward me).
These are all with NO sharpening in PP. I left them fairly large to try to show focus as much as possible. Full size they are exceptionally sharp.
Count me in as having problems since upgrading to 1.06.
1.04 gave amazing results probably the best ive ever seen for an AF system.
1.06 is a disaster.
On my 800 the AF is ａｌｌ ｏｖｅｒ ｔｈｅ ｐｌａｃｅ．
Ｃｏｎｓｉｄｅｒｉｎｇ ｉｍ ｉｎ Ｊａｐａｎ ｏｎ ａ ｐｈｏｔｏ ｔｒｉｐ ｉｍ ｎｏｔ ｈａｐｐｙ ａｔ ａｌｌ！！
rscheffler wrote:
I still observed, just as with the skier session, that if the camera goes out, it's more likely to be front focused. And every burst had frames that were out.
Doesn't sound like the Mk III to me - I don't recall seeing anyone complain of going downhill from excellent. This sounds like simply faulty firmware; that's a very different kettle of fish from faulty hardware.
I feel for you guys, but my intuition says that this will be corrected, and will prove to have been merely a ripple, compared to the Mk III's tsunami.
My experience after doing quit a bit of shooting this morning is its all my primes from 300 on that i have with me.
300/500 and 800.
100-400 seems much better but not as good as the almost perfect hit rate as before.
One would think, with all this history of the infamous A/F in the 1D3, that Canon would finally see the merit of having a competent shooter grab a Mk4 as it was being packed to ship and come back with a report after a day and night's shooting everything in sight. The latter to be done once or twice a week at least for the first several months of production and at least randomly thereafter.
They don't need Rob Galbraith for this and he wouldn't do this again anyway for the way Canon essentially 'dissed' him for all his effort with the 1D3.
Testing and verifying the performance of hi-value technical devices such as these cameras needs to be done right at the factory for all the obvious reasons. Not least of these is the language and cultural relationship between a tester and the design and production teams that would remove a lot of communication snafus.
So, yeah, this might also require the tester to be multi-cultural along with photographic competence.
Finally this proximity might even make some corner-office types more aware of and likely to keep their products in line with advertised hype......and PRICE!
OK, and if Canon claims to be doing the above already, the ball is still being dropped somewhere!!
Well I went out to St. Marks National Wildlife Refuge today after some test shooting at local parks. Some of the sea gulls I shot at the parks were OOF, but most were dead on.
Got lots of keepers at St. Marks, including a couple of additions to my life time list. Used my 500/f4 and my Sigmonster. As it got later in the day and the setting sun and clouds forced me to go to 6400 I was very pleased with the results. My 1d4 came with 1.06 and I have no complaints. Got a pix of a Yellowbelly Sapsucker in the fading dusk that was a real crowd pleaser with the birders
WELL I have a stupid question...why is Micro Adjustment needed so much on Pro bodies when the 8 different lenses Ive used on my 30D that doesnt have MA
seem to work fine,AI servo works as well as it should with this particular body with the a 100-400 Ive used at an Airshow...
It may very well be overused (and/or not the answer to a firmware issue). At least according to the respected author of the guide mentioned in this recent thread, and Canon, it's not supposed to be used routinely and has the potential of screwing things up.
digitalbug30d wrote:
WELL I have a stupid question...why is Micro Adjustment needed so much on Pro bodies when the 8 different lenses Ive used on my 30D that doesnt have MA
seem to work fine,AI servo works as well as it should with this particular body with the a 100-400 Ive used at an Airshow...
MA starts becoming more critical when you use lenses with narrow DOF. A 500 @ f/4 is pretty narrow DOF for example
I never needed MA when I used my 17-50 f/2.8 and 70-200 f/4L IS and 12-24 f/4 on my 40D, but when I upgraded to 35 f/1.4 and 85 f/1.2, these lenses are often unusable out of the box without MA or calibration
