Jeff
Offline
• • • • • •
Upload & Sell: On
|
 p.39 #16 · ['NEW Fix' UPDATE!] - MkIII AF still broken | |
inness wrote:
You failed to notice that he had said that IS was turned of, so why should we think that anything you said makes sense.
I have asked other people to tell me what is the best test, can you tell us what is the best way to test this camera.
All the good people on this forum who present us with their photos and tests, seem to get knocked on the head by people quoting reasons why the test is rubbish, so instead of knocking other peoples efforts to test, please put forward your test and photos.
Bob
I'll be honest with you, now that most of the AI Servo/AF issues revolving around the submirror have been fixed, it's much more difficult to accurately and specifically 'test' the camera. However, even the submirror 'fix' isn't fixing some cameras, and there are apparently Blue Dot cameras that have lingering issues as well (my current BD does not in relation to the submirror, though my first camera still didn't work after having gone back to Canon twice). You really need to go out and use it in a variety of conditions, with a variety of lenses (especially f/2.8 lenses), and scrutinize the images in relation to the results you feel you should have gotten given the techniques that you used. Examining the files at 100% view is required for this task, and it takes time and energy.
There are still fairly widespread reports of 'occasional' OOF images for no good reason, with no apparent plane of focus, nor any indication of front- or back-focusing. Also look for 'ghosting', or a sort of secondary image registration in one plane that looks like camera movement, but it is not (I've seen it at 1/8000th; see Rob G's most recent 'fast runner' series for examples, or the first page of this thread). I have not seen this 'ghosting' with my new Blue Dot, but I've only had the opportunity to put 400 images through it, mostly AI Servo testing (BTW, a runner coming at the camera using an f/2.8 telephoto lens is best for this test; center-point only, no expansion, default AI Tracking Sensitivity, 10 FPS). Lastly, the 'focal plane' (also referred to by some as 'fine detail smearing'?) issues that are also described in this thread are noted on Blue Dot models, mine included, thought to what extent I remain unsure, as I haven't been able to do much 'landscape' style testing, which highlights this issue rather well.
For what it's worth, many former users of MkII/n's simply say that the MkIII just doesn't quite live up to the performance of the MkIIn, and by Canon's own admission, it should have. I admit we're starting to split hairs here, because many people don't initially notice issues with their camera, especially if they use it for only certain types of shooting. My Blue Dot version is so much better than my early-production model that I am impressed, but the inconsistency in some of the results described above is hard to nail down without shooting a lot, using various techniques and lenses.
I know that probably isn't as definitive as what you'd hoped for, but that's all I've got.
-Jeff
PS: It will be interesting if someone's post about a 'harmonic balancer' being required for the fix ends up being true, as it would suggest that there is some sort of vibration issue going on...
|
