72chevelle454 Offline Dedicated FM Upload & Sell: Off
|
p.4 #15 · RE-POLL: MkIII have AF issues? | |
I noticed you tested at 8fps, I feel the problem is more noticeable at 10fps (anyone else?).
At 10fps it doesn't seem to give the camera enough time to calculate the subjects progression of speed. (I'm sure theres a technical term for this).
We shot some dogs jumping into water (dock dogs) again yesterday (Both M3's set at 10fps and had a very bad keeper rate, around 100 out of 600+ shots, most all back focused. 
As Ron pointed out earlier in this post, the 300mm f4 lenses gave us trouble on both cameras, with two different lenses.
The keepers
Chad
fraga wrote:
I received mine two days ago and didn't have much time to play with it yet.
So far, I have noticed that:
- AF in low light is MUCH improved over previous 1DM2n. Much faster.
- IQ is also a big improvment. Exposure seems better as well.
- High iso also a step forward, but everyone knows this by now! 
- My copy in single shot mode has no problems whatsoever with AF. AF is dead on everytime.
Finally, I only did one test yet regarding AI-Servo.
Today I selected 8FPS and the other settings were just "out of the box", didn't change any of them. Used a 70-200 f4IS.
The test was done shooting 3 motorcycles going down the street, as they rode in my direction. They were moving slowly, btw.
In each sequence, there was just one or two pictures that were slightly OOF.
Not much, just a tiny bit. In fact, they looked more lke soft pictures rather than OOF. I guess the result resembles a comparison between a really good L with a low-end cheap lens. They are in fact OOF of course, not soft, but you get the idea...
What is my conclusion?
If I didn't own a 1DM2n previously, I would say that my camera is faulty, since the results seem to be somewhat mimic RG's report.
However, I must say that I wouldn't expect much better results from my previous camera. I can't say that EVERY shot taken with it was perfect in terms of focus.
With my 1DM2n, in every sequence I got usually one or two OOF pics, sometimes more. At the time I atributed it to user error, not the camera's fault.
Wish I still had it, to be able to do a direct comparison.
Next Sunday I will shoot a motocross event and will be able to test it properly.
From what I've seen today, my 1DM3 seems slightly better in AI-servo.
The problem is the results are rather similar to RG's report, even if slighty better (only one or two OOF in a full sequence), and that makes me wonder if my 1DM2n was in need of a calibration or if I dind't have it set up properly...
|