A dream camera with a serious problem AF

First of all: I have withdrawn from the 4th and the 5th star at the center focus point of the camera due to the now also detected AF problems (more extensive testing), but this will immediately return with much joy when Canon not given a statement and a corresponding has offered solution.

I got my camera October 2013 as one of the first in Austria (not Amazon) and am absolutely delighted with this camera.

Top processed into a top price according to many test reports currently best APS-C camera of the present. Contrary to the EOS 550D, which I had in front of it, this camera stands out due to its sensor and its chip, the top images up to ISO 6400 (about the noise even more visible is then understandably) allowed, has a swiveling touch screen and incredibly fast at continuous shooting (7B / s!) is.

Also great are such nice features as a mode named "Backlit Low" (in truth a computed in the camera HDR) mode, or miniature recording. With the Wi-Fi I start personally not much, because it is just too slow to transfer lots of data, and therefore use almost exclusively the good old USB cable. Overall, for me a top camera for a great price.

And now for the aforementioned AF problems:

I've also tested with 4 lenses. And with
a.) a Canon 50mm 1.8,
b.) a Sigma 150mm 2.8,
c.) with the kit lens 18-135mm and
d.) with the Sigma 30mm 1.4
each at maximum aperture.

Test structure (similar to the YouTube video shown): camera on tripod (partly also freehand.) Each with at least 10 shots at full aperture per lens on
a.) the eyes of my youngest from about 1m indoor in low to moderate light,
b.) a 5m remote photo frames with built structure in low light indoor (open aperture, ISO 400, shutter speed between 1/80 and 1 / 125s) and
c.) a 15m distant aviary (outdoor) with very good lighting conditions

Result in the selected center focus point through the viewfinder: Indoor (ie in low light conditions) unfortunately disappointing as more than 50% of the photos were not perfectly sharp. A comparison with images that have been triggered via LiveView and using a different focal point via the viewfinder, showed strong differences.
It is also striking how incredibly sharp images using LiveView be.

Result in selected focus point right next to the central through the viewfinder: very good to perfect. From about 20 photos of Focus 1x sat here nurn not optimal (but what I attribute to the lens or to myself)

Result in lieveview (the eye allows the advantage of the zoom, which, for example a very precise focus, which is on the viewfinder generally not possible): Just perfect. Takes longer, but meets reliability and enables extremely accurate focusing, which is never a viewfinder as possible.

For me, these tests therefore following emerged:
a.) in aperture values> 2.8 this misbehavior is not reproducible. Only very large aperture values ​​less than 2 are affected (or at greater depth the problem is not just visible).
b.) Stark striking is to focusing can this error behavior only when really poor lighting conditions and with apparently difficult to measure focal points, which could also be seen that the live view mode in difficult situations often took a long time and in some cases several attempts (and thereafter actually to deliver crystal-clear images) very rapidly during the AF point when looking through the viewfinder loud beep - was recognized - but unfortunately incorrect. There is definitely a problem with the central focal point for focusing through the viewfinder.
c.) Other focus points (I use since the finding of misconduct now the AF point to the right of the central) show these errors do not (or at least loud Forums far not as strong)
d.) superior and impresses with an extra sharp photos in almost all situations, however, slightly longer focus search The new measuring method is the live view of the conventional measurement method apparently strong

In this respect: Yes - the problem of AF mittlerern AF point exists unfortunately in fact contradict other observations and resist my first review. Whether that's in all series so of course I do not know, but we must note that it is not easy to reproduce this error. Dim light and the lack of structures, as I had to watch as when recording the Bilderrahmes but even when shooting of the human eye (full format recording of the face and push on the eye), but unfortunately, favor the occurrence of incorrect focus.
That Canon will repair here must, therefore, which is probably out of the question.

Still, I'm saying: Has the problem with the Canon AF first released, so you will probably find currently in the upscale APS-C middle class hardly better.

for AF problem a current video: [...]