Ok, I've merged the files. This time the results are better. But you gave me underexposed pictures (next time use the liveview histogram to adjust exposure, not spot AF in the viewfinder). Nevertheless, I was still able to analyze the pictures, so no need to remake the measurements.
Like I said in the PM, the good news is that the sensitive areas are well centered horizontally, the bad is that is seems slightly shifted towards the top.
The center point between slow (f/5.6) and fast lens (f/1.8) is almost identical, so the test can be used to check for misalignment somewhere in the optical path to the AF chip.
You still haven't answered my question : during your everyday camera use, did you notice any AF issue with the centerpoint and fast lenses ? If you don't then it might not be worth sending your camera for repair. In my case, the AF sensor was decentered both vertically and horizontally, maybe that this misalignment is more problematic.
For comparison this is what I got on second repair (compared to me you're a spoiled brat) :
In the case where you decide to send the camera for repair : I've already tried to mention the AF issue by showing my colored rectangles and speaking of misaligned AF sensor.
That approach didn't get my camera fixed on the first two tries, the technician might have though that I got some misaligned focus screen and didn't fix the issue (it even made it worse).
Seeing that some people on Flickr (
https://www.flickr.com …/72157639283256254/page2/
) report that their camera are repaired,
I wouldn't mention the measurements I made, and only report a similar issue as in the first post on the AF issue thread : problem mostly with the center AF point, fast lenses and distant subject.Take some test shots to leave them on a memory stick. In my case, I found that at night the shadows under streetlights and dark surroundings works great to spot the issue (but select situations not too dim for the AF). I simply took 5-10 pictures for each focus point. Use a tripod, for each focus point recompose the picture to always have the same target in the viewfinder's marks and always turn the focus ring to infinity before snapping the picture. In my case I also noticed some trouble with my focus point N°2, but I thinks it's because my sensitive area is almost outside the VF marks so the camera couldn't focus where I was aiming at.
I'm currently testing that second approach for the third repair, hope this will work... Maybe Canon has issued a service note to their repair centers with precisely those keywords, I don't know...
Lastly, for those who still don't understand the purpose of the test read this :The main purpose of the test is to see where the camera can focus when using one AF box. But first you need to understand the distinction between the Liveview marks and the viewfinder marks, both are different :
- Liveview marks I'm speaking of the ones visible with the AF Quick mode (the marks are also visible in Digital Photo Professional by pressing Alt+L). These marks show the ideal position and they are perfectly centered on the sensor.
In order to see if there’s a misaligned AF sensor you should look at only these marks.- Viewfinder marks are on the focus screen;
unfortunately they are subject to some degree of misalignment but this doesn't impact the AF performance, it only renders framing more complicated, because you have to consider both AF sensor and focus screen misalignment when framing through the viewfinder.
For example, look at my post
https://photography-on-the.net …p=17088733&postcount=1248 . I show on the first image on the left : in red Liveview’s marks, in green viewfinder’s marks; on the next picture for Scoobert's camera on the right : In red Liveview’s marks, in black Viewfinder’s marks. You’ll notice some misalignment. Same thing for my 60D (
https://photography-on-the.net …p=17089384&postcount=1252). That sort of misalignment is normal but makes framing harder. Canon can adjust it a tiny bit but not much.
You need to understand that this mark alignment is completely separate from the AF misalignment issue. You should in first place care about the misalignment between Liveview marks and the sensitive areas.
How does this impact you in daily usage ? Well, when framing through the viewfinder, if you're trying to focus on some small high contrast subject in front of a busy background (ex: bird in a tree), you'll need to be extra careful on how you frame it. If you use Liveview marks (in AF QUICK mode), and if you put the subject you want to focus on in the right half of point N°3, you'll focus on the background; if you put the object on the left half of the focus point N°3, you'll focus on the subject. I spoke of Liveview marks here, to see how misaligned the focus screen is just follow the “To check the focus screen alignment” instructions :
https://photography-on-the.net …p=17046799&postcount=1088Here's another good example of a situation where that sort of knowledge is useful (taken with my 60D, 50% crop, better picture full res here :
http://s1262.photobucket.com …ssue_zpsd09fde73.jpg.html) :
Now compare where the AF sensitive areas are on my 60D :
To explain the problem :
- on the left , the black shadow on top of the drawing ended barely touching the sensitive zone of the used focus point, this gave me blur.
- but on the right, I just moved a few millimeters to have the black edge inside the sensitive zone, BAM ! sharp picture...
Do note that in practice, when I was aiming through the viewfinder, I had to put the black line on the right edge of the rectangular mark (represented here in green) ; that's because my focus screen is slightly misaligned, in other words, I have some parallax error.
Indeed, AF is an art !Now does everyone understand the purpose of the test ? Good thing is that these AF sensitive areas also happen to give an idea of the AF sensor misalignment.
So that may have contributed to my strange results.
