Took some landscape shots recently and noticed that the horizon was consistently tilted down on the right hand side. Further tests showed a significant discrepancy between the finder and the rear LCD. Eyeballing I'd say about 3 degrees.
Somewhat disappointing for a flagship camera.
Perhaps I shouldn't be surprised, my 1DsII had a shutter failure after a few thousand clicks and had to be replaced.
For me it's back to the DMR until Canon has solved the problem.

wouldnt it be the OVF prism thats tilted instead?
which is an adjustable component

Does it make any difference what component it is? If I were to buy an $8000 camera I would expect it to be near perfect. 3 degress isn't even close. How could Canon let that out of the factory?

Rob Riley wrote:
wouldnt it be the OVF prism thats tilted instead?
which is an adjustable component

Rob Riley wrote:
wouldnt it be the OVF prism thats tilted instead?
which is an adjustable component

Possibly. For the end user it's hard to say. But I find it unlikely that the groundglass per se is rotated unless there's a *lot* of play in the mount. It has the original focusing screen and I haven't touched it. Manual focusing btw. is spot on.

pdmphoto wrote:
Does it make any difference what component it is? If I were to buy an $8000 camera I would expect it to be near perfect. 3 degress isn't even close. How could Canon let that out of the factory?



Paul, I'm wondering about that too. A single exposure against a horizontal line is all it takes and the fault will be obvious. But such a simple test before they ship a $8000 camera seems too much to ask for. Makes me somewhat uneasy as to what else might turn up in the coming months.

photoArne wrote:
Took some landscape shots recently and noticed that the horizon was consistently tilted down on the right hand side. Further tests showed a significant discrepancy between the finder and the rear LCD. Eyeballing I'd say about 3 degrees.
Somewhat disappointing for a flagship camera.
Perhaps I shouldn't be surprised, my 1DsII had a shutter failure after a few thousand clicks and had to be replaced.
For me it's back to the DMR until Canon has solved the problem.


You are not alone in having this problem, take a look at the link below

http://www.northlight-images.co.uk/article_pages/cameras/canon_1ds3_tilt.html

There was also a piece in the BJP

http://www.bjp-online.com/public/showPage.html?page=701396

I have read in a number of places that it is a recognised fault with some 1DsIII's unfortunately Arne. Sorry to hear you've got a duff one, a bit disappointing after spending that amount on the top-of-the-range body.

One would have thought that after the 1DIII autofocus problems, Canon would have made absolutely certain about the quality of this one. Clearly their QC department needs a major kick up the Kyber.

Off topic but that BJP article surprises me. The quality of their journalism appears extremely weak with that unattributed statement in the last paragraph that Canon has found the root cause of the 1D3's focus problems and appear to be working on a new fix. Without attribution that is just blatant irresponsibility. As it stands they have just dropped a bomb and left it hanging for people to run with as a rumor.

asnapper: Thanks for the links; yes, this agrees very well with what I've found. If anything my copy is a bit worse than the samples posted.

Mine only had a .25 degree tilt. When I had to send the camera into Canon for a cleaning of the sensor, I asked that they fix it. They did so and I had the camera back in one week including shipping both ways. I was pleased.

From what I have read, other cameras have had this issue, but it does seem that the DSIII has more reports of the issue than any other camera. Possibly, this is because of the more demanding nature of the users of this camera and, possibly, due to the internet making others aware of the issue and therefore have them check for it. I am pretty sure that I would not have noticed the .25 degree discrepancy had I not actually tested for it. Even then, I had to do the test carefully to be sure. Now, 1 degree or more is clearly very noticeable and grossly unacceptable--especially for a "flagship" camera. The good news is that it seems to be an easily fixable issue.

Luvwine wrote:
Mine only had a .25 degree tilt. When I had to send the camera into Canon for a cleaning of the sensor, I asked that they fix it. They did so and I had the camera back in one week including shipping both ways. I was pleased.

From what I have read, other cameras have had this issue, but it does seem that the DSIII has more reports of the issue than any other camera. Possibly, this is because of the more demanding nature of the users of this camera and, possibly, due to the internet making others aware of the issue and therefore have them check for it. I am pretty sure that I would not have noticed the .25 degree discrepancy had I not actually tested for it. Even then, I had to do the test carefully to be sure. Now, 1 degree or more is clearly very noticeable and grossly unacceptable--especially for a "flagship" camera. The good news is that it seems to be an easily fixable issue.

That's good to hear, because in every (other) respect the IIIs is really a joy to use and a significant improvement over the 1DsII.

To finish this issue. Just got my 1DsIII back today. According to the service report they dismanteled and adjusted the viewfinder and exchanged a flexcontact (whatever that might be, reason not specified). At first glance this would indicate that the viewfinderf was at fault, but in reality it might just as well be the sensor that's tilted and the viewfinder adjusted to align with the sensor. Anyways, initial tests show a quite good ageement between viewfinder and capture, so for practical purposes the problem is solved.

Well, for practical reasons, this would mean that you will not be able to use a bubble-level in the future and will have to rely on eye-sight to align the tilted viewfinder with an upright scene so that it aligns with a tilted sensor

edwardkaraa wrote:
Well, for practical reasons, this would mean that you will not be able to use a bubble-level in the future and will have to rely on eye-sight to align the tilted viewfinder with an upright scene so that it aligns with a tilted sensor

Good point, might look up my bubble-level to find out. OTOH I hardly ever need the b-l to achieve level images. Markings in the viewfinder and eye-sight is usually suficcient.

I had mine re-aligned but before I had the time to send it in I just used live view to line up the shot(in fact without live view is why most of us have been suffering out of alignment cameras for years.)