Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: Focussing problem?  (Read 4322 times)
jule
Sr. Member
****
Offline Offline

Posts: 738


WWW
« on: September 17, 2005, 08:07:47 PM »
ReplyReply

I have had my 1DSmk2 for a couple of months now and was a little unhappy with the sharpness of my images. I thought it was my inexperience with this camera and unfamiliarity with sharpening required for the 1DSmk2. Yesterday morning I thought I would try what some have suggested to test whether the camera is focussing correctly.

I used a tripod with the camera at approx 45 degrees to a few items which were beside me on the bench.....with a tape measure alongside. I used a Canon 50mm f1.4, centre focus point, auto focus, ACR with default RAW processing and no other adjustments. I then converted to jpeg to post here.  I then used another prime - Sigma 50mm f2.8 DG Macro and a print out of a test grid. Same technique.

I examined them at 100% in RAW, but even in small jpeg it can be seen the focal plane differs from the centre point which was focussed upon. Am I correct in assuming that there is a back focussing issue with my camera, and my camera and all my lenses need to go to Canon for calibration? Before I blame the camera is there anything else I have missed?

Some feedback on my test prints would be appreciated.

I have linked to the images.  

Julie

Image 1

Image 2

Image 3
Logged

jani
Sr. Member
****
Offline Offline

Posts: 1604



WWW
« Reply #1 on: September 18, 2005, 04:49:30 AM »
ReplyReply

Your test is designed to make it difficult to judge focus ability, since the tape measure is not close to the focusing point, and the objects aren't in plane.

In image 1, the focus plane starts at the top of the paint bottle, or possibly slightly above it. It's really difficult to tell, since the rounding of the bottle makes it hard to see where the focusing plane would fall on the tape measure. You should expect that the part of the tape measure being in focus is somewhat in front of the bottle, because you've elevated the camera and tilted it downwards.

In image 2, your camera has focused in front of the object on the paper, but I'm not sure if it's reasonable to expect it to autofocus correctly there.

In image 3, your camera seems to focus okay on the label, but the same effect as in image 1 is of course apparent.

I think you need to redo your tests, but this time with something to focus on that the camera is in plane with.

Example showing clear front-focusing:

20% view
100% crop

As you can see in these images, the business card is in plane with the focus plane, and you can therefore see exactly where the ruler lines should've been in focus.
Logged

Jan
jule
Sr. Member
****
Offline Offline

Posts: 738


WWW
« Reply #2 on: September 18, 2005, 05:55:38 AM »
ReplyReply

Thanks Jan, I will redo tests on Tuesday.
Julie
Logged

jule
Sr. Member
****
Offline Offline

Posts: 738


WWW
« Reply #3 on: September 20, 2005, 07:01:28 PM »
ReplyReply

Jan, Is this a better test? Links to low res images. Focus Test 3, 4, 5 & 6 are crops of images taken at the same distance as Focus test 1, 2 & 7.

Focus Test 1

Focus Test 2

Focus Test 3

Focus Test 4

Focus Test 5

Focus Test 6

Focus Test 7

I used the same proceedure with two lenses. Canon 50 f 1.4 and Sigma 50 f2.8 macro. Examining at 100%, from what I can ascertain, is that it seems to be front focussing with both lenses, in which case may be the explanation as to why all my images appear soft and some just not in focus at all.

I haven't had a problem like this with my previous cameras, but do acknowlege that the extra sensitivity of the 1dsmk2 sensor will illuminate inadequacies in technique, so I kept perservering.

I didn't even consider it to be a camera problem until last week-end, when I noticed that an object on the far left of the image (which was closer to me than the subject) was in focus, when I am using a centre focus point.  The penny dropped then, that perhaps my camera - or lenses - were not focussing properly...hence my test, now hopefully refined and fairly accurate. I would appreciate any advice or comments.

Julie
Logged

Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #4 on: September 20, 2005, 07:49:23 PM »
ReplyReply

Your test shots appear to exhibit reasonably good test technique and a bit of front-focus, so sending your body and lenses in for focus calibration is probably a good idea. But there are two things you should keep in mind:

1. When you start placing quality demands on the 24x36mm format that used to be reserved for medium format, your definition of what's considered "in focus" becomes much more stringent. This means that DOF shrinks proportionately, and you must stop down more than you did with film to get the same DOF at the higher resolution digital offers. With the 1Ds-MkII, it's not unreasonable to expect that DOF will be about half of what you got shooting 35mm film given the same focal length, aperture,  and subject distance. This means that a slight focus miscalibration will probably go unnoticed while shooting film, but will be quite obvious when shooting high-resolution digital.

2. Do your testing at more than one distance, preferably distances you commonly shoot at when using the lens. Your test is conducted at macro range, which is fine for the macro lens, but your 50mm/1.4 may be dead-on at normal portrait distances an only off calibration at the extreme close end of the focus range. If that is the case, re-calibrating it may improve macro focus accuracy at the expense of focus accuracy at/near infinity.
Logged

jule
Sr. Member
****
Offline Offline

Posts: 738


WWW
« Reply #5 on: September 20, 2005, 08:20:06 PM »
ReplyReply

Thanks Jonathan for your reply and advice.

One of the reasons I used the 50mm 1.4 in the test was because at portrait distance the images did not seem dead on at all, and I then searched for ways to test whether it was the lens/camera or me!  The only test information was the bit of paper thing, so I thought I would use my macro as well, knowing that it should be spot on focussing with a macro lens, and if it wasn't - there is probably a problem with the camera.

What would be your recommendation for a testing proceedure for a 24 - 70 f2.8 at say 10 mtrs be? ..and should, and how do I test with my 17-40 ?

Julie
Logged

Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad