Go Back   Steve's Digicams Forums > Digital SLR and Interchangeable Lens Cameras > Canon EOS dSLR

Reply
 
Thread Tools Search this Thread
Old Jun 18, 2003, 3:23 PM   #11
Senior Member
 
Join Date: May 2003
Posts: 577
Default

Tuanokc,

Great, thanks for that link! Exactly what I was looking for. Bit pricey still though.

Eric, interesting. I can easily see that happen, just as you're about to shoot you straighten or lean back a bit. That would be user error, and not the camera's fault, obviously.

The focus tests I've done so far seem to indicate front focus for my camera setup. I'll experiment in good day light some more this weekend. I find it a bit strange that the posts on AF focus I've read all say you should test front/back focus in good light conditions. As if it is not important that the camera still manages to focus correctly in poorer (often indoor) light conditions!

Barthold
barthold is offline   Reply With Quote
Old Jun 21, 2003, 5:28 PM   #12
Senior Member
 
Join Date: Mar 2003
Posts: 3,200
Default

When I got my CANON 10D, I took some [real life] test shootings, said: Ohhhh, they are okay ... and forgot this nonsens.

Test with SIGMA 28-200
Test with SIGMA 105 macro
DonalDuc is offline   Reply With Quote
Old Jun 22, 2003, 9:25 PM   #13
Senior Member
 
Join Date: Jun 2002
Posts: 259
Default

Some issues about front/back focusing tests, and Canon AF in general are important to know before you decide your camera is off.

1) AF Precision: On the D60 (and, as far as I know, the 10D) the focus precision is "Normal," as opposed to "High" on the center spot on the 1D. High precision means that the plane of actual focus will be within +/- 1/3rd of the depth of field of the lens at its widest aperture, with a fairly well lit, contrasty subject.

"Normal" precision allows the plane of focus to lie within the entire distance of the depth of field of the lens, wide open. That means, if at the distance you are shooting (2.5 meters is standard with Canon tests), with the lens that you are using, if the depth of field is 10" then the camera can focus that far away from the target and still be considered to meet factory specs.

2) The focus target must be of high contrast, and extend beyond the little focus rectangle. If not, you will never be sure what the camera is focusing on. The target must also be parallel to the sensor plane, for the same reason.

3) There are fron/back adjustments and specifications for the lenses, as well. The interaction between the camera and lens may add the acceptable errors if they are both in the same direction, or cancel them if they are in opposite directions. This is one reason Canon prefers to look at the lens as well as the camera before making adjustments.

4) If the light is not high enough, or if the battery is low, or if the contrast of the target is not high, or if there is a higher contrast object beyond the little focusing rectangle, the AF of the camera will be fooled.

My D60 actually did front focus, and Canon adjusted the lenses as well as the camera, and the results are now absolutely on point. Prior to that, even though I would focus on the eyes for portraits of people, as well as wildlife shots, I always got the nose or the ears in focus, rather than the eyes.

While your cmaera may be, in fact, not focusing correctly, all you can tell with the test that you did is that it may have focused exactly where it should have, rather than where you think it should have.
WalterK is offline   Reply With Quote
Old Jun 23, 2003, 1:42 PM   #14
Senior Member
 
Join Date: Dec 2002
Posts: 5,803
Default

For those that don't know what AF Precision is, I recommend these two links:

http://bobatkins.photo.net/info/faq30/eos3af.html

and
http://forums.dpreview.com/forums/re...essage=4460128

Good stuff. A bit too technical if you're starting out. But if you're trying to get the most out of your camera this type of info is invaluable.

WalterK

Yes, because the focus point can be anywhere within the depth of field, this can lead to confusing focus tests. This is why it's really a good idea to refocus and retake any pictures where DOF really matters (something I forget often!)

barthold

Have you tried the Canon approved test that I posted? I'd be interested in your results. WalterK is right, though. If the target of AF is in focus, that is all you can hope for on a low end pro camera. Canon doesn't guaranty that you'll have 1/3 of the DOF before the focus point and 2/3 after (which seems to be the ideal that people want.)

It's enough to make one wish they bought the 1D. I went from manual focus to this and it seems odd/wrong... but it's a cost savings & product differentiation thing. Such is life.

Eric
eric s is offline   Reply With Quote
Old Jun 23, 2003, 6:42 PM   #15
Senior Member
 
Join Date: Jun 2002
Posts: 259
Default

Quote:
Originally Posted by eric s
Canon doesn't guaranty that you'll have 1/3 of the DOF before the focus point and 2/3 after (which seems to be the ideal that people want.)
Eric
A point of information, Eric. The distribution of the depth of field depends on the distance from the finlm plane, too. The "1/3 in front, 2/3 in back" is only true at fairly large distance from the lens. If you check out a DOF calculater, you'll find that at 2.5 meters, which is what Canon uses as a standard in their focusing tests, the DOF ratio is almost exactly equal, that is to say 1/2 front, 1/2 back.

There are a lot of "facts" in photography that people repeat, that are only true under certain circumstances.


And yes, the AF on the 1D has got to be the finest in the business, especially with wide aperture lenses.
WalterK is offline   Reply With Quote
Old Jun 23, 2003, 9:43 PM   #16
Senior Member
 
Join Date: Dec 2002
Posts: 5,803
Default

Thanks for the correction. I read about that point of focus, but it was in passing and not in a complete article on the subject (which I'd love to see!)

For the targets and situations I shoot in most, I actually want the 1/3-2/3 rule. Of course, I've never figured out how the camera every figures that out... that seems amazingly precise an expression of focus to me. From what you've said, the answer isn't that it always focuses at a certain distance into the focus area (COC?) but that it varies (based on the AF system?)

This is truly a question, not a correction. Since have expressed knowledge.... I'm learning about DOF, as I only know the troubles I've had trying to get a DOF I'm happy with. The things that effect DOF are: aperture (larger yealds larger) distance to subject (Closer means smaller? Yes? Macro is really small.) and mm of lens (No idea how this effects things.)

What DOF calculator do you use? The only one that I've seen didn't include the point where in the DOF that is in focus. (It only said how far and how wide.)

I have read that the Nikon F5's AF is really good too, but it's getting a little long in the tooth, so I wouldn't be surprised if the 1D/1Ds is better.

Eric
eric s is offline   Reply With Quote
Old Jun 24, 2003, 1:40 AM   #17
Senior Member
 
Join Date: May 2003
Posts: 577
Default

Eric, WalterK,

I did a few more tests. I've started to notice that I get better close-up portrait pictures if I focus on the eye furthest away from me, then snap the pictures. Now I know to do this, pictures have improved! On further away subjects, even with a large aperture, it doesnt matter much. This is the main reason I want to find out what is going on; is it my fault (user error) or the camera/lens combination doing closeups.

The tests I did this weekend are somewhat interesting, in that I can see (repeated) front-focus using my 28-105 lens. However, the pictures are dead on, or maybe just slightly back focus, using the 75-300 lens. The latter combination definitely does better.

Ok, then the tests. I used the test-pattern described here http://www.hkdotcom.net/Francis%20Ph...Test/index.htm (the improved test-chart at the bottom). The results are at http://www.pbase.com/barthold/focus_test

28-105 lens
---------------
IMG_6960 85mm focal length, AF, 4.5 aperture
IMG_6966 105mm focal length (camera further away), AF, 4.5
IMG_6975 105mm manual focus, 4.5

75-300 lens
--------------
IMG_7035 140mm, AF, 4.5
IMG_7036 210mm, AF, 4.5
IMG_7038 210mm, manual focus, 4.5


img_6966 is interesting, it has a huge DOF. Don't know why. Any comments?

Anyways, I'm going to enjoy my camera and take lots of pictures. If I think the focus with 28-105 lens gets too annoying maybe I'll have Canon take a look at it. I'm not looking forward to sending it in though. I might be without camera for quite a while. Plus, who knows what their adjustements (if they do any) will do to my 75-300 lens.

Barthold
barthold is offline   Reply With Quote
Old Jul 7, 2003, 1:39 AM   #18
Senior Member
 
Join Date: May 2003
Posts: 577
Default

I would love to hear some comments from all of you on this picture. Taken at 88mm, 1/250, 4.5. I used a single focus point, aimed at the right side of the nose and bottom part of the baby's left eye. Is this correct focus?

The full sized image is at: http://www.pbase.com/image/18729357/original

Thanks,
Barthold

barthold is offline   Reply With Quote
Old Jul 7, 2003, 8:50 AM   #19
Senior Member
 
Join Date: Dec 2002
Posts: 5,803
Default

That is a hard one. I can name things which aren't in focus, like the hand and the flowers on the outfit. To me, the eyes are out of focus. Was this taken using a tripod? At 88mm 1/250, it shouldn't be a problem but a tripod solves other problems.

Also, I assume this is directly out of the camera? No processing? No resizing? No up or down sampling? I loaded the picture into fileviewerutility and it showed the focus point was on the nose, not the eye. It also showed that you have sharpening set to +2. I then loaded the picture into photoshop elements and did a filter->sharpen (not USM, just straightup sharpen) and it looks much better. I'm not sure it is in focus... but if it isn't it's close. Here is the link to the reedited picture:
http://www.marx7.org/~dsmith/eric/baby_reedit.jpg

Note that the picture is 4MB in size. Please post something after you look at it so I can delete it. I can't leave that on my web server for long.

After seeing it in PS and doing the simple sharpening... I'd say that its close enough to not be a problem (assuming people think it isn't in focus.)

Eric
eric s is offline   Reply With Quote
Old Jul 7, 2003, 5:08 PM   #20
Senior Member
 
Join Date: May 2003
Posts: 577
Default

Hi Eric,

Thanks for looking at it! Yes, the hand and flowers on the dress are out of focus. The hand probably because she was moving. The flowers because of the shallow DOF?

This was not taken on a tripod. No post-processing whatsoever, other than the +2 sharpening the camera already does.

I actually didn't use the fileviewer utility to find out exactly where it focussed. Was doing it from memory. I'll trust the fileviewer utility in this case :-)

What I wondered about though is the grass. Only the grass in front of the baby is in focus, nothing beneath and behind the face. Is that normal?

Yes, your edited picture definitely looks sharper. Thanks for doing that. I did copy the file, you can delete it from your server.

I have two more shots like this, all looking similar in focus to me. I'll be happy to post them if you want to take a look at them.

Thanks,
Barthold
barthold is offline   Reply With Quote
 
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



All times are GMT -5. The time now is 5:49 AM.