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

Reply
 
Thread Tools Search this Thread
Old Jan 26, 2005, 11:33 AM   #31
Senior Member
 
Join Date: Apr 2004
Posts: 192
Default

http://www.photo.net/learn/focustest/


Madwand is offline   Reply With Quote
Old Jan 26, 2005, 3:53 PM   #32
NHL
Senior Member
 
NHL's Avatar
 
Join Date: Jun 2002
Location: 39.18776, -77.311353333333
Posts: 11,547
Default

geoffs wrote:
Quote:
However, I did think it strange that in my simple test, that under exactly the same conditions f/4, f/4.5, f/5.0 would yield very nicely focused results while f/5.6 results were obviously OOF. And then on the other side of f/5.6 things would be nicely in focus again...
How many time did you try?

Repeatability is what's important - does the lens always do that or only a few times (ie 1 front focus, 1 back focus, one spot on) :?
NHL is offline   Reply With Quote
Old Jan 26, 2005, 5:05 PM   #33
Senior Member
 
geoffs's Avatar
 
Join Date: Jun 2004
Posts: 2,025
Default

NHL, I repeated my tests 3 times, all with the same result. Still, not enough to draw a solid conclusion yet. I will do a more comprehensive set of tests this weekend in my voluminous spare time :-) I'll let you know what I find out...
geoffs is offline   Reply With Quote
Old Jan 26, 2005, 5:17 PM   #34
Senior Member
 
Join Date: Apr 2004
Posts: 192
Default

Geoffs, I'd like to mention that I don't mean to doubt what your saying -- it might be that you have made an interesting general observation that others have missed; it might be that you have a sample issue, or it might be something else. Stranger things have happened...

It would be good if you would also post some image samples, together with technique information, when you can. In my spare time, I could try duplicating some of the tests myself, or doing other tests, to see if I have the same results.
Madwand is offline   Reply With Quote
Old Jan 27, 2005, 12:31 PM   #35
NHL
Senior Member
 
NHL's Avatar
 
Join Date: Jun 2002
Location: 39.18776, -77.311353333333
Posts: 11,547
Default

Interesting reading - http://www.outbackphoto.com/workshop...y06/essay.html

:?
NHL is offline   Reply With Quote
Old Jan 27, 2005, 4:10 PM   #36
Senior Member
 
geoffs's Avatar
 
Join Date: Jun 2004
Posts: 2,025
Default

Wow, that is a very interesting article indeed, NHL! I know that I have been guilty of using focus-recompose a good percentage of the time for close subjects and that might explain some of the problems I've seen. Too bad I can't pin that on the squirrel back-focused shots I posted.

To have knowledge is also to know how to avoid situations that would produce incorrect results and this article has definitely helped in teaching me yet more. Thanks for posting that link, NHL!

To a certain extent, it has complicated my life because it suggests that it would be better to use one of the alternate focusing sensors as an alternative to focus-recompose. The extra step involved with selecting an alternate sensor introduces yet another opportunity for me to fumble with the buttons and dials... :-(
geoffs is offline   Reply With Quote
Old Jan 28, 2005, 4:15 AM   #37
Member
 
TDM_Canon_User's Avatar
 
Join Date: Jan 2005
Posts: 90
Default



Agreed - it's a good article about a common mistake made by many camera operators. But the L series focus problem goes further. My cameras all have multi-point focusing so I can set the focal point without recomposing. I still get images where the focus is all over the map and not where I know I focused. Both outside (like the article would suggest), and inside (which says it isn't the technique problem). The old ruler and tripod is a sure fire method to know the camera is or is not focusing right.

Now if I can ever decide to buy that split image screen so that when it's really important I can focus manually again... but then I give up the neat metering mode. Sigh.

--

And regarding the lens cal to a particular body. Here, if you have a good old fashioned lens that focuses on infinity (with a true infinity stop on the focus ring instead of the AF over reach), you can test your camera body for focus.

Take said lens and body, set-up on a sturdy tripod, focus on infinity, aim at the sky at night (moon works well, 1/125s f4.0 or so). Then turn the focus ring a tweak, take another picture.

Back in the house, review the images. If by chance the infinity setting isn't as clear as the tweaked setting, the body may need the mirror adjusted. This all said, there's still a micro chance the lens is out, but it's really a small chance compared to the mirror mechanism that bounces 3/4 an inch to a hard stop at each exposure.



TDM_Canon_User is offline   Reply With Quote
Old Jan 28, 2005, 9:26 AM   #38
NHL
Senior Member
 
NHL's Avatar
 
Join Date: Jun 2002
Location: 39.18776, -77.311353333333
Posts: 11,547
Default

... To add to the above mirror mechanism comment - The mirror is usually out of the way when a picture is taken but the possibility is there for the mirror to be not at its 45 degree optimum position, in which case the image in the viewfinder would be partially out of focus (either the top 1/2, bottom 1/2, or both). One can see this with a viewfinder magnifier.

I agree though "there's still a micro chance the lens is out, but it's really a small chance compared to the" mechanical tolerances of the (1)viewing screen, (2)AF sensors, and (3) the CMOS sensor itself which could be all not equidistant from the folding mirror - sometime just due to manufacturing tolerance, QC control, or 'shocks' in shipment:

1. The viewing screen could be off -> the picture you manually focus may look great, but the image projected on the CMOS sensor can be out of focus because the projected image took a slightly longer(/shorter) path! :?

2. Look in the mirror box behind the main mirror -> there's another smaller mirror behind it folding the beam down to the AF sensors on the bottom of the camera body: if this mechanism is off, the camera 'think' that the picture is in focus, but the actual image projected on the CMOS/CCD sensor is not! :O

3. Now the actual CMOS sensor itself can be out of position (look at the previous link on how the guy modified the camera by removing the spacer shim) -> now if this CMOS sensor is not on the focal plane then regardless of what the viewfinder or the AF focus mechanism think, the captured image will always be off as well! :evil:

... A lens is quite robust in contrary: it can (1)focus past infinity, (2)you can add correction diopters (+1, +2,...etc) to the front, and you can also (3)modify the lens mount distance with various extension tubes and it can still AF correctly
NHL is offline   Reply With Quote
Old Jan 29, 2005, 1:37 PM   #39
Junior Member
 
Join Date: Nov 2004
Posts: 2
Default

Thanks much. Went to the website, got the chart. My 70-220 mm lens arrived yesterday and I'll be doing some testing this weekend.
fjklein 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 1:44 AM.