SLR - autofocus appears to be inaccurate, ways to test?

I've bought a used Nikon N80 and I think the autofocus may be slightly
off. When I autofocus on a scene that I've already focused on, the lens
focus moves and the image appears slightly unsharp.
I've downloaded a test pattern from the Internet. What's the proper way
to test the focus? If it is off, would adjusting it be part of the
typical CLA?
Thanks.

THO

Re: SLR - autofocus appears to be inaccurate, ways to test?

"Norm Fleming" <nfleming@ms.umanitoba.ca> wrote
You are right in that there should be more in focus behind than
in front. It isn't necessarily 2:1 though, as fiddling with a
lens' dof scale will show.
For distances substantially longer than the focal length of the
lens, using a 50mm lens and a 0.025 mm (0.001") circle-of-confusion
[all distances in meters]:
Subject f-stop Near Far Ratio
distance dof dof far/near
zone of sharpness
1 2 0.98 1.02 1.04
5 2 4.54 5.57 1.23
10 2 8.31 12.55 1.51
1 8 0.92 1.09 1.18
5 8 3.56 8.42 2.37
10 8 5.52 53.42 9.68
10 22 3.09 infinity infinite
Wide open at close distances the depth of field in front
and in back is functionally the same; a difference of
4% is not significant. For a typical test at 1 meter/3ft
the sharpness zone should be ~2cm/1" in front and behind
the focusing line.
So, for practical matters, when checking focus accuracy
the in-focus zone should be symmetrical about the focus
point.
--
Nicholas O. Lindan, Cleveland, Ohio
Darkroom Automation
http://www.nolindan.com/da/index.htm
n o lindan at ix dot netcom dot com

Nicholas O. Lindan

Re: SLR - autofocus appears to be inaccurate, ways to test?

...
The one I had,(not a Nikon), seemed like it was guessing most of the time,
but it did work. I had to move the diopter all the way over to get any kind
of sharpness in the finder and it was always searching. I just focused
manually, cause the shutter lag was just too annoying. Bob Hickey

Bob Hickey

Re: SLR - autofocus appears to be inaccurate, ways to test?

Who started this myth and how do we get rid of it?
Take two cases, very small DOF will extend about the same in front and
in back.
Case two hyperfocal, extends all the way to infinity, clearly not 2/3
of how far it extend forward, which is half the distance to the camera.
Scott

Scott W

Re: SLR - autofocus appears to be inaccurate, ways to test?

SNIP
Why would the line be smack in the middle of the sharp focus area?
The depth of field extends twice as far behind the point of focus as it
does in front.
So if you focus on a point, everything will be in focus from a distance in
front of that point (x cm ), to twice that distance (2x cm) behind the
point, for a total depth of field of 3x cm for that particular stop?
Or does your 45 degree angle do something to cancel out that relationship?
That I would doubt, but can't begin to get my simple mind round the possible
physics.

Norm Fleming

Re: SLR - autofocus appears to be inaccurate, ways to test?

...
Is the output image slightly unsharp or is the viewfinder image slightly
unsharp when the camera focuses? If it is the latter, check your diopter
adjustment, if it has one (or check your eyesight). If the output image
seems unsharp, follow the previous poster's test.
Eric Miller

Eric Miller

Re: SLR - autofocus appears to be inaccurate, ways to test?

"THO" <tho@tho.tho.23.invalid> wrote
When the lens is set for oo and focused on a far target the camera
will be in focus [barring any gross mechanical errors]. If the
viewfinder is in focus at oo then the AF system is in error. If
the viewfinder is fuzzy but the AF system is happy then the viewfinder
is off.
You need a real infinity for this test - something at least one
mile away. The moon is coming on to full and makes a good
distant target.
For closer focusing distances the standard 'quick & easy' way
is to tape the newspaper classifieds to the wall, make a
vertical line with a magic marker, put the camera on a tripod at a
45 degree angle to the wall and focus on the magic marker line.
Use the maximum aperture [& focal length if a zoom] setting.
Take a picture with manual focus and one with automatic focus.
Examine the prints or negatives. See if the marker line is
in focus and if an equal portion of the print is in focus before
and behind the line: the line should be smack in the middle of
the sharp focus area.
Other targets are window screen, a row of pencil points on
the table, etc, etc.
No. If the visual focus is off and the AF works and you use the
AF all the time anyway the best thing may be to leave it be. There
is an engineering adage that says if you fix something long enough
you will break it.
There are posters on this group that can refer you to a good
Nikon repair shop if you need one.
--
Nicholas O. Lindan, Cleveland, Ohio
Darkroom Automation
http://www.nolindan.com/da/index.htm
n o lindan at ix dot netcom dot com