DSLRUsers.com

I know of the the massive drama regarding backfocus issues and the D70 upon its release (esp in that other forum). I'm not trying to beat a dead horse, nor am I trolling or giving Canonites ammo.

Here's the story:
Birddog has graciously loaned me his 105DC lens, and I had a proper play with it out in the backyard shooting my pet ducks today. This is a very fast lens at f/2, and the majority of my shots were within 2 stops of wide open apertures. When I downloaded the images to my computer, I was disappointed to see the main subject fuzzy. This was fairly consistent across the images with the defocus control wizardry set to neutral.

I was consciously trying to focus on the eyes, but the whole head and even body appeared out of focus. The focal point was apparently forward of its intended target (aka front focus if you wanna slap a label on it).

So I whipped out Tim Jackson's focus test chart (http://md.co.cz/d70) and set out to test the auto focus acuity with this lens versus my 18-70DX kit lens. I am aware of the criticisms of the test procedure/methodology being unscienfitic one way or another - let's ignore that for a moment and look at the result from these two lenses comparatively. The test chart was laid flat, camera on tripod, angled down estimated 45degrees, bla blah.

I tested both lenses wide open. In the case of the 18-70, the focal length was set to 70mm.

Here's the 105DC result, shot wide open:
http://www.geocities.com/choicebro2004/105dc-1.jpg

100% crop closeup of the distance scale:

Note the slight front focus, approx. between 6 to 10mm (let's call it 8mm) in front of where it should be (frontfocus).


The kit lens at 70mm, f4.5.
http://www.geocities.com/choicebro2004/1870-1.jpg

The increased depth of field made it harder to detect the sharpest point, but if we go by the numbers: eg. 14mm in front of versus 14mm behind where the focal point should be, the numbers in front drop off to apparent out of focus faster than the numbers behind the ideal focal plane, ie. the focal point appears to be in the region behind where it should be (backfocus).

100% crop of distance scale:


So I have both a frontfocusing AND backfocusing D70!

Now how would I go about correcting this issue I'm facing? I've never had a problem with backfocus in real world use with the kit lens (I guess the DOF covered it up, but apparently it is slightly backfocusing having carried out the above experiment), and currently all my other lenses are manual focus, so this 105DC is the only other AF lens I've used on this camera - and it appears to be causing frontfocusing! Now if went out and bought a new AF lens, who knows how it might behave on the D70!?

Is this a camera issue, or a lens issue?

I concluded that if it were a camera issue, both lenses would have consistently backfocused or frontfocused, not one each way. If it were a lens issue, then going by the fact that the 105DC costs more than the D70 body, it would more likely be the 18-70DX that's miscalibrated.

Your thoughts?

Test are all well and good... but the only bad shots I have had using the kit lens is when I have stuffed up!

I can get good sharp images on my kit lens so I dont care if it is a tiny bit front or back... we dont take images of rulers

you can measure all day long and I gaurentee you will find a problem, its how it goes in the field that matters

First off. Don't panic. There are many reasons why that kind of "off focus" issue can occur. Find someone else who also has a D70 and try a few shots with both cameras. Let someone else use your camera for a couple of shots. If you can then still see a problem then take the images to Maxwells along with your camera.

Remember too that the region of DoF of any lens is not equally distributed front to rear, but in fact is, IIRC, something like 1/3 in front of, and 2/3 behind the focal point.

I think that the results you're seeing from the kit lens may be displaying exactly this result, and as such, I'd say that this is perfectly acceptable.

So Gary, how'd you explain the DC lens' result? It's clearly forward of its intended target. I have real world pictures that showed it, and this artificial test confirmed it.


Anyhow, I spent the past two hours "fixing" it - adjusting the eccentrical hex screws inside the D70's mirror box, and testing and re-testing until the AF was spot on with the 105DC. I came to the justfication to use the 105DC as reference for how things SHOULD be, because 1) it's an expensive lens and I figured that's the best point of reference I have, and 2) with previous extensive shooting of the kit lens in real world conditions, focusing issues have never popped up - so I figured I could always hide it behind its comparatively large depth of field, so it should never surface as a problem.

Now to confirm that my fix has not made things worst than before, I'm gunna have to try out a different lens that has a fast aperture to see if it's consistent with the focusing results of this 105DC.

Onyx I don't think I could be that brave at this stage. Any chance of sharing the info on your adjustments which seem to indicate a very clear idea by you of what was needed to 'correct't this issue.

Onyx
You did what a man's gotta do! Waiting to hear your verdict on the adjustments.

Onyx wrote:So Gary, how'd you explain the DC lens' result? It's clearly forward of its intended target. I have real world pictures that showed it, and this artificial test confirmed it.

With the 105 lens wide open, your DoF range is very small.

I'm not too familiar with the target that you're using; I'm of the opinion that most of the so-called back-focus issues people talk about are just that - so-called, and as such I've chosen to remain ignorant of such "tests".

But your angle was an "estimated" 45 degrees, what about distance from focal plane to target? Did you measure that accurately? If not , the test would be difficult to call conclusive.

Let me add that I have similar issues with my 50mm 1.4, where images appear to be poorly focused. But with a fast lens, wide open, at relatively close distances, there is so little DoF that I'm happy to wear the issue as one of operator error.

I’m the same Gary

When I got mine the Back focus thing was at it’s peak, every one was on about it!!!!
I have never tested mine because of the endless false test results that people were coming up with
Now I have settled into complete ignorance that it ever was a problem

Cheers
Ray

Guys - I was one very vocal skeptical proponent in amongst the BF frenzy days as well. I was and still am of the opinion that the majority of cases were not due to the camera at all, compounded with the lack of understanding of depth of field (1/3:2/3 as Gary mentioned) causing needless worry, testing procedures being full of flaws and giving false positives, and that if one had to ask "do I have bf on my D70" - as was so commonly posted - then the answer to such a question wouldn't really matter.

However, in my case it's not consistent - it seems to be front focusing with 1 lens while neutral/slightly backfocusing with another. So I'm guessing it's an issue with one of the lenses, but that doesn't make sense to me, because a lens merely relays info to the body in purely dumb sense, and the body does all the focusing and so on, so I went ahead and adjusted my camera. Instructions can be found here:
http://www.leongoodman.com/d70focus.html

NB: I don't have an Australian warranty to void with my camera, and I document things well and trust myself to revert adjustments back to their original state should things go wrong. If you feel the need to do it with your D70, I suggest you seek a doctor. I STRONGLY advise against doing the things mentioned on that page, unless you know what you're doing. We're talking about differences in fractions of a millimetre here!


Firstly, these are the pics from the 105DC that suggested to me there was a problem:

Duck 1 - AF aimed at the white duck (Ralph), which is clearly OOF. The leaves on the tree, closer to camera than duck, appeared more in focus.


Duck 2 - AF aimed at head/eye of Daffy, which is less sharp than her body.


These are more illustrative of the focus issue experienced with the 105DC on my D70. I had focused on the main pink rose in all of them (at the time, they appeared to be in focus to the eye in the viewfinder - not that it would matter because I didn't focus manually).

At f8 - whole rose appears sharp.


At f4 - leaf facing camera appears more in focus than the rose itself.


At f2.5 - rose OOF, only one edge of the leaf seems in focus.



-------------------------
The fix -

Before:


After:

"This text should be perfectly in focus", and it is - after the adjustments.

I'll try to re-shoot today and see if things changed for the better.

I'll certainly be interested in seeing your results.

I don't remember seeing this through the thread - what AF modes were you using? The default setting for the D70 is to focus on the closest object in the VF, and I'm just wanting to confirm that you've turned that off.

Yes, good point to note - the default AF mode is closest subject. I have it set to single area, locked in the centre, and AF-S (focus priority). Half shutter & recomposed as necessary for all.

Onyx, you are a legend! I am very impressed you had a go at your camera. One thought is you may want a range of lenses if you want to make adjustments. You are more than welcome to borrow any of mine. I actually tested my camera about two weeks ago (when Birddy pointed out potential Maxwell backlogs over Christmas), using your test and used kit, 50 1.8 & 85 1.8 lenses. Everything was ok. One thought is that maybe the 105 DC is not the right lens to use for this test. It may be a very accurate focussing lens or possibly because of the DC they may not worry so much about focussing as there is a big ring at the end of the lens to move focus around. I have no idea, but I would probably want to make adjustments using at least 3 lenses to remove any doubt about sample variations

Onyx wrote:Yes, good point to note - the default AF mode is closest subject. I have it set to single area, locked in the centre, and AF-S (focus priority). Half shutter & recomposed as necessary for all.

Excellent. Just checking ... which raises one more question - AE/AF lock settings in your menu were ???

Thanks Glen, I'll tweak anything to make it 'better'. On the D70 - first it was custom tone curves, then image parameter settings, and how hardware adjustments... I was very meticulous however, and knew how to reset it back to "normal" if things got out of hand.

Yes, the thought had occured to me that the 105DC might not be the prime candidate for calibrating the camera. I will definitely have to try out some more lenses and see if the focus remains consistent on this camera. I have a feeling it might be my kit lens that's a dud... (but I've taken very many pics over the months with this "dud")

Gary - my AF-L/AE-L button is set to AE lock-hold. I've gotten used to my system of manipulating the shutter for focus, and the "all purpose" button as my "instant panorama" button (ie. press once, and shoot the sequence of pics with exposure held constant). Since I discovered the best stitching software, Panorama factory - I love taking multiple shots panoramas! But that's another thread...

Anyway, back to smell the roses:
It seems 105DC is tracking true!! (Please forgive the exposures, I was completely ignoring them)

f/4 shot:

Compared to before, it has improved. The petals seems sharper and more in focus.


At f2.8, the improvement is definitely encouraging!


F2 still has very shallow DOF, but the front edge of the petals facing camera are definitely in focus now (along with the facing leaf)

Success? Too early to tell, but I'm happy.