How much focus adjustment?

Before I start fiddling about, please could anyone suggest a rough amount of adjustment to correct this much back focus:

[IMG]http://i162.photobucket.com/albums/t244/chris5gd/Problems/IMGP0889.jpg[/IMG]

It's the 55-300 on the K-5, 300mm at a range of about 5 feet.

Thanks in advance!

What f/stop did you use Chris?

F5.8. It doesn't get any wider open at the long end.

Edit: PS, I forgot about the troubleshooter forum, sorry, I probably should have posted this there.

The difference is really a little notch. What are the slider marks on the K5? It seems to me that you need to bring it to the 6mm Mark. If the marks go all the way to + and - 20 I would be tempted to go for 5 to start with

I'd never heard about front and back focus until I moved to DSLR, certainly wasn't taught it back in my youth with film cameras but then I never could afford above 135mm for a lens!

I have this lens and at 300mm at maximum aperture (the equivalent of 450mm in 35mm) I expect shallow depth of focus. Is there a FAQ about all this modern angst as I'd really like to understand why so may people are fretting about it!

It may be that people look much more closely at their images than was ever done with film. There were film focusing and alignment problems, usually to do with film flatness, but they were not worried about too much.

It may also be true that modern zoom lenses are so complex that there is a certain amount of manufacturing error in their construction, especially as the use of plastics has grown and prices have fallen. These errors can now be seen on our computer screens and the technology is there to correct them post manufacture.

Of course the technology is also with us to correct distortion in lenses, so does this mean that we are on a path where manufacturers make cheaper lenses and correct the faults in camera? I hope not, but give it another 10 years and we'll see.

If the camera thinks it's focused at distance x, but actually it's focused at distance x + y, where y is some small additional distance, you'll get out of focus shots unexpectedly.

In my test shot above, the camera was pointing at where it says "this text should be perfectly in focus", and the AF was giving me the green focus confirmation hexagon.

Clearly, it's actually focused several mm further back than that, at around the 6mm mark behind where it thinks it's focused. Hence the term 'back-focus'.

In the custom menu, there's an adjustment which can be saved for several lenses separately. I've never understood why the amount of back (or front) focus would be different for different lenses, but there it is.

There's really no more to it than that, in practical terms.

Nothing to do with depth of field - it's a question of where the actual focus point is, relative to where the camera thinks it is.

Actually, I lie - it is relevant to DOF, but only in the sense that if you stop the lens down, more of the distance range appears sharp, so any back or front focusing is masked.

Hi John,

that's what I thought. I sold hundreds of lenses in my retail career back in the late seventies and do not remember ever having to replace a lens other than second hand ones. Cloth shutters failing yes but lenses not focussing correctly. Pretty sure we always blamed the user!

I note I only have one global adjustment in my KR. Very surprised though that I should even have to consider messing with it. If a lens doesn't do what it should then that's a refund under trade descriptions of 'not of mercantable quality' surely?

So far. I haven't adjusted any lenses and my images are sharp when and where required. Maybe the situation is over-fretted about, maybe I've been lucky, maybe my images are lamentably bad and I've not noticed yet.

On the other hand, maybe the lenses I haven't liked have actually been badly adjusted at the factory. In fairness though I have yet to find an SMC Pentax-M 85mm f2 lens that I do like, so I've tried a few and there is a consistency there. And every SMC Pentax-M 100mm f2.8 I've owned has been gorgeous.

As for modern lenses, I haven't rejected any since I started using DSLRs with the *istDS.

I've had a read and this proves that some are being misled into thinking they have a fault. http://www.dofmaster.com/doftable.html

I put in 300mm at F5.6 for my KR in the table and got 1.5 nearest 1.5 furthest at 1.5 metres, in other words correct focus at that distance with no latitude in depth of field. Even stopped down this stays the same. Increase to say 10 metres and much more sign of the 60/40 latitude I always understood they build into optics. Compromises are made in zoom optics (thus the prime fans on here) and logically they optimise for longer distances!

I know from using my zoom on say dragonfly and butterfly photography that it makes a good substitute macro BUT focus is often only on one part of the creature unless you are lucky to be 90 degrees to it.

Now if the focus point is clearly way out (scales on lenses are though approximate only) then back to the shop with it surely????

Thanks to Amin for his suggestion. On the K-5 it seems it goes up to +/- 10.

I've done it a bit more carefully now, and I'm quite troubled.

Here it is, pre-adjustment:

[IMG]http://i162.photobucket.com/albums/t244/chris5gd/Problems/IMGP0901.jpg[/IMG]

Here it is again, +10 dialled in and saved:

[IMG]http://i162.photobucket.com/albums/t244/chris5gd/Problems/IMGP0903.jpg[/IMG]

Both of these are the whole width of the frame.

Notice anything funny?

Um, the sensor isn't absolutely square to the chart. Very hard to achieve without lots of fiddling about.

Well, I've been doing a lot of fiddling about And I thought I had it pretty close. There are a couple of bobbles in the paper where it didn't stick to the card I've mounted it on, but I don't think they affect it where it counts.

Do you think any lack of square accounts for the differnce between left and right?

Some lenses can be very slightly off-centre as well, or it could even be a combination of that and the physical adjustment of the sensor inside the body. The sensor assembly is held by the anti-shake, and the whole enclosing assembly is on three adjusting screws. Out of the three K20D's I've converted to IR the position of those three adjusters has varied quite a bit, between bodies and between the three on a single body. This suggests that the sensor is normally adjusted to be parallel to the lens flange, but perhaps your test is showing some small residual misalignment.

I doubt you'd notice it in real pictures.

This thread on the other forum might be relevant: http://www.pentaxforums.com/forums/pentax-dslr-discussion/159611-k10d-gx10-ccd-p...

For comparison, here's the same lens on the K10D:

[IMG]http://i162.photobucket.com/albums/t244/chris5gd/Problems/IMGP3158-1.jpg[/IMG]

If there's a difference between left and right, it's clearly a lot less pronounced.