focus adjustment k10d

It's fascinating how these new problems crop up after years of it never being mentioned. If it's a problem with Pentax then I wouldn't be surprised if it isn't a problem with Nikon and Canon as well.

Fortunately my focusing seems fine. It would be IMHO intolerable to have to contemnplate constantly re-adjusting for different lenses. Any fault like that and it would be straight back to Pentax for a proper fix under warranty.

If they really are talking about adjustments of microns I can not see how the mating accuracy of the K-mount would be good enough. If it was that critical then you would need to re-calibrate every time you changed the lens – even taking the a lens off the camera and immediately re-attaching would change things.

Shaking the sensor around with SR would play havoc with any calibration if adjustments of a few microns made that much difference.

The adjustment in microns is with regard to the sensor, not the lens And it will be with regard to the AF-sensor, not the CCD too (though with the CCD effectively fixed that amounts to pretty much the same thing).

Matt

But the CCD is not fixed anymore. Is this a possible cause of this new situation? Maybe the SR system means that there is more room for error than there used to be.

I wonder if there has been a test made of absolute sharpness vs reduction in camera shake or whether there has been a trade-off.

The camera can be adjusted by hundreds of microns if desired, so we're not necessarily talking about just "a few" microns.

Very interesting point, John.

To test this, fast shutter speeds would be needed, to guarantee that there was no shake in the image when SR was off. This would be true on the tripod, too, of course, because only when the shutter is manually released, can SR be off anyway.

All my SR-related studies have been looking for SR causing shake during long exposures, not at all at sharpness at short exposures, with and without the sensor being able to move.

It would be very ironic if the real reason Pentax say leave SR off on the tripod is to do with achievable absolute sharpness, as you say, rather than the widespread and IMO unfounded claims that SR causes shake.

I don't know whether the differences we're talking about (between the sensor locked and mobile cases) are large enough, but if this were the case it may explain Don's astonishing observations too. Because if absolute sharpness achievable with the sensor fixed is higher, then images that are already shake-free might be better with SR off.

But it's a tall order getting shake-free images with long exposures off-tripod.

Interesting theory John, but if that were the case then we wouldn't see reports of exactly the same problems occuring in N/C cameras (where there is no sensor-based image stabilisation).

And is the situation really all that new? I first became aware of these problems when they were widely reported on the Nikon D70 - which also happened to be one of the first DSLR's that sold in very large quantities. The more users a camera has, the larger the base of people who are likely to find a problem.

I don't know the Pentax sales figures but I wouldn't mind betting that the K series cameras have outsold the *ist series cameras by a very wide margin. I would also imagine that the *ist series sold predominantly to Pentax die-hards with an existing lens collection, whereas the K series have sold to people (like myself) with no brand loyalty but who appreciate an exceptional product when they see it (in addition to the Pentax die-hards!).

It's worth remembering that the faults we describe here on this forum do not represent the findings of the majority of users. Two of my friends have recently bought their first DSLRs (K10D & GX20 - not that I had any influence of course!) and neither of them have any BF/FF problems. Judging by threads like this (which always have the same people contributing) it's not even the majority of users on this - or other - forums. The "Internet Effect" just magnifies a few unfortunate cases and paranoia can set in, with people who never considered their camera to be faulty suddenly blaming every OOF shot on a manufacturing defect.

It is a fact that some cameras leave the factory with mis-aligned sensors.

It is a fact that this affects all DSLR manufacturers.

It is a fact that owners of a K10D have a simple remedy open to them (without having to resort to sending their camera back).

It is a fact that K20D / GX20 owners have an even easier to access remedy at their fingertips.

It is pure supposition on my part that by the year 2018 Sony will still have not admitted any possibility of a problem - anyone care to bet against this?

We're all speculating, of course. Sort of related to this was a question in last week's AP which asked why manufacturers don't put levelling lines in the viewfinders of their DSLRs. We could then all avoid those sloping horizons!

The answer is probably not the one given in AP, but perhaps more to do with the accuracy of positioning of the viewfinders relative to the sensors. If there were lines provided then they would have to be 100% accurate and that would no doubt cost a lot more. As it is, there is probably some degree of error that is generally acceptable and, after all, we can always put a brick under one foot.

In my case the only noticeable focus problems I’ve had are with old K series lenses, but that’s probably more down to technique – relying too much on the focus indicator in the viewfinder and not concentrating on the image.

The four modern lenses I have, 24/90mm 16/45mm, DFA 100mm macro and the DA* 300mm generally produce well focused shots, especially the DFA and more so the DA*.

To get back to technique - with the old split prism screens you were forced to concentrate on the image when focusing, but with autofocus it’s all too easy to rely on the technology where admittedly a few cameras for whatever reason will be misaligned but more often the issue could be that no autofocus system is perfect depending on light levels and image content etc. Even after calibrating there will still be times when the AF will fail to focus properly.

Not having played with the debug mode I guess there are settings available which could well and truly mess things up, this is probably why Pentax are reluctant to let users access debug mode. It could be that allowing access to focus calibration while protecting the camera from damage is not so easy when the firmware was not designed with this feature originally.

I have a standard, unmodified K10D.

I can count six features in the viewfinder which are perfectly adequate as horizontal levelling lines.

Why anyone would need any more than this to get their horizons horizontal is beyond me.

It's easy to correct any mess you make in the K10D's debgug mode. There's not very much to it, actually. Most users will never need or want to use it, but it is a shame that Pentax make it so difficult to access for those who do.

As for why more users are adjusting their AF, and talking about it, than used to be the case years ago, it could just be because more people know about the debug modes now thanks to the subject being discussed a lot on the internet. Also the higher resolution of cameras now shows up smaller faults, so we can be more fussy.

Perhaps the question we should be asking is why do so many images we see have sloping horizons? Even magazines and books have plenty of examples. Does no-one notice? :

A quick update:

I noticed that my K10D was indeed 'back focussing' and have used the 'Russian' software (with firmware 1.30 on my K10D) to get into debug mode and correct this.

The English language version of the software does not appear to install correctly - when I tried to run the program, the .exe file was missing.

However, the 1.1u non-English (I presume Russian?) version installed and worked perfectly and in accordance with the tutorial.

Hope this helps if anyone is struggling to get this working.