Poll: AF Accuracy Adjustments

Tried, got confused as to best method, stopped trying after adjusting a couple of lenses.

I have a collection of methodologies, none of which are the same i.e;

printed purpose designed target at 45 degrees,

newsprint taped to a flat wall at 90 degrees to the camera,

lego bricks arranged at 45 degrees,

batteries arranged at 45 degrees,

zoom lens checked at short, middle and long focal lengths,

zoom checked at longest focal length only,

zoom checked at favourite focal length,

....

There are special set ups you can buy specifically for this purpose (I remember seeing some reviews of a couple somewhere) but this is what I do. After setting the camera and lens on a tripod and turning off SR I place a ruler on a desk so the numbers are counting away from me. The lens should be slightly above the ruler so you can distinguish the mm check marks. I also try to make sure the ruler is completely square on to the camera / lens. I then focus on a number using single AF with just the centre point and take the picture (I use the shutter release delay to minimise any shake movement caused by depressing the shutter button).

When viewing the picture if the number and the mark on the ruler that it corresponds to are in focus then all is good. If the number is out of focus compared to the marks in front or behind then you can AF adjust on the camera menu and retake and trial and error a bit until the number is at the centre of focus.

Seems to work okay for me.

Believe me, you would know if something was wrong without having to check

I had never previously felt the need but the previous threads on AF adjustment prompted me to give it a go. So last week I tested my DA 21mm using 3 of the recommended methods and taking a degree of care - mirror up, remote release etc. Try as I might I could see no discernable difference between the results even at the extremes +10, -10 however much I enlarged the image. Perhaps the adjustments are more noticeable on tele lenses?

Yes, they will be. DOF of a 21mm lens, even wide open, is much deeper than a medium to long telephoto/zoom lens.

Try one of your longer focal length lenses rather than one of your shortest to see the difference the adjustment can make.

I'm amazed that of the votes cast so far (43), more than a third (35%) have never checked their lenses/camera for AF accuracy, yet more than half (51%) have had to make adjustments for some, or all, of their lenses.

I'm wondering how many of those 35% will check their lenses/camera now ! ! ! Perhaps JR, SR, & family will check theirs now, knowing that more than half of the PU members that have so far voted, have had to make corrections?

I guess it might be that, as you said yourself, the shorter lenses have a deeper DOF so any problem simply might not be apparent. My longest lens, other than my new 300mm, is 50mm and I have never had a hint of a problem, when viewing photos taken, so have never bothered to check until now.

If you had run the poll yesterday I would have been one of the 35%. Either that or that around 40% of us are getting older and our eyesight can't tell the difference....

The lenses/cameras have been checked and show no problems.

One of the things that I have a concern about is described well by one poster above, that is, the lenses were adjusted and ended up back at zero. We shouldn't expect a lens to need adjustmnent, but the facility is there is needed. If I found all my lenses needed adjustment the camera body would go back.

Equally, with a zoom lens, the adjustment would need to be at a selection of focal lengths, checked using focusing travel in both directions, and an average decided upon. I would need to see some signs of poor focusing to want to do that.

A very wide aperture prime lens is something else, but the eye won't get it right every time either. The amount of error in manual focusing is potentially huge, actually on all lenses.

It's not something I would want to discourage people from doing, but do it knowing why you are doing it, not just because you expect to have to do it. Even the camera manual warns only to do this process if necessary.

Well that's what I thought Ken, and if it ain't broke...

See above comment .

I had problems with lenses on my K10D (thanks to Smeggypants and Anvh) I was able to effect an improvement.

Not noticed a problem with the K5, so I've not checked it.

I am somewhat bemused by the 'fact' that different lenses need different adjustments. Surely the image falling on the sensor should tally with that on the 'focusing sensor' and simultaneously with that on the screen.
Once an adjustment is done, why would changing a lens alter it?

regards
Bernard

I can understand people not checking their lenses,I never have, last week after purchasing a 31 Limited and seeing it was not any better than my other lenses,apart from the boken ,which was dreamy,it started me thinking,I know that it is dangerous,so after seeing what has been written about the 31 I was disappointed,it didn`t come good until f8 so decided to check it for focus and found it back focused and needed +8 to correct.

all of my Sigma lenses are good enough,the 70-200 2.8 is spot on,the 150-500 front focuses slightly,so didn`t do any adjusting the others have been OK so maybe I will check later.

Jack.

I selected "Correction required for one or more lenses" Msot of my lenses have needed adjustment ( both Pentax and SIgma ) on my K10D. K20D and two K-5s



There's nothing to make a hash of. It's adjust a parameter, you can always return it to Zero at any time.



I started by using focus charts, but there's no need. I just focus on something where I can see if the adjustment is not correct. Back of a cornflake packet at a non perpendicular angle for example. Stuff with print on is good.

If you like you can always compare the Contrast AF method as used by the Live view ( becasue that is spot on ) with the regular through the mirror Phase method ( which the AF adjustment parameter is applied to )

I have no idea.

Another strange one is that with my two K-5s, i have a couple of lenses that require adjustment on both K-5s, other lenses that require different adjustments on each K-5 and other lenses that require the same adjustment on both K-5s

I had horrendous issues on my first K-x, but this was across all lenses so clearly the camera was at fault. This required a massive global adjustment of +140 in the debug menu. On my second K-x, my K-m and my K-5, no adjustments have been necessary.

Regards
David

are the "set one" and "set all" on a K5 cumulative or can you only have one or the other?