AF Adjustment

I have had my da*200 for a 6 months now. Recently the images have seemed a little soft so I thought maybe I should adjust focus. Using a printed focus chart, tripod , 2 sec timer, shot at 45 degrees I found that +7 was needed and it seemed better. Then I read this post and realized that when I did the test I had been much too close to the target. 8 feet. So I did it again.

My set up basically the same but 90 degrees, at a distance of 17 feet and the target was my laptop screen diplaying a focus target at 100%. I tried the dot-tune method @ http://www.youtube.com/watch?v=7zE50jCUPhM. First attempt I found the hexagon focus indicator stayed lit in the range of -4 to + 9

This gives about a +3.5 but after rechecking several times alternating focusing from out of focus at infinity to focused and out of focus at minimum focus distance to focused I found a large variation in results eg -7 to +7, -8 to +4, -6 to +10 etc.

I rechecked the Dot-tune method and apparently the above method is not very accurate because it does not take into account mechanical tolerances of the lens.

So could this be the reason for the variable results or does the lens have a fault?

Also why does the lens have to be tested at a mimimum of 25x focal length?
I would like to test it again but at a greater distance which would mean outdoors but unfortunately the weather has been terrible of late.

Lenses are computed to be at an optimum at the distances they are expected to be used at. A macro lens at close distances, a long telephoto presumably at long distances. A portrait lens at the appropriate distance.

There's no perfect lens and some even say in the instructions if used close, use stopped down. The 16-45mm is an example of this,

With the current trend to more and more phase detect autofocus points shouldn't one be adjusting and calibrating each individual focus point separately and at every zoom and focus distance setting ?

I just set up my lenses at infinity because that is where I find mis-focus most annoying. They all seem to be within one step anyway, except the DA35 Ltd which is the same close up but 5 odd steps off at infinity - as expected from a macro.

Some people seem to get obsessed with the 'calibration' when their fuzzy photo problems lie elsewhere ...

Thanks for your reply. That makes complete sense but I would have thought that in focus would be in focus as long its within minimum and infinity. Then again if you are saying that a lenses focus drifts depending on aperture or distance to subject then after having read several threads it seems that checking the lens at a variety of distances and taking an average is best or at the distance at which the lens is used most often.

Zoom lenses in particular will drift as you zoom. There are very few true zooms these days, that is parfocal zooms that hold focus as we zoom. Most are varifocal and need to be refocused after zooming.

Prime lenses also vary as we focus and can actually also change focal length. This is especially true of internal focusing lenses.

So add together a varifocal zoom and IF and the possible drifting and errors becomes a minefield. Without computer design these lenses would not be possible.

Look at camera shake as a possible source of blurring, proper use of aperture for the required depth of field and the actual point being chosen to focus on. It should rarely be the lens that's at fault because of calibration issues.

At last a fine day. I could do the test outside at 30m on a tripod with 2 sec timer with flat high contrast target. again using the dot-tune method. This time got consistent results refocusing several times. -2 to +9 when refocused from infinity and -4 to +8 when refocused the other way which gives about a plus 3.

I am assuming that the different results when refocusing from different directions is due to mechanical tolerances in the focus mechanism and the fact that I can hear, in a quiet room, a clinking sound when the elements reach the focus stops at infinity and Zero is normal. What do you think?

Regardless of whether you can accurately calibrate a lens at all focal lengths (which you can't), calibrating your lenses to each individual camera body is a must if you are interested in ensuring you get the best results from your lenses.

I've owned 9 Pentax DSLR's and a K-01 and the only one that hasn't required any lens calibration is the K-01, purely due to the AF method it employs. There has been massive variation between K-5's and to a lesser extent with K-5iis and K-3, but all have required individual calibration of each lens. I almost sold my DA L 55-300 when I first got it as I couldn't get a sharp image from it. Two things changed that: I calibrated it and I ditched the protective UV filter - hey presto it's sharp!

I simply don't believe anyone who says they have never had to calibrate any lenses. To qualify that statement, if they do say this I firmly believe they are not getting the best from their equipment.

Regards
David

Guilty - not getting the best from my equipment

If that's the case Alan, I dread to think what your shots would be like if you did calibrate your lenses! Either you have been very lucky with your cameras and lenses, or you are an outright photographic genius!

Cheers
David

Guilty too, you can't calibrate a zoom as it changes throughout the range. My DA15mm Ltd is my only prime and works fine.

I thought calibration was introduced to adjust very old lenses. If I bought a new one and it didn't focus I'd send it back under warranty.

It depends whether you can spot an error or not in real world situations. In my experience, the more you test, the more variable your results become and it's easy to get mis-led by a erroneous test. There are so many variables that precise corrections are almost impossible to achieve more often than not.

You can calibrate a zoom to an extent. Calibrate it for an average focal length / focus distance. Some lens / camera combinations can be so far out that they will not focus accurately at any focal length or focus distance. Ask AndrewA about the situation with his K-5 and all of his lenses, primes and zooms. I had a similar situation with my first K-5, it was miles out with all lenses. The second one was great, not much adjustment needed.

I bought a mint DA 100mm WR Macro off a forum member about 8 months ago. Lovely lens, but when I tried to use it the focus was appalling, even on bodies that focussed accurately with other lenses. It needs +9 adjustment on my K-5iis and almost as much on my K-3 and now it focusses accurately. I wonder if this is why it was so 'mint'? Did the previous owner not use it because of this?

I repeat what I said before, if lenses are not calibrated to each body they will not work at their best unless you are very, very lucky. There is tolerance in both camera and lens manufacture, so this is just plain common sense. It applies to zooms as much as it does to primes, the only thing is that the calibration on zooms will not be spot on at all focal lengths.

Regards
David

A couple of years ago I calibrated the 55-300 David. Spent about 45 mins doing it. I was happy with the adjustment made on the day. Come the weekend I wasn't happy, so did it again and ended back at zero.

Haven't done any since

We seem to cover this every now and again. I haven't ever found a lens that didn't focus where it was supposed to.

Even the camera instructions caution against calibrating unless it's really necessary. There are so many variables, including length of focus travel, direction of focus, distance, zoom setting....

Call me old fashioned, but they should be set at the factory. This was never a problem until recently, and it's only become one since the mechanism to fiddle was made available to us. I think they made it available because some people were messing in the service menus anyway.

My Pentax WR 100mm F2.8 Macro, a lens that costs £800 when new, would have had to go in the bin if I hadn't calibrated it. As I bought it second hand, I could not send it back to the shop for exchange or refund. It was useless before I adjusted the AF on all my bodies, bearing in mind this is a macro lens and is used in situations where focus depth is wafer-thin. It produced plain blurred results. It's now great, there's nothing wrong with the optics, it just needed calibrating.

You are lucky Alan! I love my 55-300 and it produces very accurate and consistent results, but I have calibrated it and once calibrated I haven't had to return it back to zero, I'm very happy with it the calibrated setting.

I disagree John. How can it 'be calibrated at the factory' when the manufacturer cannot account for the combination of camera body and lens that we use? It is the interaction between the two that produces the accuracy or inaccuracy. I'm not trying to claim that you can calibrate for all eventualities, just that you can make the best of a particular camera and lens combination. How would you have dealt with the situation I have described on my Pentax 100mm macro? Would it just have been left on the shelf?

Regards
David