[Fontan]

Christine, you make enough money. Go buy a Leica. You will no longer have any AF issue whatsoever. Period. Merry Christmas.

[krebsy75]

This is a cool shot. Nice!

Well, I thought the K-5 was amazing compared to the K10D (which couldn't really focus all that well in low light, neither could the Canon 1D Mk III)

But then I discovered my NEX-5N could do this, in virtual darkness (there is so little light, I could not even see properly):


So I guess my real question is, is there any way of getting the K-5 to match the AF accuracy of the NEX-5N in anything other than Live View?

Another possibility is that somehow the AF accuracy of the K-5 has deteriorated with age. Because I do seem to recall I was fairly impressed with K-5 accuracy when I first bought it.

[krebsy75]

-1 for not being very friendly.

Christine, you make enough money. Go buy a Leica. You will no longer have any AF issue whatsoever. Period. Merry Christmas.

[pentup]

Go buy a Leica. You will no longer have any AF issue whatsoever. Period. Merry Christmas.

Rather a shame to see the gracious phrase "Merry Christmas" tacked on to one of these petulant and unhelpful "go buy a [fill in the brand]" messages that too many people entertain themselves with on forums like this.

[Christine Tham]

Big time deja vu, didn't we have this conversation almost a year ago? Except you where one of those stating that there wasn't a low light AF problem? (I could be wrong though as there where a lot of people involved in the low light AF incident. )

Well, it wouldn't be me because I only bought the K-5 in Feb.

And it's not a low light AF problem. It's a consistent lack of precision in AF, and I have confirmed that it manifests itself even in good lighting.

It's the same imprecision that drove me nuts on the K10D, and finally convinced me to buy a K-5. And now it seems I have similar issues on the K-5 after 20k shots.

It can't be adjusted away because it's not a systematic error. It's almost as if the motor is not stepping the lens focus where it needs to but slightly off, and after a while the camera gives up. It's not always front focus either, sometimes it's back focus - depends on which direction the lens barrel was rotating when seeking focus.

The misfocus is very small - it's only noticeable when shooting wide open and the DOF is razor thin. But it means I can't shoot below f4 without risking OOF.

[Christine Tham]

Christine, you make enough money. Go buy a Leica. You will no longer have any AF issue whatsoever. Period. Merry Christmas.

It's ironic you say that, because I had a lot of problems focusing on a Leica M9 - my right eye can't focus properly in the viewfinder.

I eventually manage to focus, but it was quite difficult and gave me a headache. Wish I had kept some of the lenses though, they would have come in very handy on the NEX-5N.

Here's an example, this was a practice shot focusing on the weed stalks - the lens is a Summilux 35mm f1.4:


I found it nearly impossible to focus in dark indoor shots - it took me nearly a minute to focus on this at the museum, and the M9 high ISO performance is terrible - I had to do a lot of post processing noise reduction to clean it up:


Eventually I gave up and started shooting at f4 - and really, what is the point of using a Leica to take high DOF photos???

[jolepp]

It might be an idea to shoot a static, AF friendly target (*) in good light. If you can see an inconsistency with that then the problem could be something having worn / worked loose so that the length of the optical path to the PDAF sensor varies between shots. (This in the grand tradition of idle speculation, but it would seem to explain the problem ). Should it turn out that PDAF accuracy varies as much as to result in both 'acceptable' and 'useless' results in ideal conditions you would have a very concrete complaint to present to the seller / service. As often is the case, the devil is in the details as to what would be 'acceptable' and what 'useless', some variance is to be expected. Taking shots of the same target with LV / CDAF might provide a baseline for comparision, or you could discover that both AF methods result in unacceptable variance in which case the problem could be something like the focus adjusting mechanism having developed mechanical play to affect both AF methods. Another source for a baseline would be your NEX, of course.

(') such as a focus chart, or maybe just a piece of paper with largish text on it for a contrasty target that does not have a simple repeating pattern, for best results a tripod, 2s delay, center point only with the target amply filling the center sensor area etc.

[Fontan]

It's ironic you say that, because I had a lot of problems focusing on a Leica M9 - my right eye can't focus properly in the viewfinder.

I eventually manage to focus, but it was quite difficult and gave me a headache. Wish I had kept some of the lenses though, they would have come in very handy on the NEX-5N.

Here's an example, this was a practice shot focusing on the weed stalks - the lens is a Summilux 35mm f1.4:


I found it nearly impossible to focus in dark indoor shots - it took me nearly a minute to focus on this at the museum, and the M9 high ISO performance is terrible - I had to do a lot of post processing noise reduction to clean it up:


Eventually I gave up and started shooting at f4 - and really, what is the point of using a Leica to take high DOF photos???

Ah, you did try. I am disappointed to hear your headache.

I meant what I said. I wonder what you mean when you say "properly." When I started using my M8, my first viewfinder ever, it didn't take me long to get used to it. In fact, after about a week with it I could focus spot on without my glasses because it was easy to see those two images merging together, and fine tuning had more to do with the turing of the focus barrel and pacing of it than the act of "seeing" itself. How long did you try out M9 for?? I figure you must have tried a good while as it is a substantial investment of some sort. 6 months?

I think that if there is a point at all of shooting with Leica, I don't think it is neither DOF nor sharpness. Again, if there is a point, to me it is about being descreet to the moment, being as ready as you can be if and when you happen to encounter that moment, and finally being able to respond and adapt to that opportunity before it fleets.

I find auto focusing in general a lazy act. Even if your eye sight is such that MF is difficult, it is still worth developing that sense of distance in a way of knowing where certain familiar "distances" correspond to which particular senctions of the lens barrel. Skillful photographers of the past have been shooting fast moving objects way before the invention of AF.

Now I apologize for deviating from your orignial post - K-5 and its AF. Between M9 and the most advanced AF system that Pentax has ever created, perhaps you hit a goofy sweet spot with your Sony, which may well be the most ironic. No?

[Christine Tham]

Ah, you did try. I am disappointed to hear your headache.

I meant what I said. I wonder what you mean when you say "properly." When I started using my M8, my first viewfinder ever, it didn't take me long to get used to it. In fact, after about a week with it I could focus spot on without my glasses because it was easy to see those two images merging together, and fine tuning had more to do with the turing of the focus barrel and pacing of it than the act of "seeing" itself.

I have suffered permanent damage to my right eye as a result of high cholesterol, and it can no longer focus properly. Fortunately I can still use the viewfinder on my K-5 because the apparent focus distance is at the sweet spot. Even then, it's not focusing properly because the image is never perfectly sharp.

On the M-9, unfortunately, my right eye can never focus on the viewfinder.

As a result, in terms of "trying to merge two images together" - I see (multiple) blurry images and I can never really tell when they are exactly aligned.

It's especially difficult when there is not a lot of light, because then my right fails to focus at all - at any focus distance.

On the NEX-5N, I don't have any issues because my brain favours the left eye when I look at the LCD.

[dosdan]

Christine, have you tried LV in conjunction with the Info button? Up to 6x zoom with AF & 10x with MF. At least you'd be able to confirm that the Katz Eye screen is shimmed correctly.

Dan.

[philbaum]

What's your shutter count on your cameras? Mine is well over 20k on the K-5, and well over 30k on the K10D.

I'm asking because I suspect the deterioration is based on usage, rather than age of the camera. I've had my K-5 since Feb too.

Sorry for the delay - My K5 is just shy of 10,000 shutter clicks.

[dexmus]

This is a contentious issue. I think you hit the nail on the head when you said that earlier you perceived the AF to be great when your benchmark was k10D. But if anyone has used a pro Nikon body (such as D700 and up) or a pro Canon (I haven't used a Sony 5N), he/she can quickly see that the AF in k-5 is not upto the mark, in terms of consistency. Where as the AF speed may now be close, the consistency of right focus is still not there.
Jim Radcliffe (from boxedlight.com) was posting here a while back and even though he loved the k-5, he was also quick to note the lack of AF consistency. His theory was that the AF sensors are bigger than the AF area shown in the focus screen. And therefore, sometimes Camera locks on something outside the intended spot. That sounds legitimate to me.

cheers,

Abhi

[Ex Finn.]

But if anyone has used a pro Nikon body (such as D700 and up) or a pro Canon

And the entry fee for that privilege is what?. I like to keep my expectations of body performance directly linked to cost of equipment. For a thousand bucks Pentax is not bad.
Not saying that her camera does not have a problem.
Merry Christmas.

[Christine Tham]

This is a contentious issue. I think you hit the nail on the head when you said that earlier you perceived the AF to be great when your benchmark was k10D. But if anyone has used a pro Nikon body (such as D700 and up) or a pro Canon (I haven't used a Sony 5N), he/she can quickly see that the AF in k-5 is not upto the mark, in terms of consistency.

When I first bought the K-5, I compared the AF not only with the K10D, but also with the 1D Mk III and 1D Mk IV. I remembered being pretty satisfied with the AF performance then. Okay, it wasn't as good as the Canon bodies, but it was doing pretty well.

I have taken hundreds of shots on the K-5 on the FA43 wide open, and I don't remember it missing focus as often as it seems to now.

Still debating whether I should take the K-5 in for repair, or just give up on it. I may ring them in early Jan and ask for a quote. I have a suspicion come Jan and once the NEX-7 is in stock, I may actually cave in and buy a NEX-7 - a few Leica owners I know are planning to abandon their Leica bodies for a NEX-7.

[D4rknezz]

Did you try what jolepp said?

Because if you put your camera on a tripod in good light, and aim it at an object and shoot, and then un-focus a little bit, then shoot, then un-focus a little bit, then shoot....you should get almost 100% accuracy if nothing else changed.

If your camera af really has a wear/tear problem, then it should not matter if the camera is in good light/bad light/etc, the focus would be everywhere.

The nex 7 sounds pretty good, tell us how it compares when you get it.