[godavari]

I came across this article randomly.

LumoLabs

In its conclusion section it mentions that

"The K-5 as it presently ships has a flaw in its phase detect autofocus module or software which causes it to front focus in low enough light. This applies to both v1.02 and 1.03 firmware releases."

Not sure about validity but it makes an interesting point that even with firmware update the AF issues still exist.

I own a K-5 too and I randomly see AF issues particularly in low light (with and without tripod) and multi-segment AF setting. I use DA* lenses (50-135) and not sure if it was lens or K-5 itself.

Any comments/opinions/experiences?

[imtheguy]

I recently tested Live View (contrast detect) versus the normal viewfinder (phase detect) under low incandescent light and found the Live View consistant and accurate. The Phase Detect was usually front focused.

Camera is k-5, firmware 1.11, lens was Sigma 24-70/2.8 at various focal lengths and f-stops. I had not done this check in almost a year so was hoping for better. Fortunately I am strictly an outdoor shooter and seldom deal with this.

[jaytee]

This subject has been debated quite a lot here and elsewhere and the jury is still out I think. I focus manual and find it hard to believe that AF really works better in low light for anyone. That's just me I guess but I hear other people's AF going back and forth and back and forth, it would drive me crazy. Anyway perhaps Pentax will release and even better firmware update in the future?

[Dave L]

I've just tested this with my new 55-300 at 300mm on a nearly new K5 at 4 metres range, f5.8, 1/5 at ISO 800, SR off on a tripod at 2S delay. Firmware v1.11 Normal room lighting (quite dim).

At zero fine adjustment the target (a cotton bud on a mm ruler!) is in perfect focus but is near the back of the in-focus range.

At -10 fine adjustment (the max.) the target is still in perfect focus and is nearer the middle of the in-focus range, so still a hint of front focus but nothing to worry about IMO.

Outdoors in daylight on a dull day when I tested it yesterday it was in perfect focus.

[Nicks]

My k-5 does this and it's annoying. I have to use a -7 AF correction for my Tamron 17-50 when I shoot indoor and a +5 when I go outside. Pentax AF 50mm f1.4 needs a -4 indoor and +2 outdoors. Firmware 1.03 didn't fix it. Live view works much better but it's such a hassle to use it.....

[filorp]

firmwares 1.10 and 1.11 fixed the problem as much as it was possible with the tiny af system pentax implemented inside k5 body.... i was comparing sizes of the af system from d700 and k5 - the latter is couple of times smaller - The point of the problem is that huge sensors surfaces in dim light below 4EV and f2.8 lens doesn't seems to see the contrast good enough. If the plane of the sensor is filled with big contrasty things, af is accurate even in pitch black if the contrasty things seen in the viewfinder are tiny like 1/4 of the sensor plane or less - even if the sensor sees planty of small contrasty point (but thay are too small) - than af error appears and we see nice pentax type bokeh on entire image.... ;-D The fact that Pentax uses f5.6 accurate af sensors instead of 2.8 like nikon does, adds some confusion to the af brain too + big dof on apsc especially with wide angle lenses (thats because we are waiting for Pentax FF - but for now on we have just ff problem, what means two things: lack of FullFrame and plenty of front focus errors) ;D

[godavari]

filorp, Thanks for the response. Does "AF Fine Adjustment" help offset the AF situation in any way in your experience?

[filorp]

what you mean godavari? af fine adjustment is supposed to help only when your lens has slightly shifted focus point generating constant and repeatable front or back focus...? at light conditions as i described: below 4ev and f2.8 lens focusing is so erratic and unpredictable (however always with some degree of front focus more or less) that aim not able to set af microadjustment.....