K20D AF tracking: a non scientific test

lol101 · 04-16-2008, 03:27 PM

Today, I stopped my car on the side of a small road (speed limitation 90km/h) and took 128 pics of various vehicles coming towards or away from me with my new K20 and the old FA135f2.8.

Results: 3 oof pics, 2 slightly oof, the rest is sharp.

Some moths ago, I did the same with the K10 and the same lens and had about 25% hit rate...

I think the K20D is more than slightly improved in the AF dept. compared to K10, anyone has the same conclusion?

Here are thumbnails of the shots I took:

lol101 · 04-16-2008, 03:56 PM

Here are some example of 100% crops of the center portion to explain what I judged to be oof, slightly oof and sharp:

1 oof, 2 is the following shot and appears sharp to me.

3 is slightly oof compared to 4 (last shot of the series)

5 is the last shot of the bus series and is quite sharp (as are all the other ones of this and most of other series).

As you can see, the FA 135 suffers quite a bit of CA and PF but remember that these are unprocessed jpegs.

Also I shot all these in D-range mode (hence the slight grain in shadow areas).

*Mallee Boy* · 04-16-2008, 04:25 PM

The oof shots were probably from people hard on the brakes thinking you had some other sort of hand held device

Betcha you sent a few heart rates up!!

Anecdotal evidence says that the AF is faster on the K20D than previous models, which is what you would hope for on any new/upgrade model of anything.

Interesting test though, thanks for taking the time to do it!

jgredline · 04-16-2008, 05:08 PM

Take a look here....
K20D + DA* = impresive

*Jonson PL* · 04-16-2008, 08:57 PM

Originally Posted by lol101

Here are some example of 100% crops of the center portion to explain what I judged to be oof, slightly oof and sharp:

1 oof, 2 is the following shot and appears sharp to me.

3 is slightly oof compared to 4 (last shot of the series)

5 is the last shot of the bus series and is quite sharp (as are all the other ones of this and most of other series).

As you can see, the FA 135 suffers quite a bit of CA and PF but remember that these are unprocessed jpegs.

Also I shot all these in D-range mode (hence the slight grain in shadow areas).

Hi Laurent, thanks for taking the time and posting. Seems they worked on, and improved some of the areas that the K10 was lacking.

I was so happy with my K10, now you’re all showing me what the K20 can do

What are you're thoughts on the Dynamic-range mode ?

kenyee · 04-16-2008, 09:11 PM

Must...not...buy K20D... :-)
Thanks LOL101.

ftpaddict · 04-20-2008, 02:03 PM

And now we know your license plate.

lol101 · 04-21-2008, 03:49 AM

Originally Posted by ftpaddict

And now we know your license plate.

... I am still unable to drive and take pictures of my licence plate at the same time

ftpaddict · 04-21-2008, 03:54 AM

Originally Posted by lol101

... I am still unable to drive and take pictures of my licence plate at the same time

Woops, my bad. I misread the writing on the first picture of the Scénic. I thought it said "K10D" and I thought it impossible to capture the same car from the same angle at almost the same time.

xGene · 04-21-2008, 02:30 PM

I knew it!!!!! Have both K10D and K20D and can also confirm (done similar test) that K20D has MUCH better tracking ability in continuous focus mode, at least in NORMAL DAYLIGHT.

I guess Pentax decided not to mention anything about it, it would be like admitting the K10D was inferior...

Thanks Lol101 for the confirmation.,

Gene