[yorik]

For lenses with built-in motor, the AF speed on the K-3 will be about the same as on other cameras, since the camera has no control over the motor. However, screwdrive lenses are considerably faster. Any DA or FA limited (except for the macro) focuses almost instantly

Is the screw-drive motor in the K3 stronger?

[DSims]

Is the screw-drive motor in the K3 stronger?

Absolutely. It was improved in the K-5 II/IIs, and then the K-3 made it even better. Pentax "had" to do this, because they plan on keeping many screw-drive lenses (like the Ltds) in their lineup for a long time.

[Adam]

Is the screw-drive motor in the K3 stronger?

Yes. The speed speed improvement is around 50% as measured by these tests.

My only concern is if the added torque might be harmful to lenses with a short focus throw. For example, the 40mm really feels like it's being knocked around when it hunts- it goes from minimum to infinity in the blink of an eye.

[K10Perry]

I have an old Tamron 90mm/2.8 macro. It was just functional on my K-7. I have not mounted it on my K-3 until yesterday and WOW! It is like night and day, how fast it will focus on the K-3. I was thinking about dumping it but not now, I'm keeping it.

[crewl1]

Here is a test series I did at my granddaughter's softball game today where I used AFC to shoot her running from third to home.
18 shots in 4 seconds, this equates to 4.5 FPS so the camera does slow down a bit to focus, I have it set to focus priority.
I used the lowly Tamron 70-300 which is a lens I like to take to the softball field as it is light and has good reach and is sharp @ f8 and works fine outdoors in good sun.
The settings were TAv, 70mm, f8, (ISO 800) 1/1000 and I had it set for spot focus.
I was pleasantly surprised that every shot is in focus on the entire series, as she coming straight at me for 90 feet in 4 seconds.

I down sized, then put all the shots in powerpoint to show the collection so the series displays here look muddy, but every shot looks in focus and sharp.
The last picture is one I picked from the middle of the series but they all look right on the money.
I did not process the pictures except to crop and resize for posting.
The Tamron is not the greatest lens but It gives decent results, and post processing improves the output quite a bit.
My impression is that the AF in the K-3 is very effective, and definitely an improvement over my previous K-5.

[awaldram]

Yes. The speed speed improvement is around 50% as measured by these tests.

My only concern is if the added torque might be harmful to lenses with a short focus throw. For example, the 40mm really feels like it's being knocked around when it hunts- it goes from minimum to infinity in the blink of an eye.

Does that show any improvemnt in motor capability ?

Or does it show that the K3 has a cpu 4 time faster than the k5 series so reacts faster ?
As PD is a feedbacked focusing system processing time will have at least as much effect on focus speed as physical movment times.

Given the k5 is quite capapable of stripping gears in lower quality lens then why with the K3's super torqued up motor have we not seem a raft of failing gear trains ?

[DSims]

Does that show any improvemnt in motor capability ?

Or does it show that the K3 has a cpu 4 time faster than the k5 series so reacts faster ?
As PD is a feedbacked focusing system processing time will have at least as much effect on focus speed as physical movment times.

Given the k5 is quite capapable of stripping gears in lower quality lens then why with the K3's super torqued up motor have we not seem a raft of failing gear trains ?

Oh, I don't know, perhaps camera manufacturers actually tell the truth about hardware upgrades in their products? Something that can be easily proven by looking at it. Perhaps Adam's a little bit informed? I suppose he has trouble keeping track of Pentax' extensive lineup of 3 K-mount DSLR models?

Furthermore, simple logic tells you there has to be more to it than the CPU. How has Canon had fast AF performance for so many years with the weaker CPUs of the day? Even the most crude CPU can tell the motor to move the lens focus from where it is to where it thinks it should be, at full speed. A slow CPU doesn't inherently slow this motor down. Perhaps the faster CPU is the very reason the new motor's not stripping lens gears, as you pointed out. It could better control and brake it.


For our next topic, I propose we discuss whether anyone's actually been to the moon.

---

Last edited by DSims; 04-13-2014 at 12:34 PM.