[crewl1]

In another thread a user commented that the DA*300 with HD1.4x combo did not AF-C well so I decided to do an AF test with K-3 this morning at a nearby bike path to check continuous Af with subjects advancing toward camera at good clip.

Tried a few combinations and found these to have best results for this appplication:
AF-C, Spot, First Shot Focus Priority, Continuous Action -Focus Priority, Hold Status Off

Lenses tried - Bigma 50-500, DA*300, DA*300+ HD 1.4x TC

Exposure mode used: TAv, f8, 1/500 (In retrospect should have used faster shutter speed as I think overall IQ suffered a bit from motion blur)

Overall the Bigma yielded most % in-focus shots in continuous bursts.

Both the DA*300 alone and with the 1.4x HD TC yielded similar % in focus shots, a bit lower than the Bigma.

I suspect this may be due to the SDM motor not reacting as fast to AF commands from the K-3 as the Sigma HSM system.

The DA*300 alone or with the 1.4 would sometimes lose a shot or two focus in a sequence, where the Bigma was more consistent.

Most common loss of focus was as the riders got closest the DA*300 with or without the 1.4x , perhaps as the changes required were just too big for the SDM to keep up with.

The naked DA*300 yielded the sharpest photos, while adding the 1.4x gave a small drop in IQ, with the Bigma close behind.

All combos seem usable for this application on a sunny day, as the light drops the DA*300 alone will be best option.

Here are representative sequences.

Bigma @500mm 44 shot sequence, all in focus (not all frames fit in the LR4 display window.)




DA*300 33 shot sequence, 2 shots out of focus after it picked up the cyclist.




DA*300 + HD1.4x 46 shot sequence, 6 shots out of focus.




I used a monopod to help keep the lens pointed accurately, I am sure some mis-focus was due to my moving off the target in spot mode.
I tested 9 point center bias and 9 point auto, but sometimes the focus would lock on another nearby object, it worked best if I tried to keep the cyclist in the center.

---

Last edited by crewl1; 09-27-2014 at 02:47 PM.

[danny09]

Great series. I have the exact same settings for AFC. They work fantastically well!

[JinDesu]

Great post. Goes to show that the k-3 is pretty formidable when used by someone knowledgeable, despite what some people may say.

[jbinpg]

Thanks for this, Larry. Nicely done. But since this is the K-3, the whole shebang should be replicated but this time with the various Expanded Area settings explored.

Jack

[danny09]

I dont think anyone has done the "whole shebang" tests that you mentioned. Probably yourself?

[crewl1]

Thanks for this, Larry. Nicely done. But since this is the K-3, the whole shebang should be replicated but this time with the various Expanded Area settings explored.

Jack

I tried expanded area but for this application it seemed spot was best.
When I used expanded it would lock onto a jogger or other cyclist when subject came in proximity.

I also tried hold settings but they didn't let AF change fast enough, I think that is for when subjects are moving parallel to camera.

I could not find mind read setting in the menu, that would have been ideal here

[rawr]

I could not find mind read setting in the menu, that would have been ideal here

Absolutely. Telepathic camera control would be great.

Some pre-packaged AF.C usage case menu options like the 6 Canon has for the 7DII, with nice little icons, would be nice though.

Speaking of the 7DII AF, I've noticed some handy guides to using AF.C with the 7DII that have a lot of things that will look very familiar to K-3 users.

There is a handy overview of the differences between Zone AF and AF Point Expansion with the camera.

And there is a glossy 52 page EOS 7D Mark II AF-setting Guidebook [PDF 6MB] that sets out nicely how all of the AF features work. Once again, many of those Canon AF settings seem to work similarly to the K-3, so will probably help folks drive the K-3 better too.

It's a pity Pentax have never released anything as nice as this Canon guide for the K-3's AF.

[stevebrot]

It's a pity Pentax have never released anything as nice as this Canon guide for the K-3's AF.

I totally agree. Ricoh/Pentax could do its users a huge favor by providing guides for use of some of the camera's more useful features.


Steve

[MrNPhoto]

Great series. I have the exact same settings for AFC. They work fantastically well!

Except when they don't. I've been having real issues getting the K-3 to track runners, let alone cyclists.
https://www.pentaxforums.com/forums/172-pentax-k-3/272794-k-3-af-c-focus-problems.html

Same settings for AFC, focus priority for 1st release and continuous, center point, 25-point expansion, AF Hold off. Sigma 70-200 f2.8 HSM lens. Yet there are times when it simply does not lock on to the runner, and produces entire multi-shot sequences out of focus. I'm only taking bursts of 4 to 8 frames, not anything near 40+ frames, and it is as if the camera simply cannot keep up with the runner.

I'm aching to know what I need to do to get consistent AF-C tracking on a moving subject such as a runner, and haven't come anywhere close to this kind of performance.

--Mike

[crewl1]

Why do you want to use 25 point expansion?
In my tests even 9 point caused the AF to shift to objects in the proximity of the intended subject.
I used spot.

---------- Post added 09-30-14 at 07:03 PM ----------

I will test with my Sigma 70-200 next weekend if I can.

[danny09]

Except when they don't. I've been having real issues getting the K-3 to track runners, let alone cyclists.
https://www.pentaxforums.com/forums/172-pentax-k-3/272794-k-3-af-c-focus-problems.html

Same settings for AFC, focus priority for 1st release and continuous, center point, 25-point expansion, AF Hold off. Sigma 70-200 f2.8 HSM lens. Yet there are times when it simply does not lock on to the runner, and produces entire multi-shot sequences out of focus. I'm only taking bursts of 4 to 8 frames, not anything near 40+ frames, and it is as if the camera simply cannot keep up with the runner.

I'm aching to know what I need to do to get consistent AF-C tracking on a moving subject such as a runner, and haven't come anywhere close to this kind of performance.

--Mike

I can understand it may not have worked for you. Have you tried tracking only using Center point autofocus + Spot Metering? My subjects might not have been as challenging as yours. (i.e a subject running towards me). I have tracked birds, planes and humans. But I looked at my series and all have been at a distance plane to me. They werent coming at me. I will test out some subjects coming at me.
Also I had much more keepers using the 18-135 WR compared to my screw driven 50-200.

Why do you want to use 25 point expansion?
In my tests even 9 point caused the AF to shift to objects in the proximity of the intended subject.
I used spot.

---------- Post added 09-30-14 at 07:03 PM ----------

I will test with my Sigma 70-200 next weekend if I can.

I had good rate of keepers with 9-Point focus system with 18-135 and slightly worse rate using a screw drive lens with center point focus.

[LeRolls]

Some really handy info in this thread. May have to refer back now that my K-3 has arrived.

[crewl1]

This morning I tested the Sigma 70-200 2.8 OS, Pentax DA*50-135, and DFA 100 WR (just for kicks.)

I didn't use the monopod for these tests as the Sigma is the only one with a lens tripod mount.
It is possible the less steady hold on target may have affected the results as the first set of tests above yielded better results.

Setting for AF were same as the first set, but today I used shutter speed 1/1000 with f5.6 on TAv.

The Sigma did not do as well as I expected, consistently giving about a 25% miss rate. It seemed to frequently stay slightly back or front focused for a series of frames.
I may need to calibrate it, but in practice it has performed well.

The DA*50-135 did well even though the AF feels slower than the Sigma.
With the consistent biker pace and change in distance frame to frame it kept up well.

I was surprised by the DFA100WR as I thought it would be too slow, but the K-3 screw drive motor seems to pep this lens up quite a bit.
If it goes off it does rack all the way but the K-3 brings it back quickly. Once locked in it tracks well.

I believe my tests may suffer from too small a sample, and I also think the approaching cyclist subject probably does not stress the camera AF much as the speed and change in distance from frame to frame is consistent. It does give me an idea of what to expect from my K-3 and various lenses.

Results are representative of a few tests for each lens.

Sigma 70-200OS 19 misfocus out of 75 shot series





DA*50-135, 8 misfocus out of 71 shot series





DFA100WR, 5 misfocus out of 57 shot series

[reivax]

Just out of curiousity, I see a lot of you recommending the AF Hold off. Any reason for that? It seems like a great feature that I'd like to use. Have you guys found a lot of issues with it?

[crewl1]

I think hold is useful when you don't want the AF switching to a new target too quickly, such as when following a runner moving parallel to the camera between trees.
Hold delays the camera trying to adjust to focus on each tree, in the hopes that the runner will reappear before the camera moves the focus.

In the testing above I intend to always have the subject in view without interruption so hold-off let's the camera quickly adjust as the rider advances.

I also shoot my grandkids football and I need to test how Hold could help me there.
Currently using hold-off in that situation the AF loses track of the ball carrier if another player comes into the focus area.
There may be a Hold setting that will work better in this case.