[Qwntm]

I think the tone of the thread title sort of gave that impression

That aside, this thread has carried a very constructive tone and provided opportunity for people to share their insights on what works and what doesn't.


Steve

I wanted to catch people's attention without undo drama... I could have just titled it "PENTAX IS DOOOMED". If I wanted to maintain the status quo.


Yea, I guess the title is a bit provocative...

[stevebrot]

"PENTAX IS DOOOMED".

Now THAT'S THE SPIRIT!!!


Steve

[bobbotron]

I had some of focusing issues, I found really keeping the center focus actually on the target of focus instead of near it to be a huge help - I think what can happen is that you get a bit complacent and end up aiming it at the ground near the target of motion, especially when it's filling the whole frame. When I try harder and keep the center on the subject, my hit rate goes way up. It is mentally tiring.

[Srhphoto]

Ok, I did some experimenting at my local racetrack with the AF-C ability of the K3 with my Sigma 150-500mm lens: -



I used SEL9 with AF hold set to HIGH. The car was doing around 70mph, but the slight left to right motion will have reduced the speed relative to my shooting position by some degree.

The combination can do it, but with an atrocious keeper rate (less than 30% for what is not a difficult shot), so I certainly wouldn't want to be relying on it if I was a working professional. I normally pre-focus these types of shots, and to be honest will probably continue to do so the majority of the time.

Cheers,

Simon.

[Qwntm]

Changed a few settings and now I can't get a shot out of focus... This shot was towards the end of a string of 15 on continuous medium speed. Focus is spot on and the camera has tracked since the dog was 10 from the far fence.





So basically I set AFC, 9 point, 1st shot focus priority, focus priority high. Worked like a charm. I'm 100% confident shooting a dog at full sprint coming towards me with the K3II and 70mm HD LTD. So chalk this one up to pilot error, or lack of experience/knowledge. The camera was spectacular to watch, the AF points were lighting up like a Christmas tree.


I assume the K3 would behave similarly.

[ScooterMaxi Jim]

Well done.

The K-3 is a real challenge to figure out when you think about so many viewfinder options, as well as menu options. Some folks swear by single point selection (among the three center highest accuracy points), others swear by 27-point selection, and you've shown that 9-point can work fine. It really depends on anticipation of the relative size of the object you are trying to track.

One thing seems certain - focus priority high is an essential selection for good tracking. Also, if the first shot isn't in focus, the camera is highly unlikely to catch up with the object. But as the Flores series shows, the camera can lose an object for a frame or two but regain focus a few frames later - which is pretty darn impressive. I've also been impressed by the camera's ability to track using an older (very high quality) slow zoom - f/5.6 in relatively low arena light.

[johnmflores]

But as the Flores series shows, the camera can lose an object for a frame or two but regain focus a few frames later - which is pretty darn impressive.

To be clear, I am not sure if the camera re-found the subject or if I released and re-repressed the AF to regain focus.

[slip]

Edward: did you run any shots with a narrow DOF to see if the tracking is accurate? It seems like the shot you posted has a very wide DOF which helps keeping the subject in focus

Thanks

Randy

[ScooterMaxi Jim]

Either way, it tracked nicely (responding to Flores). I do recall a review somewhere with a bicyclist moving toward the camera where the tracking was similar to what you showed. I have no idea if that one was a situation where the AF was re-pressed, but it seems unlikely given the speed of the rider nearing the camera.

[Qwntm]

Edward: did you run any shots with a narrow DOF to see if the tracking is accurate? It seems like the shot you posted has a very wide DOF which helps keeping the subject in focus

Thanks

Randy

Hi Randy, the shots that the tracking worked perfectly on were shot at F8. The shot below in which not 1 photo out of 10 was sharp except maybe the first one were shot at F6.3. So I really don't think a 1/3 stop difference in DOF has any bearing, and you can tell by this photo that there is a sharp plane of focus, it's just 15 feet behind Chloe the Wonder Puppy.


I guess to truly test the system these should be shot at 200mm and F2.8. Maybe next time...




---------- Post added 07-07-15 at 03:38 PM ----------

Either way, it tracked nicely (responding to Flores). I do recall a review somewhere with a bicyclist moving toward the camera where the tracking was similar to what you showed. I have no idea if that one was a situation where the AF was re-pressed, but it seems unlikely given the speed of the rider nearing the camera.

I guess to really test all this I need to whip out the 200mm and shoot at 2.8. I only have the Tamron 70-200 and it's not bad, but I'd love to test this with the DC Pentax 70-200, that would only be fair if we were going to make any comparison's to Canikon.


I just have this suspicion that if you put a Tamron 70-200 screw drive on a Nikon D7200 or Canon 7DII, you would probably end up with similar performance to the Pentax.

[gebco]

Can someone clarify how Focus hold works? My understanding was that if subjects were moving laterally and went say behind a tree, the camera would hold the focus on the plane of the subject.

Does it also work for objects moving toward the camera? My understanding is that it isn't designed for that.

[ScooterMaxi Jim]

Can someone clarify how Focus hold works? My understanding was that if subjects were moving laterally and went say behind a tree, the camera would hold the focus on the plane of the subject.

Does it also work for objects moving toward the camera? My understanding is that it isn't designed for that.

That's a good question. I think the former is its primarily design intention, but the latter almost certainly gains at least some residual benefit. All I can say is the K-3 is far superior to earlier designs when you have a dog running toward you - even when shooting with a slow legacy zoom. Same thing can be said for relatively low light sports action. The overall AF improvements over previous models are not well appreciated. I think the B-I-F crew has noticed, though.

[gebco]

To my logic, if it is designed to hold a lateral plane in focus, it should actually be slower to refocus on an object approaching.

That's a good question. I think the former is its primarily design intention, but the latter almost certainly gains at least some residual benefit. All I can say is the K-3 is far superior to earlier designs when you have a dog running toward you - even when shooting with a slow legacy zoom. Same thing can be said for relatively low light sports action. The overall AF improvements over previous models are not well appreciated. I think the B-I-F crew has noticed, though.

[audiobomber]

To my logic, if it is designed to hold a lateral plane in focus, it should actually be slower to refocus on an object approaching.

Except that it also actually tracks, i.e. recognizes the subject by shape and colour , and anticipates its position via algorithm.

Real Time Scene Analysis with 86,000 pixel sensor

The new Real Time Scene Analysis System features an 86,000 pixel RGB light-metering sensor and improved algorithms that further enhance accuracy of exposure, white balance and autofocusing.
PENTAX K-3?Cameras?RICOH IMAGING CANADA

[gebco]

Thanks @audiobomber, I'll try that setting and see how it works.