[monochrome]

but wait, did you set your camera on a tripod and have the motoXers go out of the AF area? Surely this is not a good test. Actually using your skills to track the subject - how absurd!

/end sarc

A test is not a test unless the aperture is f/2.8. Otherwise the plane of focus is far to deep to actually test the AF.

[totsmuyco]

First and foremost:
Think about what you want to focus on. It should make sense. It should have some good contrast. And it should ne be hidden behind some other things too much. I settled for chest of the driver or starter number plate.
Then keep pointing the combo at this as good as you can.

A monopod helped with some shots, but is no use for others like jumps.

Second:
I made good experiences pressing the AF-on button on the lens with my left thumb as soon as I decided on a target subject, so the lens could start computing the AF.C. And 0,5-1 seconds later when the driver was where it got interesting, push the shutter button to fire.

Other:

1. AF.C
2. Sel-1 or Sel-9 (since the head of the driver is on the upper end I used the topmost AF point in the center column). If I had use a F2.8 lens like the 70-200, I'd used only one step upwards as only those middle three sensor are F2.8 enabled.
3. Hold:Low
4. AF.C priority 1st image: auto
5. AF.C priority other images: auto
6. JPG L*
7. I explicitly turned off absolutely all other JPG corrections or automatisms.
8. Used TAv mode all the time to set the aperture to open and set a certain time. Shutter times I experimented with anything between 1/320 and 1/4000 but mostly was stuck with very short ones due to the light and not having brought some NDs or at least a CPL. Generally I do not like to completely freeze dynamic subjects, so I prefer some blur on the wheels and with the dirt.

Thanks for the tip. I'll give it a try on my k-3II. I know it's not at par with the k-1 but it would be worth looking into.

[jpzk]

First and foremost:
Think about what you want to focus on. It should make sense. It should have some good contrast. And it should ne be hidden behind some other things too much. I settled for chest of the driver or starter number plate.
Then keep pointing the combo at this as good as you can.

A monopod helped with some shots, but is no use for others like jumps.

Second:
I made good experiences pressing the AF-on button on the lens with my left thumb as soon as I decided on a target subject, so the lens could start computing the AF.C. And 0,5-1 seconds later when the driver was where it got interesting, push the shutter button to fire.

Other:

1. AF.C
2. Sel-1 or Sel-9 (since the head of the driver is on the upper end I used the topmost AF point in the center column). If I had use a F2.8 lens like the 70-200, I'd used only one step upwards as only those middle three sensor are F2.8 enabled.
3. Hold:Low
4. AF.C priority 1st image: auto
5. AF.C priority other images: auto
6. JPG L*
7. I explicitly turned off absolutely all other JPG corrections or automatisms.
8. Used TAv mode all the time to set the aperture to open and set a certain time. Shutter times I experimented with anything between 1/320 and 1/4000 but mostly was stuck with very short ones due to the light and not having brought some NDs or at least a CPL. Generally I do not like to completely freeze dynamic subjects, so I prefer some blur on the wheels and with the dirt.

Seems like good technique here and ... did you have the SR ON or OFF ?
Just curious.

[krazykat]

but wait, did you set your camera on a tripod and have the motoXers go out of the AF area? Surely this is not a good test. Actually using your skills to track the subject - how absurd!

/end sarc

Haha, that's right! And to be a "standard" test it's really important that you don't test all other cameras the same way. Otherwise it wouldn't be "standard".

[Gimbal]

A test is not a test unless the aperture is f/2.8. Otherwise the plane of focus is far to deep to actually test the AF.



:brap:

Is that supposed to be sarcasm?
Do you calibrate your lenses stopped down? Of course the lens should be wide open if you’re testing the AF.

[beholder3]

Seems like good technique here and ... did you have the SR ON or OFF ?
Just curious.

SR was set to OFF. As I was swinging around the camera all the time it would not help and I wanted to switch everything off that would use valuable computing power/time.
I don't know at all if it helped. The whole event was about trying and learning after all.

[redpit]

SR was set to OFF. As I was swinging around the camera all the time it would not help and I wanted to switch everything off that would use valuable computing power/time.
I don't know at all if it helped. The whole event was about trying and learning after all.

I'm still learning too* but I agree with SR turned OFF for shutter speeds (SS) 1/500 and higher. Between SS 1/160 and 1/400 SR on and off gives more or less the same results (% of keepers) on moving targets. Panning with and without SR also gives various results and mainly depends on your technique but more or less the same % of keepers. On stationary subjects SR is better to turned on even on higher SS. That's my experience up to now... and we are talking about the DFA 150-450 not other lenses.

*I'm talking mainly about shooting airshows, I haven't got any experience on shooting motocross or car racing yet.

PS: Pay attention not to try to shoot slowly moving bicycles... You will be dissapointed

---

Last edited by redpit; 08-23-2016 at 01:34 AM.

[FantasticMrFox]

As one of the whiners about Pentax AF.C I had a look at several of the posted sequences and I have to say I am very impressed. The riders are consistently sharp and in focus.

Now if I could only get my K3 to do this ...

[clackers]

As one of the whiners about Pentax AF.C

Yes, you were, FMF.

In the interests of being fair and balanced will you now go back to DPR and whine about Alex Cooke's methodology?

No, I didn't think so. 😀

But you did it to us Pentax AF-C shooters.

[monochrome]

A test is not a test unless the aperture is f/2.8. Otherwise the plane of focus is far to deep to actually test the AF.

Is that supposed to be sarcasm?
Do you calibrate your lenses stopped down? Of course the lens should be wide open if you’re testing the AF.

On a tripod.

[beholder3]

I'm still learning too* but I agree with SR turned OFF for shutter speeds (SS) 1/500 and higher. Between SS 1/160 and 1/400 SR on and off gives more or less the same results (% of keepers) on moving targets. Panning with and without SR also gives various results and mainly depends on your technique but more or less the same % of keepers. On stationary subjects SR is better to turned on even on higher SS. That's my experience up to now... and we are talking about the DFA 150-450 not other lenses.

Agreed.

Since my panning tries were done capturing jumping drivers, SR could not help (I think), as the subject moved in a curve, not a straight line.

---------- Post added 23rd Aug 2016 at 13:46 ----------

As one of the whiners about Pentax AF.C I had a look at several of the posted sequences and I have to say I am very impressed. The riders are consistently sharp and in focus.

Thanks for your feedback. Admitting that lifts you over the level of others who only complain and never step even a tiny bit away from their prejudices and predefined "truths". Discussion can only exist where people are open to changing their views.

[redpit]

Agreed.

Since my panning tries were done capturing jumping drivers, SR could not help (I think), as the subject moved in a curve, not a straight line.

I copy from the Ricoh official page here: K-1 | Ricoh Imaging

"Even when taking a panning shot, this system automatically detects the direction of the camera’s movement, and efficiently controls the SR unit to always produce the best image possible."

That made me trying both SR on and off panning shooting to conclude what works best for me. I am more confused than before after shooting two events

What I can verify though is that K-1 SR surely allows for some perfect panning shots when K-5IIs SR didn't. I don't know if it detects the horizontal panning movement and works in vertical stabilization mode only and I don't know if it can detect panning movement when handholding the camera (because slight vertical shaking-movement is inevitable).

[beholder3]

I think SR will be fine when panning say a car on a street passing by, because that is "just" horizontally.
The IS modes on lenses work just that way. They allow panning strictly horizontally on their special modes.

[FantasticMrFox]

Yes, you were, FMF.

In the interests of being fair and balanced will you now go back to DPR and whine about Alex Cooke's methodology?

No, I didn't think so. 😀

But you did it to us Pentax AF-C shooters.

In my defence, I never claimed with absolute certainty the K-1 can't do it. I said that the K3 cannot do it (and I still assume it can't, because in contrast to the K-1 I actually own the K3) and that because the AF.C behaviour DPR reported in the K-1 closely mirrors that of the K3, I assumed they are right.

And because I still really want my K3 do be able to do it I am still receptive to people doing the same with the K3 Beholder did with the K-1, i.e. demonstrate the K3 can do it and show me how to do it.

[Barry Pearson]

A test is not a test unless the aperture is f/2.8. Otherwise the plane of focus is far to deep to actually test the AF.

I know that was sarcastic, but I've given this some thought.

I've seen a request here for something like a long burst of a fast moving object coming towards the camera, to show whether or not all the frames are in focus. (The assertion being that only a few will be).

A long burst (say 12 or more frames) on the K-1 takes about 3 seconds or so. A fast moving object (for example, a racing car or bike) can go a long way in that time.

If the last frame, where the object is probably fairly close and fairly large in the frame, and so one of the most likely to be a keeper, is to be in focus over the whole of the object (car, say) being photographed, it typically needs a smaller aperture than f/2.8. Since I use TAv for this sort of work, then that smaller aperture applies throughout, including the first frames. But the first frame has the object a long way away and small in the frame. Even though I'll probably be using 200mm for the first frame, there will probably be plenty of depth of field. So that smaller aperture won't be much of a test of AF over much of the burst.

In other words, the way I set and use the K-1 for fast moving objects in the field is significantly different from the way someone would set and use the K-1 to photograph a slow moving object for a review. And vice versa.

How relevant is a stringent test of a slow moving object (for a review) for indicating how good the camera is for shooting fast moving objects in real-world situations?