Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
08-21-2016, 07:36 AM   #16
Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 26,233
QuoteOriginally posted by Ptitboul Quote
Using the last version of ExifTool (the one that knows how to decode AF information) I get slightly different results.
All three photo have
Focus Mode : AF-C (Release-priority)
AF Point Selected : Center; Single Point
AF Points Selected : 17
but only the first photo has:
AF Points In Focus :
The second and third photos have:
AF Points In Focus : 17
which means that the K-1 knew that the center point was in focus.
Thanks for the head's up. I am running ExifTool v10.04 (current April 29, 2016), which does a good job decoding AF information, but apparently not for the K-1. I tried to update just now, but apparently Queens University (host for the ExifTool project) is having a few IT problems and the server is unreachable.

As noted in other comments above, a sequence in burst mode with the system in auto-33 or zone select would have been a better representation of the AF tracking in AF-C.


Steve

08-21-2016, 08:05 AM   #17
Forum Member




Join Date: Sep 2006
Location: Central Plains of North America
Photos: Gallery
Posts: 50
QuoteOriginally posted by stevebrot Quote
As noted in other comments above, a sequence in burst mode with the system in auto-33 or zone select would have been a better representation of the AF tracking in AF-C.
Not intending to be argumentative, but a sequence in burst mode with the system in auto-33 or zone select would actually have been a better representation of the AF tracking in burst mode with the system in auto-33 or zone select. Which I agree would also have been interesting.
08-21-2016, 10:28 AM   #18
Site Supporter
GeneV's Avatar

Join Date: Oct 2008
Location: Albuquerque NM
Photos: Albums
Posts: 9,765
I have not done a scientific test, but did just complete a four day shoot of a dance camp. The K1 plus the screw drive Tammy 70-200/2.8 did everything it needed to satisfy me in AF-C catching people jumping around.
08-21-2016, 11:32 AM   #19
Forum Member




Join Date: Apr 2014
Posts: 54
QuoteOriginally posted by kenspo Quote
But since its you, i can give you my best settings for freeze movement: Spot metering, Singel point focus with focus priority
Thank you.

I guess that the metering mode does not have any influence on the behaviour of the autofocus, and you mention it because it is part of your usual settings.
My practice is to use TAv mode when I want to freeze the movement, and spot metering with a -1 to -2 EV correction when shooting artists indoor. I never used the possibility to link spot metering to the autofocus point, therefore usually I crop the resulting picture when I don't want the subject to be in the center of the resulting picture.

08-21-2016, 03:03 PM   #20
Site Supporter




Join Date: Feb 2015
Photos: Albums
Posts: 2,832
Speed isn't necessarily where Pentax K3 and Pentax K1 AF struggle. Usually, with enough speed, prediction works very well. Usually, the issue for the user is to keep the target under the AF area and this is not an easy task with sensor stab instead of optical stab. Usually, Pentax K3 and K1 AF is very bad at tracking slow motion subjects because subject motion give very poor accuracy on predictive af tracking. There is a zone of incompetence of the Pentax K3/K1 AF system and it is slow moving target too slow for AFC but too fast to be snapped for AFS. For really high speed targets, the limit is the lens actuation. You can do a trial for yourself on a relatively fast moving target in Z direction, you see that there is a range of speed of target where the camera is able to track continuously and smoothly.
08-21-2016, 04:28 PM   #21
Site Supporter
jlstrawman's Avatar

Join Date: Feb 2010
Location: Des Moines, Iowa
Photos: Gallery
Posts: 419
QuoteOriginally posted by stevebrot Quote
Thanks for the head's up. I am running ExifTool v10.04 (current April 29, 2016), which does a good job decoding AF information, but apparently not for the K-1. I tried to update just now, but apparently Queens University (host for the ExifTool project) is having a few IT problems and the server is unreachable.

Steve
Try here:

https://sourceforge.net/projects/exiftool/files/exiftool-10.25.zip/download

Hopefully the version you need.
08-21-2016, 05:52 PM   #22
Lens Buying Addict
monochrome's Avatar

Join Date: Mar 2008
Location: Kirkwood (St. Louis) MO
Photos: Gallery | Albums
Posts: 19,497
@biz-engineer Would a slow jogger on Z-axis fall in this tracking doughnut hole? I'm having a discussion on another thread who claims K-3 cannot track a slow jogger, therefore all Pentax AF is awful.
08-21-2016, 07:17 PM   #23
Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 26,233
QuoteOriginally posted by jlstrawman Quote
Thanks! It looks as if Phil Harvey moved ExifTool off the Queens University servers to SourceForge sometime this last week. At least the old site was still up as recently as August 16 according to the "wayback machine".


Steve

08-21-2016, 07:21 PM   #24
Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 26,233
QuoteOriginally posted by monochrome Quote
@biz-engineer Would a slow jogger on Z-axis fall in this tracking doughnut hole? I'm having a discussion on another thread who claims K-3 cannot track a slow jogger, therefore all Pentax AF is awful.
If the slow jogger is at 50 meters and I am shooting with a 28mm on FF, I am quite confident. The same slow jogger at 6 meters with a 135mm lens is another matter entirely. The problem with Z-axis discussions is that people are hung up on velocity without considering that the AF system does not measure or react to velocity.


Steve
08-22-2016, 01:02 AM   #25
Site Supporter




Join Date: Jun 2013
Photos: Gallery
Posts: 74
QuoteOriginally posted by stevebrot Quote
...The problem with Z-axis discussions is that people are hung up on velocity without considering that the AF system does not measure or react to velocity...
Steve,

I think this possibly nails the problem - it should measure velocity, so that it can predictively focus to where the subject will be at shutter release. Are you sure that the Pentax algorithms don't take account of vz? Or are the algorithms just not very good at it?

Regards,

Chris
08-22-2016, 01:23 AM   #26
Pentaxian




Join Date: Oct 2010
Photos: Gallery | Albums
Posts: 2,663
I was trying to shoot bursts, but they were driving too fast to have even that target in frame...So not just Oof shots, but empty ones too. It is cameras fault that I had to cherry pick these...ummm, not.




this one was more easy because he was wheeling after winning that race...

Just to mention that these were shot with that 55-300 WR and there were row of trees along that route...so there were also some bikes behind those trees, which made not here.

---------- Post added 08-22-16 at 12:04 ----------

QuoteOriginally posted by Grognard Quote
TAv mode, F4.0, 1/400 shutter, DA* 300 AF-C
No intention to smash DPReview's marks, but K-1's AF-C capability in capturing this little swift toy aircraft is satisfactory!
Pictures only cropped and shrunk.
Center image is sharp enough for DA* 300 even when wide open.





Forgot to mention that these are nice shots!
08-23-2016, 01:36 PM   #27
Site Supporter




Join Date: Feb 2015
Photos: Albums
Posts: 2,832
QuoteOriginally posted by stevebrot Quote
The problem with Z-axis discussions is that people are hung up on velocity without considering that the AF system does not measure or react to velocity.
I'm not so sure about that, especially when AF tracking is predictive such as base on a PID controller, which responds proportionally to AF error for focusing on a static subject, integrate the error to produce an acceleration of the lens actuation (so that to reduce the time to move AF ring from current position to target) and also the derivative of error variation in case of moving subject. Using the same focal length and subject distance, if the subject is moving away or towards the camera, the derivative component of the lens drive is proportional to the speed of the subject, but, if the motion is very slow the derivative is also very small signal relative to other terms, therefore, tracking is not sensitive. If the motion is fast, the derivative amount applied on lens drive is very strong be actuation torque is limited by current driver and lens motor and mass of hardware in the lens. If the way I understand is correct, then, any AF system is limited, and using it with proper setting helps. For example, for high speed subjects, I'd recommend to select FPS priority because... even if the percentage of keeper is smaller (ration between in focus and out of focus shots), the absolute number of in focus shots is higher with FPS prio because in that case, the AF servo bandwidth is higher. But for slow motion, it is the other way around and down to the point where pressing the shutter button fully in AFS mode is more effective than AFC (here I'm referring to people having had blurred wedding photos when shooting in AFC mode with only very little motion of the people being photographed).

Last edited by biz-engineer; 08-23-2016 at 01:46 PM.
08-23-2016, 01:53 PM   #28
Site Supporter
6BQ5's Avatar

Join Date: Jun 2013
Location: Nevada, USA
Photos: Albums
Posts: 3,064
Alrighty ... So let me throw some grains of my own experience here.

I buckled and bought a K-1. It is configured to engage focus with with the shutter release button. Shutter is set to focus priority. I stood on the side of the road photographing cars and trucks whizzing by. This motion is a combination of the car coming toward me and by me. The K-1 will only fire when it confirms focus. I could notice the erratic frame rate because the camera needed time to adjust focus. Otherwise I got crisply focused photographs of wheel wells, mirrors, and door handles.

I did cheat a bit. My lens was set to f/8 so I had what felt like massive depth of field helping me out. I mean, I had nearly the whole car in focus from bumper to bumper.

Focusing with back button and disabling the focus/shutter link with the shutter release button resulted in out of focus pictures. The shutter will fire whenever the shutter release button is pressed. Synchronizing focus confirmation is the photographers responsibility.


08-23-2016, 02:37 PM - 1 Like   #29
Pentaxian




Join Date: Sep 2008
Location: Phoenix, AZ
Posts: 1,417
so many threads, so many arguments about pentax AF.

and then there are those who want a test to conform to their methodology to justify their thoughts processes.

i don't shoot like you do, but i get good results. why should i shoot like you and get the same bad results just to confirm your own bias?

seriously?

I'm not real fond of the K1 multi-point algorithm. Wasn't fond of the K20 or K5iis or K3 either. But then again, most Nikon shooters i hang out with don't use multi-point either, they rely on their own tracking skills to keep the center point dot on target.

If you really want to include composition elements, use one dot offset right or left via point select.

When shooting burst mode, the results are entirely dependent are YOUR ability to track a subject during the frame blackout (shutter up) because the AF sensor is in the dark as well. Nikon does this better than any other manufacturer, but the weakest link is STILL the photog. Why do you think Nikon went to 99+ points, because the average person is horrible at tracking. How good are you? Go to a skeet range with a 20 gauge and find out.... it's the exact same technique!

Secondly, focus priority is fine if you want slower burst rate or you only want one to three frames per second. For me, release priority works so much better, because the camera more often that not catches up to the subject.... and this is key... if your tracking skills are good enough.

For really tough subjects, like diving birds, hummingbirds in flight and jets at an airshow, I actually shot AF-S and feather the shutter. I probably refocus 3-4x a second while tracking. I get nearly 100% hit rate. I actually turn off the AF <beep> because it annoys the heck out me, sounds like an alarm clock on speed...lol

My point is, if someone posts a thread with a good action sequence, learn from them, If you like the results, learn the technique. Who cares if they cherry picked, or used AF-S or AF-C or AF-A, back button or half press, focus or release priority. I'm looking at you mr
QuoteOriginally posted by enoeske Quote
nothing here shows a sequence of AF-C shots. These are cherry picked shots that happen to be in focus.
I hope that was sarcasm...

Seriously, people, let the DPR thing die. That horse is bone meal.

Instead of "I'm happy with the K1 AF-C." How about "Here's my pic and here's how I did it. You can too."?

Get over your inferiority complex pentax shooters.

Am I 100% thrilled with the K1 AF-C? No. But it's a hell of a lot better than the K3 could ever be.
Is the K1 a D5? No. But I know of at least one d810 owner who's jealous as hell i paid 1/4 what he did and i can match his output pixel for pixel.

Last edited by nomadkng; 08-23-2016 at 02:49 PM.
08-23-2016, 03:29 PM   #30
Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 26,233
QuoteOriginally posted by biz-engineer Quote
I'm not so sure about that, especially when AF tracking is predictive such as base on a PID controller, which responds proportionally to AF error for focusing on a static subject, integrate the error to produce an acceleration of the lens actuation (so that to reduce the time to move AF ring from current position to target) and also the derivative of error variation in case of moving subject. Using the same focal length and subject distance, if the subject is moving away or towards the camera, the derivative component of the lens drive is proportional to the speed of the subject, but, if the motion is very slow the derivative is also very small signal relative to other terms, therefore, tracking is not sensitive.
Are you making this stuff up? I believe you are stating multiple unsupported assumptions starting with the first phrase.


Steve
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
af-c, da*, dslr, flickr, full frame, full-frame, k-1, k1, pentax k-1, shots, trees
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Japanese blogger tests K-3 II AF.C performance against K-3: Measurable improvement bwDraco Pentax K-3 17 05-01-2016 10:04 AM
K-3's AF.C with Focus priority. skyer Pentax K-3 18 12-11-2014 05:44 PM
Pentax K-x: focus confirmation in AF-S, AF-C, not AF-A manteiv Pentax DSLR Discussion 24 12-10-2010 07:25 PM
For Sale - Sold: s-m-c takumar 85mm/f1.9, s-m-c 35mm/f2, smc-k 50 1.2 and more zhha Sold Items 14 10-03-2010 05:24 PM
K-7 and AF-C performance? Clem Nichols Pentax DSLR Discussion 16 07-20-2009 12:50 AM



All times are GMT -7. The time now is 06:42 AM. | See also: NikonForums.com, part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top