I'm getting sick and tired of unreliable AF (rant)

Anvh · 07-05-2012, 04:05 PM

ah right... so wait for the K5 to get back then and see how that works. Hopefully they improved the issue.

Zafar Iqbal · 07-05-2012, 07:25 PM

Yeah, I hope so too.

A fluked zoom lens test lead to this idea:

You can view each row in bigger sizes from the album - camera models are listed underneath the images:

The K-r focus fluctuate a lot.

The K-x on the other hand is much more even.

That fits very well with what I have been claiming. There is more to AF of cause, but since the results reflect my experience this much, I feel there very well may be something to it. This series might also explain the focus shifts I noticed as the day went. Be cause as we know, the overall color changes too. The K-r have so far behaved best in overcast. Overcast = dull colors.

stevebrot · 07-05-2012, 10:24 PM

Originally posted by Williunck

Lol did I say that? My point is some professionals depend on it to be accurate, and yes I am sure it fails them from time to time, but on the street in good light with time to compose a shot it should be spot on more often then not.

Consider the relative size of the subject in your average sports photo. Not much to miss here. 2 feet of DOF pretty much covers it for most shots and accuracy is usually not as big an issue as responsiveness. If the system can lock fast enough, chances are focus will be quite good since most sports venues are VERY well-lit and the subjects generally are well-defined and have high contrast.

As for composing the shot in action sports...the pros anticipate the action, wait for it to come to them, and then take their best shots and pray that the AF system was able to track the subject. Now Pentax pretty much sucks at tracking and predictive AF, but I suspect that trap-in-focus would yield some pretty acceptable results from a focus perspective. (I have seen some good examples of this being done with motorsports.)

Steve

stevebrot · 07-05-2012, 10:37 PM

Found it!!

This little animation and the accompanying text give one of the best descriptions of how the AF system works. Apply a little abstract thought and visualization of the real world and it is easy to see why even the most accurate AF systems will "miss" focus on subjects with complex depth profiles and why light intensity and contrast are soooo important.

Autofocus: phase detection

Steve

Anvh · 07-07-2012, 06:51 AM

Originally posted by stevebrot

Found it!!

This little animation and the accompanying text give one of the best descriptions of how the AF system works. Apply a little abstract thought and visualization of the real world and it is easy to see why even the most accurate AF systems will "miss" focus on subjects with complex depth profiles and why light intensity and contrast are soooo important.

Autofocus: phase detection

Steve

Then you throw this in the mix making it all more complicated.

Spherical aberration

Zafar Iqbal · 07-16-2012, 05:25 AM

I attended two weddings since the last time I wrote. K-x for everyday usage is still great, but I need a more capable camera for weddings and other critical stuff where there aren't many chances, if at all. K-r would have been better for the job. It keeps focusing for much longer (low light) and continuous focusing is better as well - but (my borrowed sample) is a hell to use. I had it stopped a tad down during the first wedding and fought with micro adjustment all the time. Was stressing and the outcome somewhat OK - but not good for the critical eye. On the second wedding I left it at f8 (had more light) and only used 21mm lens. Didn't pay as much attention to micro adjustments and the outcome was worse.

The success rate of the K-r equaled my manual focusing using a sloppy technique: burst fire while turning the focus ring ever so slightly.

I went to buy another K-5 some days ago (they let me try it at home before paying for it. Very nice of them). It was like the first one I bought. Good during the day and front focus in low/artificial light. Would still love to have two K-5s but I've now tried every single camera from the shop. K-30 and Photokina seems to be the only things left for now.

The local shop *should* be receiving K-30 samples later today. I'm very curious to see how it will be. Current posts about it look very promising.

Zafar Iqbal · 07-25-2012, 02:30 PM

Tried the K-30 and the focusing was rock solid. Took the chance and bought another K-5 though as the K-30 is a lesser camera for my current needs. The much quieter shutter alone is a big deal for stuff like wedding photography.

Bought it yesterday and tried it a bit during late evening/night, and then a lot more today. The impression I got was that it's a much better sample than any of the ones I tried before - was kind of ready to settle with it despite some of the issues. Then the sun went lower and hell broke lose during the golden hour. Great! :/

Back to the K-30 then. A very good camera in some aspects - but not a K-5

Parallax · 07-25-2012, 02:55 PM

Your K5 had focus issues, you tested another one that had the same issues, you tested a K30 that focused perfectly, so the logical thing to do was buy another K5?

JinDesu · 07-25-2012, 03:14 PM

Originally posted by Parallax

Your K5 had focus issues, you tested another one that had the same issues, you tested a K30 that focused perfectly, so the logical thing to do was buy another K5?

That's how I choose girls, man.

Parallax · 07-25-2012, 03:18 PM

I can understand using that logic for girlfriends; but not for something important, like a camera.

Zafar Iqbal · 07-25-2012, 03:40 PM

lol

Edit: I forgot the most important piece of information. I made sure this time, that I could purchase it, and in case of problems, return and have the money refunded. Otherwise I would have gone for the K-30 immediately.

Also removed my original reply as it doesn't really matter. Most important reason is mentioned above.

bossa · 07-26-2012, 08:23 PM

I'm not sure if this has already been posted but I felt it might be of interest:
http://www.3dnews.ru/_imgdata/img/2012/07/13/632251/7.jpg

vonBaloney · 07-26-2012, 09:44 PM

Originally posted by bossa

I'm not sure if this has already been posted but I felt it might be of interest:
http://www.3dnews.ru/_imgdata/img/2012/07/13/632251/7.jpg

What do those categories mean? "5% <75% unacceptable" -- means what?

joe.penn · 07-26-2012, 10:02 PM

The categories mean exactly how they are labled - if 100 shots were taken for the test using each camera tested then 75% (or 75 of the photos) were focused properly, 23% (or 23 photos) completely missed focus, etc etc.. So, the higher the "on focus" rate, the more reliable focus is for that camera (not sure what lens was used but certain SDM lenses would give a drop in that "on focus" percentage for fast moving subjects). In short, the K5 does quite well against the competition.

vonBaloney · 07-26-2012, 10:35 PM

Originally posted by joe.penn

The categories mean exactly how they are labled - if 100 shots were taken for the test using each camera tested then 75% (or 75 of the photos) were focused properly, 23% (or 23 photos) completely missed focus, etc etc.. So, the higher the "on focus" rate, the more reliable focus is for that camera (not sure what lens was used but certain SDM lenses would give a drop in that "on focus" percentage for fast moving subjects). In short, the K5 does quite well against the competition.

I'm not getting it -- there are two percentages. K-5 has 23% "75-90% blurry" -- 23% of what? The total images, I assume, as you were saying -- that's fine and obvious -- all adds up to 100%. I just don't know the 3 categories for each camera mean. What is "75-90% blurry"? 75-90% of what? Not of the images, because that's 23%.

07-05-2012, 10:37 PM	#94
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Found it!! This little animation and the accompanying text give one of the best descriptions of how the AF system works. Apply a little abstract thought and visualization of the real world and it is easy to see why even the most accurate AF systems will "miss" focus on subjects with complex depth profiles and why light intensity and contrast are soooo important. Autofocus: phase detection Steve

07-07-2012, 06:51 AM	#95
Anvh Veteran Member Join Date: Sep 2011 Posts: 4,616	Originally posted by stevebrot Found it!! This little animation and the accompanying text give one of the best descriptions of how the AF system works. Apply a little abstract thought and visualization of the real world and it is easy to see why even the most accurate AF systems will "miss" focus on subjects with complex depth profiles and why light intensity and contrast are soooo important. Autofocus: phase detection Steve Then you throw this in the mix making it all more complicated. Spherical aberration

07-25-2012, 03:40 PM	#101
Zafar Iqbal Veteran Member Join Date: Oct 2010 Photos: Gallery Posts: 1,229 Original Poster	lol Edit: I forgot the most important piece of information. I made sure this time, that I could purchase it, and in case of problems, return and have the money refunded. Otherwise I would have gone for the K-30 immediately. Also removed my original reply as it doesn't really matter. Most important reason is mentioned above. Last edited by Zafar Iqbal; 07-25-2012 at 04:02 PM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
This one is just really sick	boriscleto	General Talk	2	10-24-2011 10:50 AM
Machinery Fully Sick	RichardS	Post Your Photos!	11	09-06-2011 05:08 PM
My K-7 is sick	SOldBear	Pentax DSLR Discussion	5	08-10-2011 10:27 PM

07-05-2012, 04:05 PM	#91
Anvh Veteran Member Join Date: Sep 2011 Posts: 4,616	ah right... so wait for the K5 to get back then and see how that works. Hopefully they improved the issue.

07-05-2012, 07:25 PM	#92
Zafar Iqbal Veteran Member Join Date: Oct 2010 Photos: Gallery Posts: 1,229 Original Poster	Yeah, I hope so too. A fluked zoom lens test lead to this idea: You can view each row in bigger sizes from the album - camera models are listed underneath the images: The K-r focus fluctuate a lot. The K-x on the other hand is much more even. That fits very well with what I have been claiming. There is more to AF of cause, but since the results reflect my experience this much, I feel there very well may be something to it. This series might also explain the focus shifts I noticed as the day went. Be cause as we know, the overall color changes too. The K-r have so far behaved best in overcast. Overcast = dull colors.

07-05-2012, 10:24 PM	#93
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by Williunck Lol did I say that? My point is some professionals depend on it to be accurate, and yes I am sure it fails them from time to time, but on the street in good light with time to compose a shot it should be spot on more often then not. Consider the relative size of the subject in your average sports photo. Not much to miss here. 2 feet of DOF pretty much covers it for most shots and accuracy is usually not as big an issue as responsiveness. If the system can lock fast enough, chances are focus will be quite good since most sports venues are VERY well-lit and the subjects generally are well-defined and have high contrast. As for composing the shot in action sports...the pros anticipate the action, wait for it to come to them, and then take their best shots and pray that the AF system was able to track the subject. Now Pentax pretty much sucks at tracking and predictive AF, but I suspect that trap-in-focus would yield some pretty acceptable results from a focus perspective. (I have seen some good examples of this being done with motorsports.) Steve

07-16-2012, 05:25 AM	#96
Zafar Iqbal Veteran Member Join Date: Oct 2010 Photos: Gallery Posts: 1,229 Original Poster	I attended two weddings since the last time I wrote. K-x for everyday usage is still great, but I need a more capable camera for weddings and other critical stuff where there aren't many chances, if at all. K-r would have been better for the job. It keeps focusing for much longer (low light) and continuous focusing is better as well - but (my borrowed sample) is a hell to use. I had it stopped a tad down during the first wedding and fought with micro adjustment all the time. Was stressing and the outcome somewhat OK - but not good for the critical eye. On the second wedding I left it at f8 (had more light) and only used 21mm lens. Didn't pay as much attention to micro adjustments and the outcome was worse. The success rate of the K-r equaled my manual focusing using a sloppy technique: burst fire while turning the focus ring ever so slightly. I went to buy another K-5 some days ago (they let me try it at home before paying for it. Very nice of them). It was like the first one I bought. Good during the day and front focus in low/artificial light. Would still love to have two K-5s but I've now tried every single camera from the shop. K-30 and Photokina seems to be the only things left for now. The local shop should be receiving K-30 samples later today. I'm very curious to see how it will be. Current posts about it look very promising.

07-25-2012, 02:30 PM	#97
Zafar Iqbal Veteran Member Join Date: Oct 2010 Photos: Gallery Posts: 1,229 Original Poster	Tried the K-30 and the focusing was rock solid. Took the chance and bought another K-5 though as the K-30 is a lesser camera for my current needs. The much quieter shutter alone is a big deal for stuff like wedding photography. Bought it yesterday and tried it a bit during late evening/night, and then a lot more today. The impression I got was that it's a much better sample than any of the ones I tried before - was kind of ready to settle with it despite some of the issues. Then the sun went lower and hell broke lose during the golden hour. Great! :/ Back to the K-30 then. A very good camera in some aspects - but not a K-5

07-25-2012, 02:55 PM	#98
Parallax Senior Moderator Loyal Site Supporter Join Date: Apr 2008 Location: South Dakota Photos: Gallery Posts: 19,333	Your K5 had focus issues, you tested another one that had the same issues, you tested a K30 that focused perfectly, so the logical thing to do was buy another K5?

07-25-2012, 03:14 PM	#99
JinDesu Veteran Member Join Date: Jun 2011 Location: New York City Photos: Gallery Posts: 5,638	Originally posted by Parallax Your K5 had focus issues, you tested another one that had the same issues, you tested a K30 that focused perfectly, so the logical thing to do was buy another K5? That's how I choose girls, man.

07-25-2012, 03:18 PM	#100
Parallax Senior Moderator Loyal Site Supporter Join Date: Apr 2008 Location: South Dakota Photos: Gallery Posts: 19,333	I can understand using that logic for girlfriends; but not for something important, like a camera.

07-26-2012, 08:23 PM	#102
bossa Veteran Member Join Date: Jul 2011 Location: Adelaide Posts: 4,546	I'm not sure if this has already been posted but I felt it might be of interest: http://www.3dnews.ru/_imgdata/img/2012/07/13/632251/7.jpg

07-26-2012, 09:44 PM	#103
vonBaloney Pentaxian Join Date: Oct 2011 Location: Albuquerque, NM Posts: 6,029	Originally posted by bossa I'm not sure if this has already been posted but I felt it might be of interest: http://www.3dnews.ru/_imgdata/img/2012/07/13/632251/7.jpg What do those categories mean? "5% <75% unacceptable" -- means what?

07-26-2012, 10:02 PM	#104
joe.penn Veteran Member Join Date: Mar 2011 Location: Maryland (Right Outside Washington DC) Photos: Gallery \| Albums Posts: 2,902	The categories mean exactly how they are labled - if 100 shots were taken for the test using each camera tested then 75% (or 75 of the photos) were focused properly, 23% (or 23 photos) completely missed focus, etc etc.. So, the higher the "on focus" rate, the more reliable focus is for that camera (not sure what lens was used but certain SDM lenses would give a drop in that "on focus" percentage for fast moving subjects). In short, the K5 does quite well against the competition.

07-26-2012, 10:35 PM	#105
vonBaloney Pentaxian Join Date: Oct 2011 Location: Albuquerque, NM Posts: 6,029	Originally posted by joe.penn The categories mean exactly how they are labled - if 100 shots were taken for the test using each camera tested then 75% (or 75 of the photos) were focused properly, 23% (or 23 photos) completely missed focus, etc etc.. So, the higher the "on focus" rate, the more reliable focus is for that camera (not sure what lens was used but certain SDM lenses would give a drop in that "on focus" percentage for fast moving subjects). In short, the K5 does quite well against the competition. I'm not getting it -- there are two percentages. K-5 has 23% "75-90% blurry" -- 23% of what? The total images, I assume, as you were saying -- that's fine and obvious -- all adds up to 100%. I just don't know the 3 categories for each camera mean. What is "75-90% blurry"? 75-90% of what? Not of the images, because that's 23%.