Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
09-13-2014, 05:00 PM   #1
New Member




Join Date: Sep 2014
Posts: 9
K-3 AF issue

Hi. I got my K-3 Thursday, opened the box and started playing with it Friday (yesterday). As far as handling goes, it is fantastic. Such a great feeling camera.

Within about 45 minutes I had figured out most things I needed to begin, without the manual mostly, which speaks to the ease of use. I still need a little work, but I feel as though I know it pretty well already, and I like it a lot.

But I'm having trouble with AF. I took several pics today, and they all seem a little out of focus. Some were of my neighbor's dog in my yard, and that is when I noticed that the K-3 seemed to be front-focusing. I had the spot plainly on the dog, and while she was out of focus, the blades of grass in front of her were in focus.

So I did some reading. I read the other recent thread about focus issues. I also did a google search that returned a result saying that many DSLRs have this issue, and there is an adjustment (the article was from a Pentax site. I think this one.).

So I tried what was suggested. When I did that, the K-3 seemed to back focus. I only moved it to -1 (The smallest adjustment possible).

This is very unscientific so far, just using objects in and near my house. I plan to do this: How to Quickly Test Your DSLR for Autofocus Issues but I don't have time right now. Anyone have any thoughts? Is this common? I suppose it could be the camera or the 50mm 1.8 (The only AF Pentax lens I have).

Thanks,
Wayne

09-13-2014, 05:19 PM   #2
Site Supporter
Site Supporter
Ex Finn.'s Avatar

Join Date: Nov 2007
Location: Southern Maryland.
Photos: Gallery | Albums
Posts: 6,374
50mm F1.8, that is a demanding lens for nailing focus, specially if shooting with more open apertures. Post a sample image.
09-13-2014, 05:27 PM - 1 Like   #3
Site Supporter
Site Supporter
jatrax's Avatar

Join Date: May 2010
Location: Oregon
Photos: Gallery | Albums
Posts: 10,825
Please do NOT mess with the AF fine tuning without a tripod, proper high contrast target and a good knowledge of what you are doing. The margin of error with poor testing exceeds the potential real error.

Testing the AF on a dog or your family or any other such target will not often result in a good test, there are too many variables. If you want a quick test to see if you MAY have a problem, shoot a well lit stationary target like a brick wall, while using a tripod. If it is sharp then the camera / lens is not the problem. Remember if you are shooting wide open, the depth of focus @ f/1.8 will be miniscule, as in maybe a few centimeters. So shooting a dog might result in a part of the dog's ear in focus and nothing else.

If the brick test comes up not sharp, then proceed to more extensive testing but this is not something to just snap a pick and change some settings. When I do AF testing I allocate 1 to 2 hours and use a purpose designed target, a tripod, a shutter release and take 5 to 10 samples images at each setting.

You also do not mention which AF mode you are using. So it is entirely possible that the camera decided the blades of grass in front of the dog was what you wanted to be in focus. The camera does not know what you want to focus on. I almost always use single point focus mode so that I know exactly what the camera is going to lock onto.
09-13-2014, 05:49 PM   #4
Pentaxian




Join Date: Jun 2009
Location: GMT +10
Photos: Albums
Posts: 10,893
QuoteOriginally posted by jatrax Quote
Please do NOT mess with the AF fine tuning without a tripod, proper high contrast target and a good knowledge of what you are doing.
Agreed. Proceed with precision and care when doing AF fine tuning.

09-13-2014, 05:59 PM   #5
Veteran Member




Join Date: Sep 2012
Photos: Gallery | Albums
Posts: 1,728
QuoteOriginally posted by jatrax Quote
Please do NOT mess with the AF fine tuning without a tripod, proper high contrast target and a good knowledge of what you are doing. The margin of error with poor testing exceeds the potential real error.

Testing the AF on a dog or your family or any other such target will not often result in a good test, there are too many variables. If you want a quick test to see if you MAY have a problem, shoot a well lit stationary target like a brick wall, while using a tripod. If it is sharp then the camera / lens is not the problem. Remember if you are shooting wide open, the depth of focus @ f/1.8 will be miniscule, as in maybe a few centimeters. So shooting a dog might result in a part of the dog's ear in focus and nothing else.

If the brick test comes up not sharp, then proceed to more extensive testing but this is not something to just snap a pick and change some settings. When I do AF testing I allocate 1 to 2 hours and use a purpose designed target, a tripod, a shutter release and take 5 to 10 samples images at each setting.

You also do not mention which AF mode you are using. So it is entirely possible that the camera decided the blades of grass in front of the dog was what you wanted to be in focus. The camera does not know what you want to focus on. I almost always use single point focus mode so that I know exactly what the camera is going to lock onto.
That was a nice sales pitch for MILCs.
09-13-2014, 06:12 PM   #6
Site Supporter
Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 27,663
QuoteOriginally posted by jatrax Quote
Please do NOT mess with the AF fine tuning without a tripod, proper high contrast target and a good knowledge of what you are doing. The margin of error with poor testing exceeds the potential real error.
What he said.

The process is not easy and may not provide the performance boost you are expecting. If you seriously believe your camera's AF is not working within specification, you may want to consider talking to Ricoh to determine if your camera is defective.


Steve

---------- Post added 09-13-14 at 06:21 PM ----------

QuoteOriginally posted by IchabodCrane Quote
That was a nice sales pitch for MILCs.
I think I missed something...


Steve
09-13-2014, 06:58 PM   #7
Senior Member




Join Date: Mar 2012
Location: Wisconsin
Posts: 272
I just got the K3 as well and there are firmware updates out there. That appeared to fix some AF things I was having.
09-13-2014, 10:31 PM   #8
New Member




Join Date: Sep 2014
Posts: 9
Original Poster
QuoteOriginally posted by theoicarry Quote
I just got the K3 as well and there are firmware updates out there. That appeared to fix some AF things I was having.
Thanks. I updated the firmware before this.

---------- Post added 09-14-14 at 01:34 AM ----------

QuoteOriginally posted by IchabodCrane Quote
That was a nice sales pitch for MILCs.
Agreed about the sales pitch for MILCs. I love my Samsung NX1100 (used it tonight to shoot The Whigs).

I'd pay for the higher end Samsungs if I thought they would really give me something more. Maybe next generation.

---------- Post added 09-14-14 at 01:43 AM ----------

QuoteOriginally posted by jatrax Quote
Please do NOT mess with the AF fine tuning without a tripod, proper high contrast target and a good knowledge of what you are doing. The margin of error with poor testing exceeds the potential real error.

Testing the AF on a dog or your family or any other such target will not often result in a good test, there are too many variables. If you want a quick test to see if you MAY have a problem, shoot a well lit stationary target like a brick wall, while using a tripod. If it is sharp then the camera / lens is not the problem. Remember if you are shooting wide open, the depth of focus @ f/1.8 will be miniscule, as in maybe a few centimeters. So shooting a dog might result in a part of the dog's ear in focus and nothing else.

If the brick test comes up not sharp, then proceed to more extensive testing but this is not something to just snap a pick and change some settings. When I do AF testing I allocate 1 to 2 hours and use a purpose designed target, a tripod, a shutter release and take 5 to 10 samples images at each setting.

You also do not mention which AF mode you are using. So it is entirely possible that the camera decided the blades of grass in front of the dog was what you wanted to be in focus. The camera does not know what you want to focus on. I almost always use single point focus mode so that I know exactly what the camera is going to lock onto.

Thank you for your reply. I've used several AF modes. I started with spot, but they all seem to not focus properly. I may try the brick test, but these shots are clearly out of focus. Repeatedly. With different subjects. I have not yet used a tripod, but I can (and have) take the same pic with my other cameras and they will be in focus. Every time.

I didn't methodically test AF using a test chart and such, but I did take several pictures of different subjects. None of them in focus.

And yes, it was at 1.8. Yes, that depth of field is miniiscule. But it does exist, and it exists in a different location than where the AF point said it was.

09-13-2014, 11:56 PM   #9
Site Supporter
Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 27,663
QuoteOriginally posted by waynesworld Quote
And yes, it was at 1.8. Yes, that depth of field is miniiscule. But it does exist, and it exists in a different location than where the AF point said it was.
...and you are sure about that.

I know that sounds sarcastic and it is a little bit, but one of the issues of AF is just simply determining what it was you intended to be in focus and figuring out good ways to let the camera in on the secret. That is why you test against a stationary flat target at a constant distance. The intent is to remove all variables but the AF calibration. AF speed and tracking ability are not pertinent here, nor is your perception of where the focus point should be at any point in time. Hand-held, even a steady hand shifts several millimeters per second. Dogs and children continue moving after the exposure cycle starts.

You may not be willing to take advice, but I am a K-3 owner (DA 50/1.8 owner too) and a bit of a fiddly sort and went down the lens calibration path on a whim. What you will find is that the sensitivity of the PDAF system is such that it will provide the same focus precision at f/2.8 (best case with the center focus points) as at f/2 or f/1.8 or f/1.4 or f/1.2. Other than that array of center column points, the best case precision is the same as at f/5.6. This is not an urban legend. It is the design specification for your camera.

There is another aspect that is not well documented. The calibration for a particular lens may vary by subject distance and/or (for zoom) focal length. A calibration bias set at 1 meter may not work well at 5 meters. I won't speculate further because I don't have a firm explanation. What I can say is that after considerable time with a controlled setup, I found that I could not dial in per-lens calibrations that were consistently better at all distances than the factory settings.

Having fully warned you, here is my recommendation:
  • Evaluate using a high-contrast target attached to a flat wall. I use Bob Atkins' lens test target, but something as simple as black printed text glued to a piece of hardboard is adequate. What you want is a good number of nice sharp edges.
  • Test at 10-20x the focal length with camera on tripod
  • Confirm focal plane of camera is parallel to target
  • Natural light (AF systems are prone to error with some lighting)
  • AF at center point only
  • Disengage AF and manually move the focus ring to infinity
  • Engage AF system again
  • Use shutter button or AF button to acquire focus
  • Without touching the lens, switch to live view and press OK for magnification. Correct focus may be confirmed by visual inspection or using focus peaking.
  • I suggest doing this at least five times and accepting the best three of five
If AF agrees with live view manual focus, you should stop here. If not, you must determine (by trial and error) how much bias to apply and repeat the sequence above until you get consistent results at 10-20x the focal length. As a final confirmation, check again at 200x the focal length. If your calibration value does not work at distance, you will have to figure out what distance you want to test at.

The camera recognizes four ranges (unknown, macro, close, distant). What these mean varies by lens and is recorded in the exif. If you can't get agreement across ranges, calibrate for the one where most of your shooting happens.

As I mentioned in my initial comment, you may want to gather data indicating the problem and perhaps negotiate for a replacement camera. If the AF is as bad as your accounts indicate, that may be your best route of action and will certainly take less time.


Steve

Last edited by stevebrot; 09-14-2014 at 12:06 AM.
09-14-2014, 07:16 AM   #10
New Member




Join Date: Sep 2014
Posts: 9
Original Poster
QuoteOriginally posted by stevebrot Quote
...and you are sure about that.

I know that sounds sarcastic and it is a little bit, but one of the issues of AF is just simply determining what it was you intended to be in focus and figuring out good ways to let the camera in on the secret. That is why you test against a stationary flat target at a constant distance. The intent is to remove all variables but the AF calibration. AF speed and tracking ability are not pertinent here, nor is your perception of where the focus point should be at any point in time. Hand-held, even a steady hand shifts several millimeters per second. Dogs and children continue moving after the exposure cycle starts.

You may not be willing to take advice, but I am a K-3 owner (DA 50/1.8 owner too) and a bit of a fiddly sort and went down the lens calibration path on a whim. What you will find is that the sensitivity of the PDAF system is such that it will provide the same focus precision at f/2.8 (best case with the center focus points) as at f/2 or f/1.8 or f/1.4 or f/1.2. Other than that array of center column points, the best case precision is the same as at f/5.6. This is not an urban legend. It is the design specification for your camera.

There is another aspect that is not well documented. The calibration for a particular lens may vary by subject distance and/or (for zoom) focal length. A calibration bias set at 1 meter may not work well at 5 meters. I won't speculate further because I don't have a firm explanation. What I can say is that after considerable time with a controlled setup, I found that I could not dial in per-lens calibrations that were consistently better at all distances than the factory settings.

Having fully warned you, here is my recommendation:
  • Evaluate using a high-contrast target attached to a flat wall. I use Bob Atkins' lens test target, but something as simple as black printed text glued to a piece of hardboard is adequate. What you want is a good number of nice sharp edges.
  • Test at 10-20x the focal length with camera on tripod
  • Confirm focal plane of camera is parallel to target
  • Natural light (AF systems are prone to error with some lighting)
  • AF at center point only
  • Disengage AF and manually move the focus ring to infinity
  • Engage AF system again
  • Use shutter button or AF button to acquire focus
  • Without touching the lens, switch to live view and press OK for magnification. Correct focus may be confirmed by visual inspection or using focus peaking.
  • I suggest doing this at least five times and accepting the best three of five
If AF agrees with live view manual focus, you should stop here. If not, you must determine (by trial and error) how much bias to apply and repeat the sequence above until you get consistent results at 10-20x the focal length. As a final confirmation, check again at 200x the focal length. If your calibration value does not work at distance, you will have to figure out what distance you want to test at.

The camera recognizes four ranges (unknown, macro, close, distant). What these mean varies by lens and is recorded in the exif. If you can't get agreement across ranges, calibrate for the one where most of your shooting happens.

As I mentioned in my initial comment, you may want to gather data indicating the problem and perhaps negotiate for a replacement camera. If the AF is as bad as your accounts indicate, that may be your best route of action and will certainly take less time.


Steve
Thanks so much for your detailed reply. I'm not sure what I'll do. I may just return it and pay the return shipping. That's a pretty cheap rent of a camera, though I didn't intend for that to happen.

I just checked, and it looks like I deleted all my test photos. Sorry about that. I know everyone wants confirmation, and I understand that.

I currently have these settings: 15, 17 and 18 set to 1. 16 set to 3. In SEL mode for AF, with a medium-sized red square, though I've tried several AF settings.
09-14-2014, 08:16 AM   #11
Site Supporter
Site Supporter
jatrax's Avatar

Join Date: May 2010
Location: Oregon
Photos: Gallery | Albums
Posts: 10,825
QuoteOriginally posted by waynesworld Quote
Thank you for your reply. I've used several AF modes. I started with spot, but they all seem to not focus properly. I may try the brick test, but these shots are clearly out of focus. Repeatedly. With different subjects. I have not yet used a tripod, but I can (and have) take the same pic with my other cameras and they will be in focus. Every time. I didn't methodically test AF using a test chart and such, but I did take several pictures of different subjects. None of them in focus. And yes, it was at 1.8. Yes, that depth of field is miniiscule. But it does exist, and it exists in a different location than where the AF point said it was.
It is impossible to judge another poster's experience level, so if my reply seemed simplistic, my apologizes. We get nearly every week a new owner who snaps a couple pics, finds the AF fine tuning and starts fooling around with it. 95% of the time it was user error, which gets progressively worse once they start messing around with AF fine tuning. AF tuning is useful and necessary in some cases but not something to start with until all other possibilities are exhausted.

On my k-5 nearly every lens required an adjustment. On my k-3 only one lens does and that is so small (-1) that it probably is not really needed.

A couple of thoughts:
The AF 'points' are not points but areas and are much bigger than you might think. It is quite possible for the red dot to be on what you think should be in focus but the camera actually focused on something else.
It is possible you have a problem, though it is most likely a lens issue not a camera issue. However, experience tells us that in most cases the issue is unfamiliarity with a new camera or inexperience with DSLRs in general.
Run the tests as Steve noted above, it is a good process and should tell you if there is a problem. If there is then you can use the AF fine tuning to bring things into correct adjustment

If you cannot bring the camera / lens into correct focus using AF tuning then the best route is send it back as defective and ask for a replacement.
09-14-2014, 08:23 AM   #12
Site Supporter
Site Supporter




Join Date: Sep 2006
Location: WV
Posts: 1,495
QuoteOriginally posted by waynesworld Quote
And yes, it was at 1.8. Yes, that depth of field is miniiscule. But it does exist, and it exists in a different location than where the AF point said it was.
Keep in mind the camera does not know what you want to focus on. It may select a point within the focus plane that is different from the point you want in focus. That is not a defect, and it's one of the reasons many Pentax lenses have a quick-shift focus feature.

Nevertheless, due to manufacturing tolerances, it is certainly possible that a lens may front focus or back focus. Read this article: LensRentals.com - "This lens is soft" and other myths

I find this method of testing for and correcting focus issues works best: AF microadjustment for the 1Ds mark III, 1D Mk3, 5D Mk2, 7D, 1D X
09-14-2014, 09:07 AM   #13
New Member




Join Date: Sep 2014
Posts: 9
Original Poster
QuoteOriginally posted by jatrax Quote
It is impossible to judge another poster's experience level, so if my reply seemed simplistic, my apologizes. We get nearly every week a new owner who snaps a couple pics, finds the AF fine tuning and starts fooling around with it. 95% of the time it was user error, which gets progressively worse once they start messing around with AF fine tuning. AF tuning is useful and necessary in some cases but not something to start with until all other possibilities are exhausted.

On my k-5 nearly every lens required an adjustment. On my k-3 only one lens does and that is so small (-1) that it probably is not really needed.

A couple of thoughts:
The AF 'points' are not points but areas and are much bigger than you might think. It is quite possible for the red dot to be on what you think should be in focus but the camera actually focused on something else.
It is possible you have a problem, though it is most likely a lens issue not a camera issue. However, experience tells us that in most cases the issue is unfamiliarity with a new camera or inexperience with DSLRs in general.
Run the tests as Steve noted above, it is a good process and should tell you if there is a problem. If there is then you can use the AF fine tuning to bring things into correct adjustment

If you cannot bring the camera / lens into correct focus using AF tuning then the best route is send it back as defective and ask for a replacement.
No worries. Thanks so much for your assistance.

If the camera won't focus on the things I was shooting (large stationary dog, tree) then I don't know what good the AF is. You're right, it could be focusing on something else, but if it is, at least in most of the cases in which I was testing, then it is far off.

---------- Post added 09-14-14 at 12:10 PM ----------

QuoteOriginally posted by MPrince Quote
Keep in mind the camera does not know what you want to focus on. It may select a point within the focus plane that is different from the point you want in focus. That is not a defect, and it's one of the reasons many Pentax lenses have a quick-shift focus feature.

Nevertheless, due to manufacturing tolerances, it is certainly possible that a lens may front focus or back focus. Read this article: LensRentals.com - "This lens is soft" and other myths

I find this method of testing for and correcting focus issues works best: AF microadjustment for the 1Ds mark III, 1D Mk3, 5D Mk2, 7D, 1D X
Thank you. I may do this.

I didn't expect point and shoot, and I was willing to do some work to get things right, but I'm rather frustrated at the moment, and I'm not yet sure I'll continue down this path.
09-14-2014, 10:02 AM   #14
Site Supporter
Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 27,663
QuoteOriginally posted by waynesworld Quote
Thanks so much for your detailed reply. I'm not sure what I'll do. I may just return it and pay the return shipping. That's a pretty cheap rent of a camera, though I didn't intend for that to happen.
Contact Ricoh. They may be willing to calibrate your camera to your lenses as part of the warranty.

I agree that there is a good chance that your AF is badly off or defective. My experience with the K-3 is that I have been fairly amazed at how well the AF works. I am a manual focus shooter at heart, but for some subjects just simply switching all the AF points on in AF-C and letting the camera "drive" has resulted in results that are more than I would have asked for or expected.


Steve

---------- Post added 09-14-14 at 10:06 AM ----------

QuoteOriginally posted by MPrince Quote
I find this method of testing for and correcting focus issues works best: AF microadjustment for the 1Ds mark III, 1D Mk3, 5D Mk2, 7D, 1D X
I too have found this to be a good method and equivalent to magnified focus peaking on the K-3.


Steve

Last edited by stevebrot; 09-14-2014 at 10:11 AM.
09-14-2014, 05:40 PM   #15
Inactive Account




Join Date: Aug 2014
Posts: 76
QuoteOriginally posted by waynesworld Quote
Thanks so much for your detailed reply. I'm not sure what I'll do. I may just return it and pay the return shipping. That's a pretty cheap rent of a camera, though I didn't intend for that to happen.

I just checked, and it looks like I deleted all my test photos. Sorry about that. I know everyone wants confirmation, and I understand that.

I currently have these settings: 15, 17 and 18 set to 1. 16 set to 3. In SEL mode for AF, with a medium-sized red square, though I've tried several AF settings.
Check the return policy of the store you purchased it from.
If you bought it at B&H, they have a ten day return/exchange policy.
They will email you return postage label for UPS ground shipping at their expense.

I believe return policy states:
You must NOT have filled out warranty cards.
You must NOT have removed the UPC from the box.
All original contents of box must be returned. (double check that you put all accessories, manuals etc back in the box)
You must return an exchange authorization form (that they will EMAIL to you) with your camera.

Package MUST be scanned by UPS within 10 days of receipt of your original camera shipment from them.

If your tests show that your camera is indeed faulty, my suggestion is you call your retailer, ask what return policy is before you wait too long.

Also, if it is truly faulty (after above mentioned tests) please post the serial number on this thread, it might be helpful in future for others if they should experience similar issues;

I think if you have filled out warranty cards, it must be sent in for repair to mfg.

So everyone who purchases a new camera, should put it through ALL of the modes and functions and take a LOT of pictures upon receipt. Sometimes cameras behave well, until switched to a problematic mode or function. Best to find out about potential issues before your exchange/return period is up!

Last edited by Pentaxaholic; 09-14-2014 at 05:44 PM. Reason: error
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!
50mm, adjustment, af, autofocus, camera, dog, dslr, error, firmware, firmware updates, focus, issue, issues, k-3, k3, lens, milcs, pentax, pentax k-3, pitch, sales, steve, test, tripod
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
K-3 Movie-SR issue tava Pentax K-3 45 03-18-2016 07:03 AM
K-3 AF issue PixelGeek Troubleshooting and Beginner Help 80 04-14-2014 10:00 PM
K-3 USB connectivity issue peterjcb Pentax K-3 8 03-20-2014 08:12 PM
K-3 flash issue konraDarnok Pentax K-3 68 03-13-2014 06:07 AM
How good is the K-3's AF-C tracking AF? Jeff Charles Pentax K-3 6 12-09-2013 03:41 PM



All times are GMT -7. The time now is 05:40 AM. | See also: NikonForums.com, CanonForums.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