Softer in LiveView?

mee · 12-12-2015, 05:48 PM

I suppose the question now is, (outside of Does the top image look sharp for f/2.8 with this lens?) Why is this particular lens erratic in focusing when using contrast detection AF in liveview when the other lenses are not erratic?

mee · 12-13-2015, 04:01 PM

No one?

**UncleVanya** · 12-13-2015, 07:57 PM

Originally posted by mee

No one?

My suggestion was to compare the lens files. My guess is that they would show differences. BUT I don't know how CDAF could be worse and less reliable than PDAF. I'm at a loss.

mee · 12-13-2015, 08:21 PM

Originally posted by UncleVanya

My suggestion was to compare the lens files. My guess is that they would show differences. BUT I don't know how CDAF could be worse and less reliable than PDAF. I'm at a loss.

I suppose by seeing the lack of responses after posting more info, others are too ?

don't really want to play with debug menu atm .. really hoping Precision can do something further.. I mean what is the next step? Replacing element groups?

It behaves in the same manner on 2 different Pentax bodies so (unless both bodies have the same problem) it seems more like there is an issue still with the lens..

stevebrot · 12-14-2015, 09:15 PM

Originally posted by UncleVanya

BUT I don't know how CDAF could be worse and less reliable than PDAF. I'm at a loss.

My thoughts as well. There is something strange about these photos. The top one has somewhat better sharpness despite the plane of focus not being apparent, but the image overall does not reflect the expected quality for this class of lens. For the bottom image, the focus point is on the chair at lower right and again the image quality is sub-par even within the plane of focus.

Exif would be helpful (straight from camera JPEG) as would an attempt at manual focus. What is the history of this lens before it was sent to Precision?

Steve

mee · 12-15-2015, 06:59 AM

Originally posted by stevebrot

My thoughts as well. There is something strange about these photos. The top one has somewhat better sharpness despite the plane of focus not being apparent, but the image overall does not reflect the expected quality for this class of lens. For the bottom image, the focus point is on the chair at lower right and again the image quality is sub-par even within the plane of focus.

Both are aimed at the brick pylon in the center of the photograph using single focus point straight on.

From the tests, it appears the lens just occasionally goes beserk in CDAF mode and doesn't know what is in focus. In this situation, it seems to have happened to focus closer up and the chair is within it's plane of focus.

stevebrot · 12-15-2015, 02:06 PM

Originally posted by mee

Both are aimed at the brick pylon in the center of the photograph using single focus point straight on.

From the tests, it appears the lens just occasionally goes beserk in CDAF mode and doesn't know what is in focus. In this situation, it seems to have happened to focus closer up and the chair is within it's plane of focus.

Neither photo is representative of what I would expect of the DA* 16-50/2.8. The best areas of both look as if there is a vaseline-coated filter or something similar in the optical path. That aside, CDAF searches for the contrast peak at the intended focus point. If the target is low contrast or lacks detail, focus should rack (lock-to-lock) at least once before giving up and should not lock unless contrast has peaked above a preset threshold.

If the CDAF system attains focus and the subsequent image is out of focus (e.g. bottom example), I can only think of two possibilities:

CDAF processing in the body is faulty
Shift of optical focus AFTER focus has been attained

The position of the distance scale on the focus ring is one way to differentiate the two. If the distance scale is consistent with the actual distance, the CDAF is working properly, at least for gross focus. BTW...be sure your focus target is unambiguous (flat and parallel to the sensor) and sufficiently large to extend well beyond the boundaries of the camera's focus zone. It also helps to use moderate distance. A subject at 95 feet with a 50mm lens is essentially the same as infinity (FWIW, DOF is 181' at 90' and f/2.8 on APS-C. Near focus = 55.6'). A distance less than 30 feet might be more appropriate.

If the CDAF system is not at fault (focus attained, but intended subject is OOF), there are multiple possibilities:

Inadvertent movement of quick shift ring
Internal mechanical failure resulting in a shift in the focus mechanism
Damage to the optical system that allows elements to shift position independent of the focus mechanism
Movement of the focus motor outside the CDAF control loop
Something else...

Of the above, only #1 has much likelihood and typically would not be repeatable. #2 and #3 are possible if the lens has been physically damaged (dropped) and/or there has been a botched service attempt. #4 is highly unlikely, though not impossible.

Steve

Last edited by stevebrot; 12-15-2015 at 02:36 PM.

mee · 12-15-2015, 04:11 PM

Originally posted by stevebrot

Neither photo is representative of what I would expect of the DA* 16-50/2.8. The best areas of both look as if there is a vaseline-coated filter or something similar in the optical path. That aside, CDAF searches for the contrast peak at the intended focus point. If the target is low contrast or lacks detail, focus should rack (lock-to-lock) at least once before giving up and should not lock unless contrast has peaked above a preset threshold.

If the CDAF system attains focus and the subsequent image is out of focus (e.g. bottom example), I can only think of two possibilities:

CDAF processing in the body is faulty

Two different camera bodies? Possible but highly unlikely.

Originally posted by stevebrot

Shift of optical focus AFTER focus has been attained

Not sure what this means.

Originally posted by stevebrot

The position of the distance scale on the focus ring is one way to differentiate the two. If the distance scale is consistent with the actual distance, the CDAF is working properly, at least for gross focus. BTW...be sure your focus target is unambiguous (flat and parallel to the sensor) and sufficiently large to extend well beyond the boundaries of the camera's focus zone. It also helps to use moderate distance. A subject at 95 feet with a 50mm lens is essentially the same as infinity (FWIW, DOF is 181' at 90' and f/2.8 on APS-C. Near focus = 55.6'). A distance less than 30 feet might be more appropriate.

I'm not sure you've kept up with the saga but the issue is present at infinity.. hence why I'm shooting so far away. I'm deliberately choosing a distance where the lens has issues.

Originally posted by stevebrot

If the CDAF system is not at fault (focus attained, but intended subject is OOF), there are multiple possibilities:

Inadvertent movement of quick shift ring

Nope.

Originally posted by stevebrot
Internal mechanical failure resulting in a shift in the focus mechanism

stevebrot · 12-17-2015, 04:32 PM

Originally posted by mee

until I set LiveView to Contrast detection.

CDAF is the only AF system for live view on Pentax cameras.

As for your "saga"...I have followed your experiences with this lens, including your original post regarding problems with softness at infinity.

https://www.pentaxforums.com/forums/10-pentax-slr-lens-discussion/306846-16-5...-infinity.html

I also recall your original example photos that you removed from that thread. According to that thread, you disassembled the lens in an unsuccessful attempt to service the focus system and were in discussion with Precision regarding possible repair (4 Nov 15). A couple of weeks later you sent the lens to Precision and repair was delayed for several reasons including a wait for parts. You got the repaired lens back on 9 Dec (replaced motor and controller for $214...good deal) and discovered on receipt that the SDM now squeaks and results were softer than your DAL 18-55 kit lens with the same camera and subject (not such a good deal). This thread was started that same day.

At this point, it appears your intended remedy (posted on another thread) is another repair attempt with Precision under the terms of their warranty. Good luck

Steve

stevebrot · 12-17-2015, 04:33 PM

Originally posted by mee

Gremlins?

Sounds good to me. Something may be slipping out of place after focus locks or the focus motor is receiving spurious commands after focus lock. Both of which would be a lens problem.

Steve

**luftfluss** · 12-17-2015, 04:36 PM - **1 Like**

You Tube

mee · 12-17-2015, 05:04 PM

Originally posted by stevebrot

CDAF is the only AF system for live view on Pentax cameras.

Is this true? On my K-5 II [page 147 of the manual] and the K-r , I see Smile Detection ( form of Contrast AF), Contrast AF, and Phase Difference AF in the Live View settings.

Originally posted by stevebrot

As for your "saga"...I have followed your experiences with this lens, including your original post regarding problems with softness at infinity.

https://www.pentaxforums.com/forums/10-pentax-slr-lens-discussion/306846-16-5...-infinity.html

I also recall your original example photos that you removed from that thread. According to that thread, you disassembled the lens in an unsuccessful attempt to service the focus system and were in discussion with Precision regarding possible repair (4 Nov 15). A couple of weeks later you sent the lens to Precision and repair was delayed for several reasons including a wait for parts. You got the repaired lens back on 9 Dec (replaced motor and controller for $214...good deal) and discovered on receipt that the SDM now squeaks and results were softer than your DAL 18-55 kit lens with the same camera and subject (not such a good deal). This thread was started that same day.

At this point, it appears your intended remedy (posted on another thread) is another repair attempt with Precision under the terms of their warranty. Good luck

That is it in a nutshell.

I will say when it came back initially it is a bit sharper than it was so they did something to the focusing system but no where to where it should be. In any case, Precision sent me a shipping label and their lead Pentax tech will be looking at the lens on round 2. It is in UPS' hands now on it's way to Precision and I'm hopeful for a desired outcome this time!

stevebrot · 12-17-2015, 06:39 PM

Originally posted by mee

and Phase Difference AF in the Live View settings.

Originally posted by mee

Is this true? On my K-5 II [page 147 of the manual] and the K-r , I see Smile Detection ( form of Contrast AF), Contrast AF, and Phase Difference AF in the Live View settings.

Well, that is interesting

I learn something new every day. From what I can tell, this option exists for the the K-7, K-x, K-5/K-5II, and K-r, but not for the K-30/K-50 or any current model Pentax dSLR.

For the benefit of those not having the manual, the text for the Phase Difference AF option for the K-5II reads as follows:

"Cancels Live View and performs autofocus with the AF sensor.
If the shutter release button is pressed halfway during Live View, the Live View image will disappear and the autofocus system operates. Once Focused, the Live View image will be displayed again."

In short, live view switches off, the mirror flips down and focus is done using the PDAF sensor with return to live view to finish the cycle.

"Fascinating..."

Steve

**UncleVanya** · 12-18-2015, 07:23 AM

Originally posted by stevebrot

Well, that is interesting

I learn something new every day. From what I can tell, this option exists for the the K-7, K-x, K-5/K-5II, and K-r, but not for the K-30/K-50 or any current model Pentax dSLR.

For the benefit of those not having the manual, the text for the Phase Difference AF option for the K-5II reads as follows:

"Cancels Live View and performs autofocus with the AF sensor.
If the shutter release button is pressed halfway during Live View, the Live View image will disappear and the autofocus system operates. Once Focused, the Live View image will be displayed again."

In short, live view switches off, the mirror flips down and focus is done using the PDAF sensor with return to live view to finish the cycle.

"Fascinating..."

Steve

Wild. Since the pdaf normal way of using the lens optical viewfinder and all seems fine.

12-14-2015, 09:15 PM	#35
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by UncleVanya BUT I don't know how CDAF could be worse and less reliable than PDAF. I'm at a loss. My thoughts as well. There is something strange about these photos. The top one has somewhat better sharpness despite the plane of focus not being apparent, but the image overall does not reflect the expected quality for this class of lens. For the bottom image, the focus point is on the chair at lower right and again the image quality is sub-par even within the plane of focus. Exif would be helpful (straight from camera JPEG) as would an attempt at manual focus. What is the history of this lens before it was sent to Precision? Steve Last edited by stevebrot; 12-14-2015 at 09:33 PM.

12-15-2015, 02:06 PM	#37
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by mee Both are aimed at the brick pylon in the center of the photograph using single focus point straight on. From the tests, it appears the lens just occasionally goes beserk in CDAF mode and doesn't know what is in focus. In this situation, it seems to have happened to focus closer up and the chair is within it's plane of focus. Neither photo is representative of what I would expect of the DA* 16-50/2.8. The best areas of both look as if there is a vaseline-coated filter or something similar in the optical path. That aside, CDAF searches for the contrast peak at the intended focus point. If the target is low contrast or lacks detail, focus should rack (lock-to-lock) at least once before giving up and should not lock unless contrast has peaked above a preset threshold. If the CDAF system attains focus and the subsequent image is out of focus (e.g. bottom example), I can only think of two possibilities: CDAF processing in the body is faulty Shift of optical focus AFTER focus has been attained The position of the distance scale on the focus ring is one way to differentiate the two. If the distance scale is consistent with the actual distance, the CDAF is working properly, at least for gross focus. BTW...be sure your focus target is unambiguous (flat and parallel to the sensor) and sufficiently large to extend well beyond the boundaries of the camera's focus zone. It also helps to use moderate distance. A subject at 95 feet with a 50mm lens is essentially the same as infinity (FWIW, DOF is 181' at 90' and f/2.8 on APS-C. Near focus = 55.6'). A distance less than 30 feet might be more appropriate. If the CDAF system is not at fault (focus attained, but intended subject is OOF), there are multiple possibilities: Inadvertent movement of quick shift ring Internal mechanical failure resulting in a shift in the focus mechanism Damage to the optical system that allows elements to shift position independent of the focus mechanism Movement of the focus motor outside the CDAF control loop Something else... Of the above, only #1 has much likelihood and typically would not be repeatable. #2 and #3 are possible if the lens has been physically damaged (dropped) and/or there has been a botched service attempt. #4 is highly unlikely, though not impossible. Steve Last edited by stevebrot; 12-15-2015 at 02:36 PM.

12-15-2015, 04:11 PM	#38
mee Veteran Member Join Date: May 2010 Posts: 7,403 Original Poster	Originally posted by stevebrot Neither photo is representative of what I would expect of the DA* 16-50/2.8. The best areas of both look as if there is a vaseline-coated filter or something similar in the optical path. That aside, CDAF searches for the contrast peak at the intended focus point. If the target is low contrast or lacks detail, focus should rack (lock-to-lock) at least once before giving up and should not lock unless contrast has peaked above a preset threshold. If the CDAF system attains focus and the subsequent image is out of focus (e.g. bottom example), I can only think of two possibilities: CDAF processing in the body is faulty Two different camera bodies? Possible but highly unlikely. Originally posted by stevebrot Shift of optical focus AFTER focus has been attained Not sure what this means. Originally posted by stevebrot The position of the distance scale on the focus ring is one way to differentiate the two. If the distance scale is consistent with the actual distance, the CDAF is working properly, at least for gross focus. BTW...be sure your focus target is unambiguous (flat and parallel to the sensor) and sufficiently large to extend well beyond the boundaries of the camera's focus zone. It also helps to use moderate distance. A subject at 95 feet with a 50mm lens is essentially the same as infinity (FWIW, DOF is 181' at 90' and f/2.8 on APS-C. Near focus = 55.6'). A distance less than 30 feet might be more appropriate. I'm not sure you've kept up with the saga but the issue is present at infinity.. hence why I'm shooting so far away. I'm deliberately choosing a distance where the lens has issues. Originally posted by stevebrot If the CDAF system is not at fault (focus attained, but intended subject is OOF), there are multiple possibilities: Inadvertent movement of quick shift ring Nope. Originally posted by stevebrot Internal mechanical failure resulting in a shift in the focus mechanism Possible. Originally posted by stevebrot Damage to the optical system that allows elements to shift position independent of the focus mechanism Possible. Originally posted by stevebrot Movement of the focus motor outside the CDAF control loop Possible but unlikely. It seems to 'lock' and its done - the distance scale doesn't move (I was watching it the entire time) when it both properly focuses and misfocuses. Originally posted by stevebrot Something else... Gremlins? Originally posted by stevebrot Of the above, only #1 has much likelihood and typically would not be repeatable. #2 and #3 are possible if the lens has been physically damaged (dropped) and/or there has been a botched service attempt. #4 is highly unlikely, though not impossible. Yet 1 is the least likely. Because, like I've said over and over and over now, the other lenses I tested in the same location, with the same tripod placement, at the same focal length (and with the 50-135mm the same aperture) did not do this. The K-r did not do this in Liveview initially, like it misbehaved on the K-5II, until I set LiveView to Contrast detection. Interesting thoughts though. Thank you for sharing and I will be interested in knowing what it is eventually... but none of those possibles are fixable by me. I just got off of work and I see Precision finally contacted me (by phone?) to get a shipping label to send it back. ---------- Post added 12-15-15 at 05:36 PM ---------- Oh and thank you, Steve, for confirming that doesn't look right -- even the 'in focus' one for that lens at f/2.8 -- mucho appreciated.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
LiveView and Video modes noice in K-3	Adrián Hesto	Troubleshooting and Beginner Help	3	10-09-2014 05:14 PM
What causes the whirring noise in LiveView?	Laurentiu Cristofor	Pentax DSLR Discussion	48	04-08-2012 09:52 PM
K20D and DA*16-50mm in Liveview	Woolly	Troubleshooting and Beginner Help	2	01-08-2012 06:18 AM
K-x... Slow to focus in Liveview	DaveHolmes	Pentax DSLR Discussion	9	06-07-2011 08:05 AM
Pentax K-x Audible Noise in LiveView	omardalian	Pentax DSLR Discussion	14	06-03-2010 11:45 AM

12-12-2015, 05:48 PM	#31
mee Veteran Member Join Date: May 2010 Posts: 7,403 Original Poster	I suppose the question now is, (outside of Does the top image look sharp for f/2.8 with this lens?) Why is this particular lens erratic in focusing when using contrast detection AF in liveview when the other lenses are not erratic?

12-13-2015, 04:01 PM	#32
mee Veteran Member Join Date: May 2010 Posts: 7,403 Original Poster	No one?

12-13-2015, 07:57 PM	#33
UncleVanya Loyal Site Supporter Join Date: Jul 2014 Photos: Gallery \| Albums Posts: 28,401	Originally posted by mee No one? My suggestion was to compare the lens files. My guess is that they would show differences. BUT I don't know how CDAF could be worse and less reliable than PDAF. I'm at a loss.

12-13-2015, 08:21 PM	#34
mee Veteran Member Join Date: May 2010 Posts: 7,403 Original Poster	Originally posted by UncleVanya My suggestion was to compare the lens files. My guess is that they would show differences. BUT I don't know how CDAF could be worse and less reliable than PDAF. I'm at a loss. I suppose by seeing the lack of responses after posting more info, others are too ? don't really want to play with debug menu atm .. really hoping Precision can do something further.. I mean what is the next step? Replacing element groups? It behaves in the same manner on 2 different Pentax bodies so (unless both bodies have the same problem) it seems more like there is an issue still with the lens..

12-15-2015, 06:59 AM	#36
mee Veteran Member Join Date: May 2010 Posts: 7,403 Original Poster	Originally posted by stevebrot My thoughts as well. There is something strange about these photos. The top one has somewhat better sharpness despite the plane of focus not being apparent, but the image overall does not reflect the expected quality for this class of lens. For the bottom image, the focus point is on the chair at lower right and again the image quality is sub-par even within the plane of focus. Both are aimed at the brick pylon in the center of the photograph using single focus point straight on. From the tests, it appears the lens just occasionally goes beserk in CDAF mode and doesn't know what is in focus. In this situation, it seems to have happened to focus closer up and the chair is within it's plane of focus.

12-17-2015, 04:32 PM	#39
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by mee until I set LiveView to Contrast detection. CDAF is the only AF system for live view on Pentax cameras. As for your "saga"...I have followed your experiences with this lens, including your original post regarding problems with softness at infinity. https://www.pentaxforums.com/forums/10-pentax-slr-lens-discussion/306846-16-5...-infinity.html I also recall your original example photos that you removed from that thread. According to that thread, you disassembled the lens in an unsuccessful attempt to service the focus system and were in discussion with Precision regarding possible repair (4 Nov 15). A couple of weeks later you sent the lens to Precision and repair was delayed for several reasons including a wait for parts. You got the repaired lens back on 9 Dec (replaced motor and controller for $214...good deal) and discovered on receipt that the SDM now squeaks and results were softer than your DAL 18-55 kit lens with the same camera and subject (not such a good deal). This thread was started that same day. At this point, it appears your intended remedy (posted on another thread) is another repair attempt with Precision under the terms of their warranty. Good luck Steve

12-17-2015, 04:33 PM	#40
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by mee Gremlins? Sounds good to me. Something may be slipping out of place after focus locks or the focus motor is receiving spurious commands after focus lock. Both of which would be a lens problem. Steve

12-17-2015, 05:04 PM	#42
mee Veteran Member Join Date: May 2010 Posts: 7,403 Original Poster	Originally posted by stevebrot CDAF is the only AF system for live view on Pentax cameras. Is this true? On my K-5 II [page 147 of the manual] and the K-r , I see Smile Detection ( form of Contrast AF), Contrast AF, and Phase Difference AF in the Live View settings. Originally posted by stevebrot As for your "saga"...I have followed your experiences with this lens, including your original post regarding problems with softness at infinity. https://www.pentaxforums.com/forums/10-pentax-slr-lens-discussion/306846-16-5...-infinity.html I also recall your original example photos that you removed from that thread. According to that thread, you disassembled the lens in an unsuccessful attempt to service the focus system and were in discussion with Precision regarding possible repair (4 Nov 15). A couple of weeks later you sent the lens to Precision and repair was delayed for several reasons including a wait for parts. You got the repaired lens back on 9 Dec (replaced motor and controller for $214...good deal) and discovered on receipt that the SDM now squeaks and results were softer than your DAL 18-55 kit lens with the same camera and subject (not such a good deal). This thread was started that same day. At this point, it appears your intended remedy (posted on another thread) is another repair attempt with Precision under the terms of their warranty. Good luck That is it in a nutshell. I will say when it came back initially it is a bit sharper than it was so they did something to the focusing system but no where to where it should be. In any case, Precision sent me a shipping label and their lead Pentax tech will be looking at the lens on round 2. It is in UPS' hands now on it's way to Precision and I'm hopeful for a desired outcome this time!

12-17-2015, 06:39 PM	#43
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by mee and Phase Difference AF in the Live View settings. Originally posted by mee Is this true? On my K-5 II [page 147 of the manual] and the K-r , I see Smile Detection ( form of Contrast AF), Contrast AF, and Phase Difference AF in the Live View settings. Well, that is interesting I learn something new every day. From what I can tell, this option exists for the the K-7, K-x, K-5/K-5II, and K-r, but not for the K-30/K-50 or any current model Pentax dSLR. For the benefit of those not having the manual, the text for the Phase Difference AF option for the K-5II reads as follows: "Cancels Live View and performs autofocus with the AF sensor. If the shutter release button is pressed halfway during Live View, the Live View image will disappear and the autofocus system operates. Once Focused, the Live View image will be displayed again." In short, live view switches off, the mirror flips down and focus is done using the PDAF sensor with return to live view to finish the cycle. "Fascinating..." Steve

12-18-2015, 07:23 AM	#44
UncleVanya Loyal Site Supporter Join Date: Jul 2014 Photos: Gallery \| Albums Posts: 28,401	Originally posted by stevebrot Well, that is interesting I learn something new every day. From what I can tell, this option exists for the the K-7, K-x, K-5/K-5II, and K-r, but not for the K-30/K-50 or any current model Pentax dSLR. For the benefit of those not having the manual, the text for the Phase Difference AF option for the K-5II reads as follows: "Cancels Live View and performs autofocus with the AF sensor. If the shutter release button is pressed halfway during Live View, the Live View image will disappear and the autofocus system operates. Once Focused, the Live View image will be displayed again." In short, live view switches off, the mirror flips down and focus is done using the PDAF sensor with return to live view to finish the cycle. "Fascinating..." Steve Wild. Since the pdaf normal way of using the lens optical viewfinder and all seems fine.