AC130: Video footage of the Auto Focus in action
- Thread starter ELN614
- Start date
LanceOregon
Well-known member
I believe that the consensus is that the older standard definition cameras never had these problems.
Two issues that the Autofocus seems to have problems with are low contrast situations ( especially if the scene does not have good lighting ), and scenes that have rapid movement in them.
Why it took so long to focus on that pole in your video, though, I can't explain.
If I thought that the 130a hardware upgrade would help out with any of these situations, I would buy it. However, I've yet to see anything posted by Panasonic indicating that would indeed be the case.
Two issues that the Autofocus seems to have problems with are low contrast situations ( especially if the scene does not have good lighting ), and scenes that have rapid movement in them.
Why it took so long to focus on that pole in your video, though, I can't explain.
If I thought that the 130a hardware upgrade would help out with any of these situations, I would buy it. However, I've yet to see anything posted by Panasonic indicating that would indeed be the case.
Last edited:
ChrisHarding
Well-known member
Hi Lance
It does seem strange that Panasonic have kept deathly quiet about autofocus yet quite vocal on other issues. I can vouch for the fast/lots of movement but only the subjects not the camera!! Move with the camera over fairly slow moving or stationary subjects and the AF is lightning fast and accurate too!! Stand in one place and it acts up... I actually had an instance at a wedding on a dance floor where I had the focus set to "feet" and was 15' away from a bunch of gyrating dancers and the distance display was going absolutely wild varying from 3 ' to infinity despite the fact that the dancers were roughly in the same "zone" and I was standing still.... the moment I started either panning or walking around the dance floor and introduced so camera motion the problem simply vanished. I would have expected it to be the other way around!!!
If ELN614 had panned the camera across the pole I reckon that AF would have handled it perfectly.... might be worth a try to see if the AF prefers a moving camera???
Chris
It does seem strange that Panasonic have kept deathly quiet about autofocus yet quite vocal on other issues. I can vouch for the fast/lots of movement but only the subjects not the camera!! Move with the camera over fairly slow moving or stationary subjects and the AF is lightning fast and accurate too!! Stand in one place and it acts up... I actually had an instance at a wedding on a dance floor where I had the focus set to "feet" and was 15' away from a bunch of gyrating dancers and the distance display was going absolutely wild varying from 3 ' to infinity despite the fact that the dancers were roughly in the same "zone" and I was standing still.... the moment I started either panning or walking around the dance floor and introduced so camera motion the problem simply vanished. I would have expected it to be the other way around!!!
If ELN614 had panned the camera across the pole I reckon that AF would have handled it perfectly.... might be worth a try to see if the AF prefers a moving camera???
Chris
kwkeirstead
Well-known member
I found with some testing of the 160A that when going from close to far (30 feet to 1.5 miles) I could either get slow lockin or fast lockin on demand.
All I had to do was pan slowly enough and have an intermediate distance object the camera could see as we went from close to far. The other factor was to ease into the far object via a high contrast area at the distant object. C
Coming back to the near object, I never saw slow lock in.
In the horse/tree video here, it does not look like the horse was that far from the tree (would be good to know what the zoom was) and secondly, the tree is not the distant object, the tree was the close object, so things are pretty much the reverse of my testing.
Now, the problem with my observations is I went back to the exact same site on a different day and got fast lockin no matter what I tried. So my testing was inconclusive.
But I did another test at another site with one object at 30 feet and my "far" object at 80 feet and the resutls seemed conclusive (no problem going close to far or coming back).
It's easy in non-scientific testing setups to convince yourself that the behavior is explained by this or that - maybe it's true but maybe it's not.
A proper way (I suspect) to test static focus when transitioning from close to far and return would be to place high contrast cards on stands organized in a V, say 10 x 2 cards over perhaps at a starting depth of 1,000 feet and do consecutive recordings as you gradually increase the distance between the cards going back from the camera.
Even this is likely to be problematic as the experiment design is based on lack of understanding of how the focus works as per camera optical design specifications.
I could see it taking 2 days to do these measurements. You would then have to repeat at least once to prove consistency.
All I had to do was pan slowly enough and have an intermediate distance object the camera could see as we went from close to far. The other factor was to ease into the far object via a high contrast area at the distant object. C
Coming back to the near object, I never saw slow lock in.
In the horse/tree video here, it does not look like the horse was that far from the tree (would be good to know what the zoom was) and secondly, the tree is not the distant object, the tree was the close object, so things are pretty much the reverse of my testing.
Now, the problem with my observations is I went back to the exact same site on a different day and got fast lockin no matter what I tried. So my testing was inconclusive.
But I did another test at another site with one object at 30 feet and my "far" object at 80 feet and the resutls seemed conclusive (no problem going close to far or coming back).
It's easy in non-scientific testing setups to convince yourself that the behavior is explained by this or that - maybe it's true but maybe it's not.
A proper way (I suspect) to test static focus when transitioning from close to far and return would be to place high contrast cards on stands organized in a V, say 10 x 2 cards over perhaps at a starting depth of 1,000 feet and do consecutive recordings as you gradually increase the distance between the cards going back from the camera.
Even this is likely to be problematic as the experiment design is based on lack of understanding of how the focus works as per camera optical design specifications.
I could see it taking 2 days to do these measurements. You would then have to repeat at least once to prove consistency.
Last edited:
ChrisHarding
Well-known member
Hi Walter
My main issue was items fairly close together (like the horse and pole) I do a stedicam routine where I place the couple next to a tree and that used to give me issues on the 130 as the camera refused to focus once I moved away from the couple and "swooped" around the tree and then as I came back to the couple the AF failed....also where weddings were in a gazebo the camera would shift from the couple to the back pillars of the gazebo for no reason but that was easy to solve with manual focus but on a stedicam you have to have auto so I ended up with using manual and shooting further away so the DOF covered the scene. If you want to replicate the troublesome area then you need the camera at say 15' from the subject and a second subject/object at around 20 -25' to "fool"
the AF ...it handles large distances perfectly .... I also found that doing Church ceremonys cutaway of the guests and zooming tighly on say, a group of 3 in the pews, the AF would often lock on to the people behind my group...it might just be a depth of field issue too??? With the iris quite wide and the camera zoomed a bit your effective DOF is small ..... It would be interesting to know what zoom was used with the horses as I seldom had any AF issues at full wide cos even if the focus was out both subjects were still within the DOF range. This might be what you are seeing??? At 15' and at F2 your DOF at 28mm is at least 4' whilst at 55mm it drops to just 12" so at full wide you can be in visual focus between two objects regardless of what the camera has locked onto but zoom to 55m and one will be fuzzy!!!
Chris
My main issue was items fairly close together (like the horse and pole) I do a stedicam routine where I place the couple next to a tree and that used to give me issues on the 130 as the camera refused to focus once I moved away from the couple and "swooped" around the tree and then as I came back to the couple the AF failed....also where weddings were in a gazebo the camera would shift from the couple to the back pillars of the gazebo for no reason but that was easy to solve with manual focus but on a stedicam you have to have auto so I ended up with using manual and shooting further away so the DOF covered the scene. If you want to replicate the troublesome area then you need the camera at say 15' from the subject and a second subject/object at around 20 -25' to "fool"
the AF ...it handles large distances perfectly .... I also found that doing Church ceremonys cutaway of the guests and zooming tighly on say, a group of 3 in the pews, the AF would often lock on to the people behind my group...it might just be a depth of field issue too??? With the iris quite wide and the camera zoomed a bit your effective DOF is small ..... It would be interesting to know what zoom was used with the horses as I seldom had any AF issues at full wide cos even if the focus was out both subjects were still within the DOF range. This might be what you are seeing??? At 15' and at F2 your DOF at 28mm is at least 4' whilst at 55mm it drops to just 12" so at full wide you can be in visual focus between two objects regardless of what the camera has locked onto but zoom to 55m and one will be fuzzy!!!
Chris
I would like to add my 2 cents. My Grandson came to visit. All the FAST soccer moves focused well with the 130. However, once indoors under good light, the shot was just a static shot of him sitting in a chair-no movement.
The camera went out of focus for NO REASON at all.
I normally don't use auto-focus except for very fast action. I forgot to disengage the auto-focus once indoors.
I just don't get it.
Why in the world would a camera go out of focus on a steady, well-exposed shot?
The updates are expensive and it appears to me that there are two more focus assists added but no addressing "WHY" the focus itself malfunctions. I never noticed this before until I used auto-focus. Having been a camera person since 1982 with pro cameras, I always utilized full manual Iris,shutter,focus and WB. This is a first and last time this will happen. I am going back to full manual or maybe purchasing another brand camera.
The camera went out of focus for NO REASON at all.
I normally don't use auto-focus except for very fast action. I forgot to disengage the auto-focus once indoors.
I just don't get it.
Why in the world would a camera go out of focus on a steady, well-exposed shot?
The updates are expensive and it appears to me that there are two more focus assists added but no addressing "WHY" the focus itself malfunctions. I never noticed this before until I used auto-focus. Having been a camera person since 1982 with pro cameras, I always utilized full manual Iris,shutter,focus and WB. This is a first and last time this will happen. I am going back to full manual or maybe purchasing another brand camera.
Last edited:
kwkeirstead
Well-known member
OK. Chris . . . I will do a static test where . . . camera is 15' from the subject and a 2nd object is at around 20 -25' to "fool".
where the subject at 15' stays put at that position for say 10 mins and we see if the camera decides to look elsewhere.
You OK with this as a valid test to see the problem or does the near subject need to sit there for 20 mins?
This is quite different from my 2nd test scenario where the subject leaves after xxx minutes and then comes back into the scene.
where the subject at 15' stays put at that position for say 10 mins and we see if the camera decides to look elsewhere.
You OK with this as a valid test to see the problem or does the near subject need to sit there for 20 mins?
This is quite different from my 2nd test scenario where the subject leaves after xxx minutes and then comes back into the scene.
Sorry it has taken me some time to respond. First of all it's a telephone/electrical pole. The pole was about 10 feet from me. The horse was about 75-100 feet from me. I think the zoom was close to all the way in. I will try to play the clip in the camera and see if I can see the read outs.
Since the pole was in the shot and the horse was standing there I immediately thought of this DVXuser board and the focus. SO I inched the pole into the shot to see how long it would take for the camera to say "hey make that edge of the pole sharp" or "new focus needed-- something new is in the center-- adjusting" BUT that NEVER happened. As you see I covered the area by 75% or more and the lens never focused until I had to pan to catch the moving horse.
If a pole and corn crops is not considered contrast enough for this camera -- I don't know what does. Secondly, I was outside on a sunny day and had plenty of lighting -- NO EXCUSES for a bad auto-focus, EVER.
Notice the other clip -- how EVERYTHING was out of focus when the horse ran. It wasn't until I pull out the focus motor woke up.
I have about a 3% clips that have rendered that section of footage useless. Sadly I never have this happen with the other auto-focus Panasonic cameras I owned. (DVC30, DVX100b (was my 450 auto focus?)
I tested the AC130 for the infinity setting and manual setting and neither gave me enough "keepers" to use reliably. Infinity setting never produced a focused image when the horse got close. Manual setting lacked the DOF that I would usually find in an ENG lens.
I would go back to DVC30 if I could, but my clients want HD footage.
Looks like Panasonic cameras are no longer made for action -- they are made for controlled environments of contrast. Looks like I need to see what this new JVC 600 will do.
-------------
I got a good one for you. I'll post another. I took video footage of a barn (~100 feet away) I zoomed in to allow the auto focus to catch it's focus and then I pulled out --- well guess what the auto-focus decided to change the focus to
3.3feet GIVE ME a break. Tell me how anything in this framing is 3.3 feet away. Heck the grass on the bottom is around 15-25 feet away. In the end I switched to manual, focused, pulled out and got my shot.
Since the pole was in the shot and the horse was standing there I immediately thought of this DVXuser board and the focus. SO I inched the pole into the shot to see how long it would take for the camera to say "hey make that edge of the pole sharp" or "new focus needed-- something new is in the center-- adjusting" BUT that NEVER happened. As you see I covered the area by 75% or more and the lens never focused until I had to pan to catch the moving horse.
If a pole and corn crops is not considered contrast enough for this camera -- I don't know what does. Secondly, I was outside on a sunny day and had plenty of lighting -- NO EXCUSES for a bad auto-focus, EVER.
Notice the other clip -- how EVERYTHING was out of focus when the horse ran. It wasn't until I pull out the focus motor woke up.
I have about a 3% clips that have rendered that section of footage useless. Sadly I never have this happen with the other auto-focus Panasonic cameras I owned. (DVC30, DVX100b (was my 450 auto focus?)
I tested the AC130 for the infinity setting and manual setting and neither gave me enough "keepers" to use reliably. Infinity setting never produced a focused image when the horse got close. Manual setting lacked the DOF that I would usually find in an ENG lens.
I would go back to DVC30 if I could, but my clients want HD footage.
Looks like Panasonic cameras are no longer made for action -- they are made for controlled environments of contrast. Looks like I need to see what this new JVC 600 will do.
-------------
I got a good one for you. I'll post another. I took video footage of a barn (~100 feet away) I zoomed in to allow the auto focus to catch it's focus and then I pulled out --- well guess what the auto-focus decided to change the focus to
3.3feet GIVE ME a break. Tell me how anything in this framing is 3.3 feet away. Heck the grass on the bottom is around 15-25 feet away. In the end I switched to manual, focused, pulled out and got my shot.
Mark Williams
Veteran
After doing considerable research for a new camera purchase I learned that with the AC130, AC160 and HPX250 the auto-focus issue just "is what it is". People who don't use it mostly don't understand why those that do gripe about it. I know where you are coming from with the DVC-30 reference. Mine always worked well. Even the HPX170 which I now own has a fairly good auto-focus. That's why I will never accept the lack of acknowledgment by Panasonic of the problem. There is just no way to shoot a fast action event without decent auto-focus. If you need it then these cameras are probably not the right choice. As far as I am concerned Panasonic should not even have an auto-focus feature on these cameras if it won't work correctly.
ChrisHarding
Well-known member
Hi Mark
They actually did produce a video saying that the AF was "different" but made no apologies for it...I have used Panasonic since the early 90's and all had good AF!! The 130's were the first that simply never worked correctly in the field ... ELN614?? I was filming people on a dance floor and the distance meter ran continously back and forth from 3' to infinity!! I was about 15' away so it's not only you!!
I kept my 130's for 3 months and shot 80 hours on each and then sold them and reverted back to HMC82's ...OK, smaller chips but the autofocus AND autoiris are magnificent ...before the 130's I had a pair of HMC70's then a pair of HMC80's and never had a shot out of focus in auto.
Chris
They actually did produce a video saying that the AF was "different" but made no apologies for it...I have used Panasonic since the early 90's and all had good AF!! The 130's were the first that simply never worked correctly in the field ... ELN614?? I was filming people on a dance floor and the distance meter ran continously back and forth from 3' to infinity!! I was about 15' away so it's not only you!!
I kept my 130's for 3 months and shot 80 hours on each and then sold them and reverted back to HMC82's ...OK, smaller chips but the autofocus AND autoiris are magnificent ...before the 130's I had a pair of HMC70's then a pair of HMC80's and never had a shot out of focus in auto.
Chris
LanceOregon
Well-known member
I've noticed this behavior sometimes too at the long end of the zoom range. As you report, pulling out and going to a wider zoom setting then acts to "wake up" the autofocus.
Chris, when I do dance performances (recitals mostly) I stay on manual focus and set my focus on 3/4 back on the stage and when I pull out everything appears in focus during the entire performance. fortunately they all stay on stage and they never stray far from my focus point. DVC30, if on auto focus, would lose focus during extreme low lighting. So even on the DVC30 I would use manual.
I too have the HMC80s (82 for non-USA model) I use it strictly for depositions. The zoom range on the 80's lens is not enough (similar to DVX100).
Jan_Crittenden
Veteran
Hi,
for those of you that said that we have not responded, that isn't entirely true. We have. I have tried on many occasions to explain why the AC130/160 auto-focus is different that those on the HPX170 and even more dramatically than the DVC30 or DVX100. First is the length of the lens, over twice as long as any of the above mentioned cameras, twice as heavy and the ballistics are very different. It is a completely different design than what we have delivered in the past. The second part is that with the HPX170, it is a 620,000 pixel imager, it resolves around a spatial offset and it has less resolution that that of the 130/160 which uses 2 million pixel imagers, more resolution, the more critical the focus and likely the more critical the camera is going to be. Quaduple that for the SD camera comparisons. Much easier to nail auto-focus on an SD camera. That said, we did put together a video clip that offers strategies on how to work with the Auto-focus of the AC130/160.
The link is here:
http://www.youtube.com/watch?v=I0to4OqIrBg&list=UUAtY7XCdtu016Uhyc4ngCSw&index=5&feature=plcp
Probably the most critical is the setting the area focus. And couple that with Push to Auto.
Best regards,
Jan
for those of you that said that we have not responded, that isn't entirely true. We have. I have tried on many occasions to explain why the AC130/160 auto-focus is different that those on the HPX170 and even more dramatically than the DVC30 or DVX100. First is the length of the lens, over twice as long as any of the above mentioned cameras, twice as heavy and the ballistics are very different. It is a completely different design than what we have delivered in the past. The second part is that with the HPX170, it is a 620,000 pixel imager, it resolves around a spatial offset and it has less resolution that that of the 130/160 which uses 2 million pixel imagers, more resolution, the more critical the focus and likely the more critical the camera is going to be. Quaduple that for the SD camera comparisons. Much easier to nail auto-focus on an SD camera. That said, we did put together a video clip that offers strategies on how to work with the Auto-focus of the AC130/160.
The link is here:
http://www.youtube.com/watch?v=I0to4OqIrBg&list=UUAtY7XCdtu016Uhyc4ngCSw&index=5&feature=plcp
Probably the most critical is the setting the area focus. And couple that with Push to Auto.
Best regards,
Jan
edit: Deleted inaccuracy on my part
The PBTS video you linked to has every example in LOW lighting situations. Which is understandable if it has trouble focusing!!!
Mine was in ideal lighting situation. Piezio Electric Auto Focus obviously has a lot of room for improvement.
re PBTS video: none of the strategies in your video show how to achieve focus while in auto-focus for capturing moving objects, outside in normal contrast and some may say high contrast- perfect lighting conditions with one point of light...all while the subject is centered over the predetermined Panasonic focus area.:badputer:
The PBTS video you linked to has every example in LOW lighting situations. Which is understandable if it has trouble focusing!!!
Mine was in ideal lighting situation. Piezio Electric Auto Focus obviously has a lot of room for improvement.
re PBTS video: none of the strategies in your video show how to achieve focus while in auto-focus for capturing moving objects, outside in normal contrast and some may say high contrast- perfect lighting conditions with one point of light...all while the subject is centered over the predetermined Panasonic focus area.:badputer:
Last edited:
The auto-focus is not as good as other cameras. I think that has been proven. On a plus side, I have been pleasantly surprised how well the "Face Recognition" focus options works. Whenever I want to use the auto-focus, I turn on this feature, it does seem to help a bunch.
Mark Williams
Veteran
"For the record the DVC30 has a 22x lens"
Incorrect, it had a 16x lens.
Incorrect, it had a 16x lens.
Barry_Green
Moderator
The DVC30 was also standard-def. And it was also 1/4". There's a huge difference between autofocusing a camera that resolves 0.3 megapixels per frame, and one that resolves 2.2 megapixels per frame. And there's also a huge difference in depth of field between 1/3" and 1/4", to the point where the DVC30 only had to be "vaguely accurate" for it to look in-focus, and an AC130/AC160 etc has to be hyper-accurate.
Hey, for a lark, how about putting the AC130 in DV mode and testing the autofocus? I wonder if it would appear more like what you'd expect from a DVC30?
Hey, for a lark, how about putting the AC130 in DV mode and testing the autofocus? I wonder if it would appear more like what you'd expect from a DVC30?
You are correct and I am VERY wrong. I thought it was a 20x or 22x -- I did a google search and I thought I read 22x. (camera is in use.. so it wasn't in front of me)
Barry, I will have to see about the SD I did, with the AC130 and see what it looks like.
So are you saying all other manufactures with this zoom range and chip size is having auto focus problems in HD??
Barry_Green
Moderator
I don't know of any other cameras with native HD chips and a 22x lens. I do believe that Jan posted that they did some tests against a Canon or Sony or something that had about an 18x or 20x lens, and that they found issues of comparable severity with those cameras, but I wasn't involved with that and didn't see that so I don't know what the particulars are.