AC130: Video footage of the Auto Focus in action

Hi Mike

Most 1/3rd chip cameras have manual ND's ..I never considered the tint auto-iris range as a defect or a fault ..it simply never suited my application as a solo shooter. As with Cairns, Perth in Summer when we have tons of outdoor weddings can have huge light changes and the bride seldom considers the lighting situations for either the photog or video guy..It simply "This is a pretty spot, we will get married here!" The auto-ND and shutter range of the HMC82 simply is more suitable when you have to have unattended cameras ..I would suspect the the AC-90 would also do a good job on it's own. Yep, I have seen posts regarding the AC-90 already saying "Where are the ND filters" so many consider the lack of manual ND's to be non-professional.
It's the old story ... you need the tool that works best for the job.
On the JVC front it will be indeed interesting to see if the camera actually has the same 160 autofocus issue (not that I'm getting one)

Chris
 
It doesn't struggle 90-95% of the time. It is in focus!!! I said in this thread it works MOST of the time!
Just perhaps 2% of the time is just plain wacks out. Like a broken camera with an intermittent problem. If I send this into Panasonic.. they will do what car dealerships do. Test it once and see that it works and say "nothing wrong with it" and send it back. Intermittent problems are hard to show when they are just that-- intermittent. You notice how if car dealerships can't duplicate the problem they won't fix it. Same thing would happen with Panny. Heck, they already came out and said this is normal and it is suppose to lose focus while in auto focus. (granted they said that in a very convoluted / beat-around-the-bush kinda way.)

Two points here....from my experience with Panasonic service in Hollywood anyways, I do not think they would treat you like that. My bet is they would do their level best to find the problem and fix it. I would therefore encourage you to take it in sometime and see what they say. Could not hurt. Panasonic will pay attention to the issue if they get a lot of cameras back for service of the AF system.

My other point is, if that is true that your camera goes out of focus only 2% of the time, and achieves focus 98% of the time, well...that's actually extremely good. The HMC cannot do that! Nor can the Canon XHA1. The HMC won't go out in the same way, but it will occasionally focus on something it shouldn't just to keep you on your toes. Sometimes the focus is just "soft".

Sometimes, at concerts or sports like soccer, it will focus on whatever is closest, seemingly just because it can. Very difficult to manage that, especially when shooting Soccer!

I know of no cam that I have ever used that can get focus right in AF 98+% of the time. Is it really frustrating that the AC130/160 can go wonky on focus sometimes? Sure! No denying that.

I do not mean any disrespect, I am just saying--asking, rather-is the expecation that any cam achieve a greater than 98% success rate in AF mode really reasonable? 'cause I know of no cam that can actually do that. Not the HMC150, Canon XHA1, Sony NX5U, etc.

I do hope that manufacturers wil continue to refine their offering's AF performance, as it is a very crucial ability to have.

Just my two cents....
SW
 
Last edited:
sewolla said:
from my experience with Panasonic service in Hollywood anyways, I do not think they would treat you like that. My bet is they would do their level best to find the problem and fix it. I would therefore encourage you to take it in sometime and see what they say. Could not hurt. Panasonic will pay attention to the issue if they get a lot of cameras back for service of the AF system.

Just my two cents....
SW
Click to expand...

From my point of view I hope so- time will tell........
 
sewolla said:
My other point is, if that is true that your camera goes out of focus only 2% of the time, and achieves focus 98% of the time, well...that's actually extremely good. ...

I know of no cam that I have ever used that can get focus right in AF 98+% of the time. Is it really frustrating that the AC130/160 can go wonky on focus sometimes? Sure! No denying that.
Click to expand...
I'm sorry can you point at some auto focus samples from others that are equally as bad as mine? (from the models you just mentioned) I would like to see them -- I just may change my whole position.

roughly 2% of the time the shot turns out horrible. still another 3-8% of the time I have shots that go out of focus for a small amount and for my application is still usable but for other's it wouldn't be. leaving the average to about 1 in 10 shots being bad in some way. Which is why we use manual -- but again, in some shooting environments auto is necessary.
 
Sorry but I tend to agree with ELN614 here too!!

If you had a car that 2% of the time wouldn't change into top gear would you live with it?? You would have it back at the dealership pronto and complain bitterly that there is something wrong with the gearbox and hopefully they would fix it!!
In my books either something works 100% or it doesn't work 100% ..there is no way I would put up with an issue that might ruin a maybe unrepeatable shot. If auto doesn't work at all zoom ranges then surely it shouldn't have auto at all?? or at least a proviso that auto is only available up to 10X zoom or whatever. I had a wedding where the focus went soft right in the middle, an unrepeatable event, and I could do nothing about it. Nothing in the manual says "Don't use autofocus" (Just for the record that was a bright sunny day, lots of contrast and only about 3X zoom)....the next time I used manual which of course was perfect but why should I have to use manual when auto is available ?
I still think he has a very valid issue about the camera's autofocus ... all the technical jargon and physics aside ..if it's there it should work. On that day I was pretty upset with the camera, trust me, so I feel for ELN614 big time!!!

Chris
 
ChrisHarding said:
If you had a car that 2% of the time wouldn't change into top gear would you live with it?? You would have it back at the dealership pronto and complain bitterly that there is something wrong with the gearbox and hopefully they would fix it!!
Click to expand...
Thank you Chris!
I little different analogy .. If 2% percent of the time your car would stall-out completely when driving on the highway in cruise control. I suspect the second your car stalls-out you would be heading right to the mechanic.
And then 3-8% of the time, when on cruise control, the car would just drop out of cruise and slow down and then on it's own jump back into cruise control. I suspect you'd be saying something is wrong.
Instead, many of you are saying "Cruise is for amateurs... true men don't use cruise."
 
Easy to see from the "horse" video that unless you have three hands, it's pretty important to have AF working consistently, assuming you are able to track the horse as it is moving.

This is quite a different situation from the non-practical one where we try to stress the camera by quickly going from a distant object to a near object.

The horse is not running so fast that the camera could be excused for dynamically failing to adjust as it gets further away/closer.
 
Last edited:
ChrisHarding said:
If you had a car that 2% of the time wouldn't change into top gear would you live with it?? You would have it back at the dealership pronto and complain bitterly that there is something wrong with the gearbox and hopefully they would fix it!!
Click to expand...

That's a bad analogy. It is quite easy for an auto transmission in a car to select the correct gears. There is no such thing as an auto focus that works perfectly all the time. This situation is a case of degree. Some people think the number of times it fails is acceptable, considering the lens and sensor, whereas others think it fails too often. For all but a couple of these sample videos, I think I can see why the AF failed and they don't surprise me. But those couple of samples are a mystery.

Did anyone here ever use a camera with IR or ultrasonic AF? Terrible systems, both of them. :)
 
Unmanned tripod mounted b roll- stedicam in group movement- crane shots- horses- sports- motor racing- wedding track shots- group scenes etc- All of these things are requiring auto to work and it should!

this is not a cheap camera, and we as professionals come across these events where auto is essential.(I have recently done squash- badminton- and football with my Sony on auto- all good. I had to abandon the Panasonic because it was three times as expensive and couldn't follow focus in close zoom fast action scenes reliably).

In the squash court I was actually in the court! ! bloody scary I can tell you! It was foot action we needed and the Panasonic just couldn't do it. Luckily I had the Sony as a backup. The Panasonic would sometimes lock and just not come back-(but of course just couldn't keep up- we know and accept it is a slow focusser.) this, in a well lit court with buckets of movement and contrast.
 
Last edited:
Hi Mike

My very first Panasonic camera had an ultrasonic focus..Model WVP-100 ..no CCD chips either!! It had a Saticon tube!!! It fed back into a NV100 VHS recorder slung over my shoulder.
Now I have to question your statement ...I have been on HMC82's now for 2010/2011 and 2012 (previous I had 2 x HMC72's) I have never seen one bit of footage that is out of focus and I use AF extensively. In the wedding season I am shooting a wedding every weekend otherwise I doing up to 7 Realty shoots a week (all on auto BTW) and I can safely say that my auto focus is 100% reliable...on the 130's it failed miserably on the 2nd wedding shoot!! Some autofocus cameras actually do work all the time with 100% success rate and at wedding the lighting at the end of the night is often really bad yet ..the camera locks on perfectly... Not just lucky cameras either..I change them every year to offset my tax so I'm on the third pair now!!

Chris
 
ELN614 said:
I'm sorry can you point at some auto focus samples from others that are equally as bad as mine? (from the models you just mentioned) I would like to see them -- I just may change my whole position.

roughly 2% of the time the shot turns out horrible. still another 3-8% of the time I have shots that go out of focus for a small amount and for my application is still usable but for other's it wouldn't be. leaving the average to about 1 in 10 shots being bad in some way. Which is why we use manual -- but again, in some shooting environments auto is necessary.
Click to expand...



Well if you are saying that focus is out up to 8% to 10% of the time, that is a different matter. I thought you were saying it goes out a total of 2%. I still say it would be very difficult to find a camera by any manufacturer that could achieve dead-on AF performance more than 98% of the time. I have never seen any camera that I have ever worked with going back to Sony VX2100/PD170's, etc.

I cannot point to the same condition occuring in the HMC, or Canon, as you have occuring in the horse shots. Never seen that before not even in my AC160. You really should send it in, in my humble opinion.

It would not be too hard to find examples where it (Canon XHA1, HMC150) focused on the closest object, not necessarily what I was aiming at or intending it to focus on. In those cases, it just looks soft. Not as bad as your examples with the horse. But bad enough that you don't want your client seeing it.
 
Last edited:
Right, in the interests of objectivity I have just completed another near /far test which will be uploaded shortly.

This time I have used a high res full contrast black and white test chart, again on full auto with lens open. I used full frame analysis this time to ensure the lens saw every shape and tone available. Mike, this addresses the zero contrast issue you mentioned as a possible cause - it don't get much contrastier than black and white. Results coming up shortly.
 
ChrisHarding said:
My very first Panasonic camera had an ultrasonic focus..Model WVP-100
Click to expand...

You have a good memory. I don't remember the model numbers of most of the cameras I've owned, but that model number does ring a bell.

<searches>

Yep, that was it:
http://www.oldvcr.tv/collection/images/106-WVP-100.jpg

I had one of those too.

I also remember one other model I had: the AG450, which I heavily modified (variable speed zoom and improved power and audio functions).
http://mike-warren.net/play/060-web.jpg

Now I have to question your statement ...I have been on HMC82's now for 2010/2011 and 2012 (previous I had 2 x HMC72's) I have never seen one bit of footage that is out of focus and I use AF extensively.
Click to expand...

In my defense, I've not used the 70 or 80 series, so that's probably a combination of Panasonic getting the AF quite good on those coupled with the deep DOF inherent in those cameras. Also, I have not used a large number of different video cameras in the last 10 years or so. But those I have used have not been all that much better than what I remember. From about the mid 80s to the late 90s I used to repair Panasonic, JVC and Sony cameras.

There are a couple of videos posted by ELN614 that I can't explain so I'm looking forward to Rod's new test.
 
Last edited:
Here is a further test. Maybe not lab quality but contains and covers more of the suspected missing factors such as shape, contrast vertical, horizontal and diagonals.

 
Do you still have it set up? Can you put the charts the correct way up and test again. No need to post the video, but does it behave the same? The problem I have is that there aren't any vertical lines within the focus area, exactly the sort of situation I think would fool the AF.

Did you notice how it does find focus as you move off the close chart and a strong vertical contrast line appears in the focus area?

ac160focus05.jpg
 
Last edited:
Mike there are vertical lines in these charts- Crikey! how finnicky is this focus system? Sometimes it gets there, sometimes it doesn't. In the real world we have to shoot what is in front of us generally. And we don't know what the focus area is - I have asked Barry but he said Panasonic have not published that information. That's why I filled the frame and switched off the area function so that it would not just see the centre area as a grey card.

However, I will set this up and try again in the interests of science as it were, but these are easy targets and if it can hit them once it should be able to repeat.
 
Last edited:
I understood that with the area function switched off it still gets the data from the centre of the frame, but I could be wrong. It's frustrating me that I have to make so many assumptions, not having access to a camera.

However, if it does the same thing with everything identical in your test apart from the card rotated I'd have to agree that the AF is significantly worse than it needs to be. But it simply can not be as good as a smaller lens on a 1/4" chip camera. The mass of the focus element is probably 5 to 10 times greater in this lens compared to a 12x on a 1/4" chip. (Educated guessing again).
 
I just tried to find some more information in the manual, but it's pretty thin on the ground.

However, this sentence on page 101 amused me:

Depending on the lens, the auto focus function may not be available.
Click to expand...

That could mean pretty much anything.
 
Mike Warren said:
I understood that with the area function switched off it still gets the data from the centre of the frame, but I could be wrong. It's frustrating me that I have to make so many assumptions, not having access to a camera.

However, if it does the same thing with everything identical in your test apart from the card rotated I'd have to agree that the AF is significantly worse than it needs to be. But it simply can not be as good as a smaller lens on a 1/4" chip camera. The mass of the focus element is probably 5 to 10 times greater in this lens compared to a 12x on a 1/4" chip. (Educated guessing again).
Click to expand...

Hi again Mike, Yes, but the focus area is still an unknown-have you any further data on that?- it would help.

Regarding mass etc, we have discussed this point. I think we all accept that it is a massy lens (because we are told so I guess) and we all know it is slowwwww to focus.


However,the whole issue is inconsistent repeatability on identical parameters. If it can focus the mass 66% of the time why not every time?
I used to write path programs and training manuals for six axis industrial robots. These could repeat a movement within thousanths of an inch. I know they are vastly bigger servos etc but the manufacturers were also in a competitive market and reliability, durability and repeatability were what sold their machines, as well as having to be price competitive. I guess the point of my ramble is that whatever it is , it has to perform as described and meet expectations of the specification. (fit for purpose).

But yes, as I said, I will carry out some more testing. ( but I can't help but feel there is an elephant in the room!!)
 
Rod Pike said:
Hi again Mike, Yes, but the focus area is still an unknown-have you any further data on that?- it would help.
Click to expand...

No, but cameras typically only use an area in the center of the frame and I remember (possibly falsely (old brain)) someone saying that the 130/160 also did this.


However,the whole issue is inconsistent repeatability on identical parameters.
Click to expand...

However, these tests are not repeating the same actions. Similar, but not identical, so the test, although demonstrating something is amiss, is not supplying enough data to say exactly what the problem is.

I'm not trying to be argumentative, just trying to help you all get to the bottom of this, which is not easy without access to a sample camera. Yet again I feel I should have just remained quiet. I hate trying to get answers without enough data.

If the HPX250 had done 1080p60 I'd already own one, but while I was procrastinating on getting an AC160 I noticed that the lens goes out of focus on fast zooms even in manual focus mode and that's unsuitable for what I would be using it for (wildlife).
 
You must log in or register to reply here.
Back
Top