JVC HM100 Glitch in OIS at DVinfo.net

Anthony Shera · July 24th, 2009, 02:55 AM

I have found a glitch in what I assume is the OIS in the HM100. Initially I thought it must just be a recording glitch caused by the SDHC cards (they're 16GB Transcend).

However I now know this is not the case because it has occurred on different cards and I have seen the glitch occurring LIVE on the LCD screen as it's happened.

The glitch seems to appear occasionally when you have the camera at the full 10x zoom. It has occurred while I have had the camera mounted on the tripod, and also when it's been handheld. What will happen is that the image will very suddenly shift to one side by about 500 pixels and then back again, all taking place in about one tenth of a second. This is recorded onto the SDHC cards.

I believe that the OIS is way overcorrecting for very slight "micro" movements in the camera. When this issue occurred while the camera was mounted on a tripod, I was doing a couple of "pull focus" shots with the ring. The slight movement this created was obviously interpreted by the OIS as a movement that was going to be very large, but in fact was not.

Another time I had the camera resting on my knee at the full zoom. My knee wasn't as steady as I would have liked, but nonetheless it was nowhere near shaky enough to warrant a 500-pixel correction.

Fortunately this problem has thus far been quite rare. I have only observed it three times in the month I have used the camera, and in that time I've done a LOT of 10x zoom shots. I will have to do tripod shots in future with the OIS turned off and see if the problem ever arises again.

I would be interested to hear if I am the only one with this issue or if it's more widespread.

If it becomes annoyingly frequent in future, I guess I'll need to put my warranty to good use.

Chris Hurd · July 24th, 2009, 08:55 AM

Hi Anthony, judging from your description I'd have to say that there's nothing at all wrong with the camera. Instead, the symptoms described sound like a classic case of OIS fighting the tripod. In other words, OIS is fighting the intentional movement of the pan (or zoom). The remedy is to switch OIS to Tripod Mode (which is an admittedly facetious way of saying that OIS should be turned off when the camera is on a tripod or any other form of mount that isn't handheld).

OIS bumps and jumps have been a frequent topic of discussion here ever since I started this site back in 2001. An OIS bump (or jump, whatever you want to call it) occurs at the end of a zoom and the end of a pan or tilt. The greater the focal length (that is, the more telephoto the zoom), the bigger the jump appears.

You must understand that the objective of OIS is to dampen any slight unwanted movements that occur when holding the camera by hand (and it works only when you are doing your best to hold the camera as steadily as possible to begin with). The best OIS mechanisms are designed to compensate for a particular range of frequencies, such as the blood coursing through the veins of your hands and the motion your chest and shoulders make when you inhale and exhale normally. In some cases the frequency range is broad enough to compensate for the normal vibrations felt in an automobile when driving down a smooth road (such as Canon's "Super Range" VAP for instance).

When the camera is mounted on a tripod, there is no unwanted motion for OIS to fight... therefore the only motion it can compensate for is intended motion, such as a deliberate pan or zoom. OIS tries hard to counteract this motion and this is the cause of the jump at the end of a pan or zoom. OIS is simply trying to do its job, but the only motion it can attempt to "correct" on a tripod is that motion which you wanted in the first place -- and as a result it is working against you in that particular circumstance. That's why the manual says "turn off OIS when shooting from a tripod." That's why we say it too.

In situations where the tripod is outdoors in the wind, or if the tripod is lightweight and not very sturdy, or if it is a monopod, then usually you will not see an OIS bump in those conditions. Also some Sony camcorders offer various degrees of OIS effectiveness via menu selection, and the softer ones won't show a bump while the hard setting will. And a lower-grade OIS mechanism might not show it at all because it wasn't very effective in the first place.

Handheld in daylight, OIS is a very good thing -- but using OIS on a tripod can cause stuttering in pans and zooms. Bright lights at night can introduce unwanted "fireflies" if OIS is on. There is a proper time and place to use OIS, and a proper time and place to leave it off.

OIS does nothing to assist intentional camera movement (such as panning to follow a moving subject); the purpose of OIS is to dampen *unintended* camera movement (see below) and it can't do that from a tripod (except maybe in breezy or windy conditions, maybe).

Here's what the "OIS + tripod" artifacts look like --when panning: a "bump" in the image at the end of the pan. For example when panning left to right, at the end of the pan, you'll get a little bounce in the image back to the left, followed by a slight slide as it returns to the right and settles down. When changing focal length: a "bump" or a hiccup in the image at the end of the zoom, especially when going in to telephoto. At the end of the zoom, you'll get a slight pulse in the image as if it were going slightly wide and back in again. This is unwanted movement of the image (not the camera, but the image) in both cases.

Remember that OIS is intended to dampen a particular frequency range, most commonly the type of low-frequency vibration that's transmitted from your hands to the camera by the blood coursing through the veins of your hands, but also other types of situations where the camera is subjected to vibration within a certain frequency range.

However, on a stable mounting platform such as tripod, these vibration frequencies don't occur... but if OIS is left on, it goes to work anyway, and now it's working against you, because the *only* sort of vibrations it can detect now are the *intentional* motion inputs you're giving via panning and zooming. Having nothing else to fight, OIS does its best to counter that movement, resulting in the little glitchy stutter that occurs at the end of a pan and the end of a zoom. For this reason, you turn OIS off when shooting from a tripod.

Except, as has been pointed out, in breezy conditions outdoors, but in those situations you're better off experimenting with it on or off to determine what the best results are. Hope this helps,

Leonid Yegoshin · July 24th, 2009, 01:05 PM

Chris, you description is good but it doesn't take into account that camcoder have a knowledge of zooming and it can suppress a jumping after zoom just because it knows about zoom process in action. Moreover, it can predict the picture pixels basing on zoom value.

The panning is another story, of course.

Chris Hurd · July 24th, 2009, 01:37 PM

Quote:

Originally Posted by Leonid Yegoshin

...camcoder have a knowledge of zooming and it can suppress a jumping after zoom just because it knows about zoom process in action.

I know of no camcorder with that capability... your source, please?

Leonid Yegoshin · July 24th, 2009, 02:29 PM

Quote:

Originally Posted by Chris Hurd

I know of no camcorder with that capability... your source, please?

Difficult to say without knowledge of code but I am software engineer and worked with videocodecs and it is possible to compensate a movement like zooming. Even panning may be compensated to some extent - MPEG works with it each time and actually it is an MPEG tool for compression (motion prediction). Please take in mind that MPEG codec itself has no knowledge of camcoder zooming (it does it like electronic IS - estimate it) but it may be supplied from general camcoder control which knows the exact lens magnification factor.

However, my old Canon Optura 100MC doesn't have this kind of problem for sure. Moreover, I found in hard shaking situation it is better to use a slight zoom to smooth the overall picture.

Keith Moreau · July 24th, 2009, 05:47 PM

I had the same exact thing happen on a tripod the other day (but not in full zoom, just a little zoomed in.) It is much different than the 'latency' effect that is often one of the few problems of OIS while on a tripod. It was basically over in a flash, but it ruined the take. I immediately shut the OIS off and proceeded.

I've never experienced such a wild OIS artifact with any other camcorder, I think this is what Anthony was trying to say, and I agree with him. It's too bad, because it's another thing to remember to do, and sometimes I like to use OIS on a tripod to smooth out floor shakes or micro movements in windy and/or telephoto mode or even just smoothing out pans using a cheaper tripod head. It's also too bad that turning OIS on and off is so many button presses in the HM100 menu system, makes it inconvenient to deactivate a feature that might need to be deactivated in more situations than other camcorders.

The JVC OIS seems effective but it's different than other OIS systems I've used. If you have it on, look at the lens, there are in fact 2 glass panes in front of the real lens that are moving around ALOT. Because of this, I've had some problems with dancing lens flares when shooting with OIS handheld. I didn't notice it when I was shooting, but reviewing later it was like UFOs in the background (I don't think they were actual UFO's.)

Anyway, one more thing be aware of when using this camcorder.

Anthony Shera · July 24th, 2009, 06:56 PM

Thanks Chris and others. Unless I'm misunderstanding you Chris, you're saying that these glitches are known to occur at the end of a pan or zoom. I can't say that's occurred at all with my JVC HM100, and I've done a lot of pans and zooms. My problem has occurred when no zooming or panning is happening at all -- only an adjustment to the focus using the focal adjustment ring. But I suppose that the small movement that often happens during this procedure is sometimes enough for this OIS problem to manifest.

About manuals stipulating that one should turn off the OIS while using a tripod: yes, most camera manuals do say this. My last camera manual advised this, but the manual for the JVC says no such thing.

Keith, I was hoping this problem would only manifest at full zoom, but it seems your experience has shown otherwise, which is annoying. You said that the JVC's OIS is different from other systems you've used. Perhaps this is because it uses a "rotary OIS" as advertised in the brochure for this camera. I have no idea what that is or how it works, but perhaps that explains the difference you've noticed. But in any case, like you said, the most annoying part is having to navigate through the menu to turn OIS off.

Chris Hurd · July 24th, 2009, 11:37 PM

Quote:

Originally Posted by Anthony Shera

My problem has occurred when no zooming or panning is happening at all -- only an adjustment to the focus using the focal adjustment ring. But I suppose that the small movement that often happens during this procedure is sometimes enough for this OIS problem to manifest.

Anthony, can you please do a test and try the same focus movements from a tripod with OIS on and
again with OIS off. Your results will bear out whether or not this is OIS related, but I suspect that it is.

Tim Dashwood · July 25th, 2009, 03:38 AM

Quote:

Originally Posted by Anthony Shera

About manuals stipulating that one should turn off the OIS while using a tripod: yes, most camera manuals do say this. My last camera manual advised this, but the manual for the JVC says no such thing.

Actually it does. The HM100 manual advises to turn OIS OFF "when the camcorder is secured, such as when using a tripod." Page 66.

I also bring this up in the HM100 Tutorial DVD (already done... ready to start selling soon!)

Anthony Shera · July 27th, 2009, 02:26 AM

Quote:

Originally Posted by Chris Hurd

Anthony, can you please do a test and try the same focus movements from a tripod with OIS on and
again with OIS off. Your results will bear out whether or not this is OIS related, but I suspect that it is.

Well I could do a test, except that the problem only manifests VERY rarely!!! I have decided not to get too worried about it unless it manifests every time I do focus shots.

Tim, I look forward to that DVD. Also I overlooked that bit in the instruction manual obviously.

Tom Hardwick · July 27th, 2009, 04:09 AM

For the very first time last Friday, my Z1 started to exhibit the same fault. I had Steadyshot turned on, but the camera was at a wide focal length, absolutely rock solid on my Manfrotto. I watched in amazement as the image on my screen jerked about at irregular intervals. I had SS pipped to assign button 3, so it was easy to switch off, and from then on all was well.

I was filming about 25 metres from an electrified railway line, so could this have caused the problem? Of course I couldn't even use SS when I was hand-held - the bumps 'n' jumps were all too obvious. All SS settings (wide converter, soft, etc) were affected.

Later in the day I was filming miles away and all was well. I've never experienced this before, and I've been using SSSS (OIS) for 14 years now.

tom.

July 24th, 2009, 02:55 AM	#1
Anthony Shera Regular Crew Join Date: Jul 2009 Location: Brisbane, Australia Posts: 27	JVC HM100 Glitch in OIS I have found a glitch in what I assume is the OIS in the HM100. Initially I thought it must just be a recording glitch caused by the SDHC cards (they're 16GB Transcend). However I now know this is not the case because it has occurred on different cards and I have seen the glitch occurring LIVE on the LCD screen as it's happened. The glitch seems to appear occasionally when you have the camera at the full 10x zoom. It has occurred while I have had the camera mounted on the tripod, and also when it's been handheld. What will happen is that the image will very suddenly shift to one side by about 500 pixels and then back again, all taking place in about one tenth of a second. This is recorded onto the SDHC cards. I believe that the OIS is way overcorrecting for very slight "micro" movements in the camera. When this issue occurred while the camera was mounted on a tripod, I was doing a couple of "pull focus" shots with the ring. The slight movement this created was obviously interpreted by the OIS as a movement that was going to be very large, but in fact was not. Another time I had the camera resting on my knee at the full zoom. My knee wasn't as steady as I would have liked, but nonetheless it was nowhere near shaky enough to warrant a 500-pixel correction. Fortunately this problem has thus far been quite rare. I have only observed it three times in the month I have used the camera, and in that time I've done a LOT of 10x zoom shots. I will have to do tripod shots in future with the OIS turned off and see if the problem ever arises again. I would be interested to hear if I am the only one with this issue or if it's more widespread. If it becomes annoyingly frequent in future, I guess I'll need to put my warranty to good use.

July 24th, 2009, 08:55 AM	#2
Chris Hurd Obstreperous Rex Join Date: Jan 2001 Location: San Marcos, TX Posts: 27,366 Images: 513	Hi Anthony, judging from your description I'd have to say that there's nothing at all wrong with the camera. Instead, the symptoms described sound like a classic case of OIS fighting the tripod. In other words, OIS is fighting the intentional movement of the pan (or zoom). The remedy is to switch OIS to Tripod Mode (which is an admittedly facetious way of saying that OIS should be turned off when the camera is on a tripod or any other form of mount that isn't handheld). OIS bumps and jumps have been a frequent topic of discussion here ever since I started this site back in 2001. An OIS bump (or jump, whatever you want to call it) occurs at the end of a zoom and the end of a pan or tilt. The greater the focal length (that is, the more telephoto the zoom), the bigger the jump appears. You must understand that the objective of OIS is to dampen any slight unwanted movements that occur when holding the camera by hand (and it works only when you are doing your best to hold the camera as steadily as possible to begin with). The best OIS mechanisms are designed to compensate for a particular range of frequencies, such as the blood coursing through the veins of your hands and the motion your chest and shoulders make when you inhale and exhale normally. In some cases the frequency range is broad enough to compensate for the normal vibrations felt in an automobile when driving down a smooth road (such as Canon's "Super Range" VAP for instance). When the camera is mounted on a tripod, there is no unwanted motion for OIS to fight... therefore the only motion it can compensate for is intended motion, such as a deliberate pan or zoom. OIS tries hard to counteract this motion and this is the cause of the jump at the end of a pan or zoom. OIS is simply trying to do its job, but the only motion it can attempt to "correct" on a tripod is that motion which you wanted in the first place -- and as a result it is working against you in that particular circumstance. That's why the manual says "turn off OIS when shooting from a tripod." That's why we say it too. In situations where the tripod is outdoors in the wind, or if the tripod is lightweight and not very sturdy, or if it is a monopod, then usually you will not see an OIS bump in those conditions. Also some Sony camcorders offer various degrees of OIS effectiveness via menu selection, and the softer ones won't show a bump while the hard setting will. And a lower-grade OIS mechanism might not show it at all because it wasn't very effective in the first place. Handheld in daylight, OIS is a very good thing -- but using OIS on a tripod can cause stuttering in pans and zooms. Bright lights at night can introduce unwanted "fireflies" if OIS is on. There is a proper time and place to use OIS, and a proper time and place to leave it off. OIS does nothing to assist intentional camera movement (such as panning to follow a moving subject); the purpose of OIS is to dampen unintended camera movement (see below) and it can't do that from a tripod (except maybe in breezy or windy conditions, maybe). Here's what the "OIS + tripod" artifacts look like --when panning: a "bump" in the image at the end of the pan. For example when panning left to right, at the end of the pan, you'll get a little bounce in the image back to the left, followed by a slight slide as it returns to the right and settles down. When changing focal length: a "bump" or a hiccup in the image at the end of the zoom, especially when going in to telephoto. At the end of the zoom, you'll get a slight pulse in the image as if it were going slightly wide and back in again. This is unwanted movement of the image (not the camera, but the image) in both cases. Remember that OIS is intended to dampen a particular frequency range, most commonly the type of low-frequency vibration that's transmitted from your hands to the camera by the blood coursing through the veins of your hands, but also other types of situations where the camera is subjected to vibration within a certain frequency range. However, on a stable mounting platform such as tripod, these vibration frequencies don't occur... but if OIS is left on, it goes to work anyway, and now it's working against you, because the only sort of vibrations it can detect now are the intentional motion inputs you're giving via panning and zooming. Having nothing else to fight, OIS does its best to counter that movement, resulting in the little glitchy stutter that occurs at the end of a pan and the end of a zoom. For this reason, you turn OIS off when shooting from a tripod. Except, as has been pointed out, in breezy conditions outdoors, but in those situations you're better off experimenting with it on or off to determine what the best results are. Hope this helps, __________________ CH Search DV Info Net \| 20 years of DVi \| ...Tuesday is Soylent Green Day!

July 24th, 2009, 01:05 PM	#3
Leonid Yegoshin New Boot Join Date: Jun 2009 Location: Santa Clara, CA Posts: 20	Chris, you description is good but it doesn't take into account that camcoder have a knowledge of zooming and it can suppress a jumping after zoom just because it knows about zoom process in action. Moreover, it can predict the picture pixels basing on zoom value. The panning is another story, of course.

July 24th, 2009, 05:47 PM	#6
Keith Moreau Major Player Join Date: Nov 2006 Location: Brisbane, California Posts: 530	I had the same exact thing happen on a tripod the other day (but not in full zoom, just a little zoomed in.) It is much different than the 'latency' effect that is often one of the few problems of OIS while on a tripod. It was basically over in a flash, but it ruined the take. I immediately shut the OIS off and proceeded. I've never experienced such a wild OIS artifact with any other camcorder, I think this is what Anthony was trying to say, and I agree with him. It's too bad, because it's another thing to remember to do, and sometimes I like to use OIS on a tripod to smooth out floor shakes or micro movements in windy and/or telephoto mode or even just smoothing out pans using a cheaper tripod head. It's also too bad that turning OIS on and off is so many button presses in the HM100 menu system, makes it inconvenient to deactivate a feature that might need to be deactivated in more situations than other camcorders. The JVC OIS seems effective but it's different than other OIS systems I've used. If you have it on, look at the lens, there are in fact 2 glass panes in front of the real lens that are moving around ALOT. Because of this, I've had some problems with dancing lens flares when shooting with OIS handheld. I didn't notice it when I was shooting, but reviewing later it was like UFOs in the background (I don't think they were actual UFO's.) Anyway, one more thing be aware of when using this camcorder.

July 24th, 2009, 06:56 PM	#7
Anthony Shera Regular Crew Join Date: Jul 2009 Location: Brisbane, Australia Posts: 27	Thanks Chris and others. Unless I'm misunderstanding you Chris, you're saying that these glitches are known to occur at the end of a pan or zoom. I can't say that's occurred at all with my JVC HM100, and I've done a lot of pans and zooms. My problem has occurred when no zooming or panning is happening at all -- only an adjustment to the focus using the focal adjustment ring. But I suppose that the small movement that often happens during this procedure is sometimes enough for this OIS problem to manifest. About manuals stipulating that one should turn off the OIS while using a tripod: yes, most camera manuals do say this. My last camera manual advised this, but the manual for the JVC says no such thing. Keith, I was hoping this problem would only manifest at full zoom, but it seems your experience has shown otherwise, which is annoying. You said that the JVC's OIS is different from other systems you've used. Perhaps this is because it uses a "rotary OIS" as advertised in the brochure for this camera. I have no idea what that is or how it works, but perhaps that explains the difference you've noticed. But in any case, like you said, the most annoying part is having to navigate through the menu to turn OIS off.

July 27th, 2009, 04:09 AM	#11
Tom Hardwick Inner Circle Join Date: Mar 2003 Location: Billericay, England UK Posts: 4,711	For the very first time last Friday, my Z1 started to exhibit the same fault. I had Steadyshot turned on, but the camera was at a wide focal length, absolutely rock solid on my Manfrotto. I watched in amazement as the image on my screen jerked about at irregular intervals. I had SS pipped to assign button 3, so it was easy to switch off, and from then on all was well. I was filming about 25 metres from an electrified railway line, so could this have caused the problem? Of course I couldn't even use SS when I was hand-held - the bumps 'n' jumps were all too obvious. All SS settings (wide converter, soft, etc) were affected. Later in the day I was filming miles away and all was well. I've never experienced this before, and I've been using SSSS (OIS) for 14 years now. tom.