Image Stabilization

[Lucinda Luvaas]

I read on another forum that Image Stabilization, having it on, can cause stuttering and flickering while shooting in 24p or 30p. Has anyone else experienced or heard this?

Lucinda

[Pete Bauer]

I just shot some hand-held 30p this morning in very difficult low-light conditions and didn't see any OIS problem, and don't recall ever having a gripe with the 20x lens OIS; I think it rocks.

The general advice for any video camera is to turn off image stabilization if you are trying to do smooth pan starts and stops on a tripod...might that be what they're referring to?

[Lucinda Luvaas]

No, I don't think so.

The question posed was something that was posted here some time back about fluttery footage and interlacing using 30p, etc. and the advice given on this other forum was to turn off the image stabilization function. I haven't heard that before, so I was just wondering what others might think.

[Chris Hurd]

I haven't heard of that. You do have to be careful about OIS during night shots where bright lights are present, though. The light will reflect off of the internal prism within the OIS mechanism, and create little swirly fireflies in the image, dancing around the on-screen light sources. Outside of shooting on a tripod with OIS on (don't do it), that's the biggest OIS "gotcha" that I'm aware of.

When you mention "interlace" and progressive frame rates, be advised that when you shoot in 24p, whether it's with the 2:3 or 2:3:3:2 pulldown method, you are creating interlaced frames every so often, because that's how 24p works in these camcorders. Not related to OIS, though.

[Dan Vance]

Originally Posted by Chris Hurd

You do have to be careful about OIS during night shots where bright lights are present, though. The light will reflect off of the internal prism within the OIS mechanism, and create little swirly fireflies in the image, dancing around the on-screen light sources.

But, but, but...
You said the OIS "doesn't degrade the image." I'd call "little swirly fireflies dancing around on-screen" a degradation...
:)
But, semi-seriously, I've been trying to find out just how these stabilizers work, and whether they are as optically pure as claimed. One thing that is a puzzle is the mention of a prism. Not sure how a prism could be used in this application unless it's essentially being used as a positionable mirror. In any case, it is going to introduce chromatic aberration whenever it moves.
The other explanation of the system is that it uses two glass plates, filled between them with a high refractive index liquid. One plate moves in the yaw axis and the other in the perpendicular axis, to correct horizontal and vertical motion errors. The same problem would plague this design. Changing the angles of the plates introduces a prism effect, which would cause chromatic aberration.
I think the claim of "no degradation because it's optical" really means "the degradation is small, so most people won't notice."
And, all other things being equal, the two extras pieces of glass and the liquid that the light has to travel through have to be "worse" than not having those elements in there.
I say: Turn it off, leave it off...

[Martin French]

Not sure what type of stabilizer you are using but how about a LANC control mounted somewhere convenient to handle focus issues (which would be my main issue with the XL2)

http://www.b-hague.co.uk/Remote_camcorder_control.htm

On a curly cable it may be possible to have it strapped to your left had in some way. Always wanted to give that a shot with stabilizer work.

But i would agree, have fun with the Stedicam (or variant) and learn to compose and shoot good moving shots.

And be careful of on coming traffic! ;op

[Chris Hurd]

Originally Posted by Dan Vance

But, but, but... You said the OIS "doesn't degrade the image." I'd call "little swirly fireflies dancing around on-screen" a degradation...

Heh. I don't know about that, Dan... I'd call it an enhancement!

Okay, let's put it this way. Just humor me for a minute. Let's say that we define degradation in a rather strict way. Let's say (once again for the sake of argument) that when we say "degradation," what we're really talking about is a loss of resolution. That is, a loss of pixels.

Now, OIS may introduce the occasional *anomaly* such as the fireflies at night, or the stuttery zooms and pans on a tripod. After all, technological enhancements usually come bundled with a drawback or two, and OIS is no exception. It has a couple of key drawbacks, as noted above. But -- it doesn't cost you any *pixels* to use it, as can certain flavors of EIS (electronic image stabilisation).

But, semi-seriously, I've been trying to find out just how these stabilizers work, and whether they are as optically pure as claimed.

There are two basic flavors of OIS. You have the VAP, or Vari-Angle Prism, as found in the Canon XL series auto lenses. Then there's the Shift-Method OIS, as found in the smaller palmcorders such as the Canon Optura single-chippers. Neither method costs pixels to use.

OIS-VAP explained: http://www.canon.com/technology/dv/01.html

OIS Shift-Method explained: http://www.canon.com/technology/dv/02.html

I think the claim of "no degradation because it's optical" really means "the degradation is small, so most people won't notice."

If it's not noticeable, then why should it matter? Video is all about what's noticeable. Consider what's at stake and what is noticeable with OIS. Handheld at a very long focal length with OIS off equals shakycam. Not very watchable. Handheld at longer focal lengths with OIS on produces a much more watchable image. Outside of the instances where you shouldn't turn it on, I think the benefit of having OIS far outweighs any minute change in image quality which may or may not be there and which people can't detect even if it is there.

And, all other things being equal, the two extras pieces of glass and the liquid that the light has to travel through have to be "worse" than not having those elements in there.

I have seen the difference between having OIS turned on, and turned off, as viewed on a high-quality studio monitor, and I'd like to describe that difference for you. I could detect no change whatsoever in the "quality" of the image. What I could detect, however, and detect rather immediately as most people do, is that while handheld at just about any focal length other than full wide, OIS goes a long way to dampen any camera shake, and I'd rather watch the handheld video that had OIS turned on rather than off.

The tools are on the camera for a reason. Know the right conditions in which to use them and in which to leave 'em alone. The benefit of OIS is a tangible one. It produces pleasing results. I say, use it when you need it. That's what it's there for.

[Dan Vance]

...grr...grumble...grumble...
yeah, well okay, when you put it that way...

[Lucinda Luvaas]

I know that I must turn the stabilizer off when using the tripod and on when not. Unfortunately, due to fatigue and heavy heat, I forgot to turn it off when on the tripod today and did a reshoot...sometimes I forget and I know that when I am completely accustomed to all the settings on my new XL2 I won't, but what happens, I'm curious, to the footage if you leave the stabilizer on when the cam is on the tripod? Just curious.

[Chris Hurd]

The job of OIS is to counteract handheld camera jitter. When your camera is mounted on a tripod, the only kind of movement you have left over is intentional movement -- tilts, pans and zooms. If OIS is still turned on, it tries to fight those intended movements, resulting in a stuttery pan or a "bump" or pulsation in the image at the end of a zoom.

[Patrick King]

Chris,

I know how you feel about stabilization On while tripod mounted, but I have to say something in favor of it.

We shoot live with three XL1 cams to a stage far enough away that we are about 3/4 of the way through the optical zoom range of the 16x lens to frame a Medium Close-Up. We have inexperienced, volunteer camera operators, including myself.

We get consistently better results leaving OIS On. It takes out the shake when the camera operator places his/her hand on the tripod pan bar, it takes out the shake from little kids jumping down the balcony steps coming back the restroom, it takes out the shake from the shakes.

It does requires the camera operator to accomodate for the 'drift' you see when OIS is on while tripod mounted. We've bagged hundreds of hours doing this and it works out best for us with OIS On.

[Chris Hurd]

Patrick, I'm a firm believer in somebody using the gear as it best suits them. I suppose the point I'm trying to make is for the person who sees that image drift, and notices that little bump at the end of a zoom, they should know why that's happening and how to avoid it if they want to. If OIS on a tripod produces good results for you, then I say more power to ya.

[Steve House]

Unfortunately when OIS is on when the camera is on a tripod, it also sees rapid subject movement as camera shakes and tries to damp that out as well. Not good, since "moving" is the operant word in the phase "moving pictures" <grin>.

[Patrick King]

Steve,

In actual practice there is a threshold where the lens tries to dampen movement, like on a pan, and then if it breaks the threshold it doesn't fight the operator and then settles nicely on the new image.

[Stephen Duke]

Basicly the GENERAL rule is:

If your on sticks or changing the zoom, turn off the image stabalizer and auto focus.

The processor will contantly fight to adjust for the movement. If you have the stabalizer on a pan, then at the end of the pan you will have an elastic finish to your shot. And if it is a slow pan, it can judder. The stabalizer shouldn't effect the zoom unless its dark, and then there is a possibilty of the light refraction that Chris was talking about.

If you are on auto focus, then a slow pan or zoom, you will get focus 'breathing'.

Unless your picking up shots you have no control over, set the shot and shoot it manual. Always the best results.

IMHO

steve