On Thu, 21 Nov 2002 01:40:59 -0500, "Sanman" wrote:

>>
>> I have seen no difference in bright light with
>> the stabilizer on or off on Sony 1-chip cameras
>> (all have EIS). In low light, you can trick it into
>> using 1/60th, and then also see no difference (as
>> I recall, switch to manual exposure, locked, then
>> turn on the stabilizer, then unlock the exposure...).
>> David Ruether
>> d_ruether@hotmail.com
>> http://www.David-Ruether-Photography.com
>> Hey, check out www.visitithaca.com too...!

>David;
>
>I'm not sure what Sony camera you are using. In the TRV-103, and that
>series of cameras, there is a slight blurring effect that comes and goes
>after you have moved the camera around (stabilization ON). I first noticed
>it shortly after I bought the camera and thought it was my auto focus
>hunting. Turning the auto focus off did not eliminate the effect, which is
>very slight I might add. Turning off the SteadyShot stopped it. Thinking
>something was wrong with my camera, I went to Sears near by, where they have
>all the cameras hooked up to TVs so you can play, and all the Sony models
>did the same thing.

I was referring to Sony DV models - is the 103 8mm? Hi-8?
The older TRV400 Hi-8 did lose resolution with the
stabilizer engaged, but the TRV700 did not, and no
Sony DV model I've seen (a LOT! ;-) did either...

>It kinda makes sense, given how this stabilization works. The image from
>the lens fills the entire area of a 460,000 pixel CCD. All of the
>photo-diodes on the chip are exposed as you would expect. A square area in
>the centre that is smaller than the total size of the chip is what is used
>for the video. When you move the camera, sensors translate the movements
>and in turn electronically shift that smaller square area around,
>compensating for the movements. You jerk the camera up, the square of
>pixels that are being used for video moves down at the same time, attempting
>to counter your movement, kinda like when you turn your head, but keep
>looking at the same object. It's like an electronic pupil that moves around
>to smooth out the bumps. All this requires that pixels be switched on and
>off, as the square moves around. When you stop moving the camera, the
>square slowly moves back into the centre of the CCD and rests there until
>the nest movement. It's during THAT period that I notice the blurring, as
>the square of "ON" pixels slowly sneaks back to the middle of the chip.
>Once it's settled, everything looks fine again. Just turning pixels on and
>off might cause jerky movement, so they slowly shift this square of ON
>pixels, sort of fading or interpolating between rows of pixels and that's
>what causes the effect. One thing; it's not as bad when zoomed all the way
>out. That's because the SteadyShot is practically inactive on wide shots.
>As soon as you even start to zoom in, the SteadyShot kicks in full force
>and, well all of the above. Try focusing your camera on a brick wall, or
>something with sharp edges. Put it on a tripod. Shoot a few seconds and
>then bump it a few times. Make sure you're zoomed in slightly. You'll see
>it, I'm sure.
>
>As I say, if it wasn't for my experiment in Sears, I would have returned my
>camera and asked for a new one.
>
>I was disappointed when I found this out. My previous camera, a TR-101,
>which is sitting in my basement with a broken roller guide, has mechanical
>SteadyShot, using a fluid filled prism that changes shape slightly when the
>camera moves, thus refracting the light to counter the movements. It didn't
>suffer any of these anomalies and I, in fact, much prefer that over the
>electronic version.
>
>I hope I haven't sent hundreds of ppl rushing out to check this and to be
>disappointed. Sometimes what you don't know won't hurt.

;-) I will give it a look, but I have not seen this effect
in DV cameras so far...

>Oh, and the little trick, fooling the camera into 1/60th shutter speed while
>SteadyShot is in WORKS! Thank you!!

Ah, good - I had trouble recently explaining it to
someone in person, glad my memory served better here...!;-)