On Tue, 19 Feb 2002 12:22:53 -0500, "RGBaker" wrote:

>> I have run into motion artifacts using DV. Serious enough that I could
>not
>> use the footage. I tape machinery for various reasons and found that a
>> crank handle rotating under power at a slow speed caused the dv codec to
>go
>> nuts. Large blocks followed the handle around the circle. Other than
>that,
>> only satellite tv has an visible artifacts that I've seen.

>Interesting. I have seen DV recordings collapse because of vibration caused
>by large machinery, i.e. standing on the deck of a ship, but those failings
>were caused by the mechanical vibrations affecting the actual recording
>mechanism, not the DV codec. The conclusion of the SMPTE in their study of
>DV25 compared to BetaSP was that the 'potential' for motion artifacts would
>be found in images somewhat different that you describe, i.e. a wide shot of
>multi-coloured leaves blowing gently in the wind would be expected to stress
>the codec, not the sort of gross motion you describe. Neither the SMPTE nor
>other studies I have seen have actually generated the motion artifacts
>described.

Hmmmmm........;-)
As the probable "originator" of the "artifacting with
motion" issue, let me try to clear the air a bit...
- D25, having a frame-by-frame compression scheme,
does not show the gross compression-effects we
hated with early broadcast digital ("steelwool
eyes" with blinking, crowd scenes that got weird
with pans).
- 3-chip D25 camcorders often show some obvious
motion problems (aliasing on near-horizontal lines,
and with the worst, on near vertical lines also).
- other compression effects, like the "mosquito"-type,
can look worse with slight motion with 3-chip
cameras due to the abrupt frame-by-frame changes
in the appearance of the mosquito effects.
- with 1-chip cameras, there are additional issues,
and motion-effects can range from moderate (as
with most good 3-chip cameras) to extreme - most
evident in the sharpest "megapixel" 1-chip
cameras. With these, good as their images can look
when there is no motion (hmmm.... - this *is*
video, not still-photography...;-), some "mangle"
some types of subjects, producing VERY unpleasant
motion-artifacting (try shooting clapboard-sided
houses, cityscapes, leaves floating in dark water,
light grasses against dark backgrounds, rock walls
with parallel layering, etc. with these - and if
the wild busy "flapping" and "stairstepping" in
the images don't drive you nuts, well...;-)
The above-described effects generally disappear when
motion stops. With good analogue 1-chip camcorders
(Sony TR700, TR200, for instance...), the image is
***CONSIDERABLY*** "quieter-looking" (though it is
also less sharp than D25 1-chip can be, more subject
dropouts, and more subject to generation losses).
Sometimes I wonder about using a TR700 as the camera
end, and a PC9 as the recording deck to combine
some of the advantages without having some of the
disadvantages of Mini-DV one-chippers - but then I
"snap out of it" and grab my VX2000....;-) Basically,
except for special-use needs, I have given up on
1-chip camcorders for all but casual use... (See:
www.David-Ruether-Photography.com/camcorder--comparison.htm
for Sony Mini-DV imaging-type comparisons - without
motion, some of those 1-chippers look MIGHTY good! ;-)
As for the "crank" effect described above, this
looks to me like a case of Mini-DV dropouts, or
possibly a case of the camera electronics mistaking
the motion for dropouts, and "correcting" them...
I will try to reproduce the effect...;-)