RJT> WC> I've seen 1,000 hour camera tape machines advertized.

RJT> Oh really?  And these use standard tapes?  I wonder how they do that...

 Usually they use a stepper motor that advances
the tape one step, pauses, then one more step,
etc. This writes over the same spot on the tape
a few times. The last time being "saved" and
then advancing to the next step, pause, etc.

 You lose frames, but you get a much longer
record time. You get to choose the speed, and
therefore the number of frames to "throw away".

 Better systems will monitor the video and only
advance the tape when a change is detected. If
no change is detected within a short time, the
heads stop spinning. When a change is noticed,
the heads are brought up to speed and the
pictures are recorded from memory until the
tape recording "catches up", then they go on to
the tape directly.

 Even better systems will monitor several
cameras in this way and record only the ones
where changes are detected. The cameras are
recorded by time division, with memory saving
one or two cameras as another is going onto
the tape. The memory is recorded during the
appropriate time slices as the first camera
goes into memory.

 Camera "robots" can take this even farther by
letting you decide which part of a camera's
"vision" to monitor for changes and how big of
a change must occur. They can also give
presidence to one or more cameras over all of
the others. They can save more detail from some
cameras at the expense of all others, too. A
typical robot can control 16 cameras.

 We currently use a computer system, with video
going directly onto hard drives. It can (does)
do all of the above and more (alarm by time of
day, access control camera call-up, playback
during recording, etc.). Our system currently
supports 64 cameras, but we're building another
six story section, so we'll be expanding this a
bit.

 TTYL.

 Robert

 * SLMR 2.1a * Wreck the halls with cows on Harleys, fa-la-la-la-la.....
--- Platinum Xpress/Win/WINServer v3.0pr5