[comp.graphics] WORM drives as animation output devices

jbn@glacier.UUCP (12/05/87)

     Anyone trying to use a WORM drive as a buffering medium for the creation
of video one frame at a time?  Ideally, one would like to create CD-V disks
frame by frame, something which should not be fundamentally impossible.
Next best would be to use a WORM disk as a buffering device for the
digital accumulation of frames to be played back later at high speed by
a suitable play program, at which time, of course, one could record the
output on videotape.

					John Nagle

siegel@svax.cs.cornell.edu (Alexander Siegel) (12/05/87)

In article <17231@glacier.STANFORD.EDU> jbn@glacier.STANFORD.EDU (John B. Nagle) writes:
>
>     Anyone trying to use a WORM drive as a buffering medium for the creation
>of video one frame at a time?  Ideally, one would like to create CD-V disks
>frame by frame, something which should not be fundamentally impossible.
>Next best would be to use a WORM disk as a buffering device for the
>digital accumulation of frames to be played back later at high speed by
>a suitable play program, at which time, of course, one could record the
>output on videotape.
>
>					John Nagle

Sure.  Sony makes a video disk recorder and player which do exactly this.
I've used them to record a movie one frame at a time over a period of
several days.  Very nice.  You can roll forward or back at nearly any useful
speed, and the single frame can be manually and computer controlled.  High
quality too.  Unfortunately it cost $$$!  20k is a good starting price.

-- 
Alex Siegel
Computer Science at Cornell University, Ithaca NY  14853
siegel@svax.cs.cornell.edu (ARPAnet)
siegel@CRNLCS (BITNET)
{uw-beaver,ihnp4,decvax,vax135}!cornell!siegel (UUCP)

dave@onfcanim.UUCP (12/07/87)

In article <17231@glacier.STANFORD.EDU> jbn@glacier.STANFORD.EDU (John B. Nagle) writes:
>
>     Anyone trying to use a WORM drive as a buffering medium for the creation
>of video one frame at a time?  Ideally, one would like to create CD-V disks
>frame by frame, something which should not be fundamentally impossible.
>Next best would be to use a WORM disk as a buffering device for the
>digital accumulation of frames to be played back later at high speed by
>a suitable play program, at which time, of course, one could record the
>output on videotape.

There are analog WORM video disks, typically with 13 or 20 minutes
recording time on a disk.  These can record single frames under
computer control if they have an RS-232 command interface, provided you
can give them NTSC-encoded images.  Very little computer hardware is
capable of real NTSC video, by the way.  (To see what is involved in
producing NTSC video, see the EXCELLENT article in a recent IEEE CG&A
by DeFanti and friends; I don't have it here so can't give an exact
citation).

Unfortunately, these discs are expensive (several hundred dollars each)
and can only be played back on the recorder, or other players designed
for that disk format, *not* on standard laserdisk or CD-V players.
The WORM recorders use special disks that are pre-formatted with track
markings on the disk, which the recorder uses to determine where to write.
The optical head positioning is done in essentially the same manner used
by CD and Laserdisk players.

Laserdisk and CD-V, on the other hand, have only the data itself on the
disk, which makes the mastering machines very expensive since they have
to contain measuring equipment capable of positioning the beam with an
accuracy in the micron range with no help from the disk.  This is why
you can't write Laserdisk or CD-V format disks unless you are willing to
spend a very large amount of money.

There are relatively inexpensive digital WORM disks available, but they
do not have the bandwidth to support real-time playback.  If NTSC video
is digitized as a composite waveform, it is usually done at 4 times
the subcarrier frequency with at least 8 bits/sample.  That's a 14 Mb
(megabyte)/second data rate, while WORM drives seem to do about 500 Kb/s.
If you store the images in digital form before NTSC encoding, that's
(say) 640x486x3 = 933 Kb/image; at 30 frames/sec that's 28 Mb/s.
Digital WORM disks are a great way to store digital images between the
rendering and recording phases (so you don't tie up your single-frame
recorder for weeks, recording one frame every hour), but you still need
something else to do the single-frame video.

Another device starting to appear that should be great for intermediate
storage is a drive that stores several Gb of data on a videocassette.
It doesn't allow random access, but that's fine for animation.  It
holds about as much as the digital WORM drives, so you have to change
media far less often when recording than with 1/2 inch tape (2400 ft
@6250 BPI = 140 Mb/tape).  But the media cost is perhaps 1/50 that of
the WORM drives and also much less than for 1/2 inch tape, and you can
re-use it when your images have been safely transferred to film (oops, video).

I don't have one of these yet, but when the controllers become readily
available....

	Dave

king@dciem.UUCP (Stephen King) (12/09/87)

In article <17231@glacier.STANFORD.EDU> jbn@glacier.UUCP writes:
>
>     Anyone trying to use a WORM drive as a buffering medium for the creation
>of video one frame at a time?  

A WORM drive designed for data probably uses SCSI interface. Not enough
bandwidth for moving video images. Teac makes a laser video WORM system
that will do the job, but the price is somewhat prohibitive for most
people ( >$20,000 for the recorder, >$100 each disc (Canadian dollars)).
The discs from the Teac system are not playable on anyone elses equipment.
McDonnell-Douglas builds a write-once laser FILM recorder which uses a
photographic medium. This is a transmission system, instead of reflection,
and the discs are floppy. Again, the price is high, although the media
should be cheaper, and it's not compatible with anything else.  ...sjk

-- 
 * Defence & Civil Institute *		...!{utzoo | mnetor}!dciem!king 
 * of Environmental Medicine *		Stephen J King
- Simulation & Training Group -		(416) 635-2149