Newsgroups: sci.image.processing,comp.sys.ibm.pc.hardware.video,rec.video
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!gatech!newsfeed.pitt.edu!uunet!zib-berlin.de!ceres.fokus.gmd.de!nntp.gmd.de!news.rwth-aachen.de!news.rhrz.uni-bonn.de!mpifr-bonn.mpg.de!speckled.mpifr-bonn.mpg.de!mlelstv
From: mlelstv@speckled.mpifr-bonn.mpg.de (Michael van Elst)
Subject: Re: True 30 fps sustained video capture solution
Message-ID: <1995Jan19.011019.14197@mpifr-bonn.mpg.de>
Sender: news@mpifr-bonn.mpg.de
Nntp-Posting-Host: speckled
Organization: Max-Planck-Institut f"ur Radioastronomie
References: <lukner-1701951904400001@bingham.che.utexas.edu>
Date: Thu, 19 Jan 1995 01:10:19 GMT
Lines: 30
Xref: glinda.oz.cs.cmu.edu sci.image.processing:11987 comp.sys.ibm.pc.hardware.video:23111

In <lukner-1701951904400001@bingham.che.utexas.edu> lukner@che.utexas.edu (Ralf B. Lukner) writes:

>can capture 1Kx0.5K pixel video images (8bit data) at 30 fps directly to a
>PC based acquisition system.

That's 15MByte/s.

>disk.  From the preliminary information I have, the bottleneck is the
>permanent storage (e.g., hard disk).  A Seagate ST-1250W 2.1 GB drive has
>an internal transfer rate of 4.5 to 7.1 MB/s, with the SCSI-2 fast wide
>controller limited to 20 MB/s.

You can try RAID arrays and/or multiple SCSI controllers.

>rates, the CPU should carrying out "real-time" data compression before
>sending the data to permanent storage, and more than one hard disk may be
>needed.

Difficult. Images don't compress nicely unless you start throwing away
some data.

We use 4 Exabyte streamers in parallel to record 12bit data. But that's
just 512x512 at 4fps.

Regards,
-- 
                                Michael van Elst

Internet: mlelstv@mpifr-bonn.mpg.de                mlelstv@serpens.rhein.de
                                "A potential Snark may lurk in every tree."
