🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.multimedia
Article #1793

Re: Video Capture framerate results on 660AV

#1793
From: Lawrence D'Oliveiro, Waikato Uni...
Date: Mon, 01 Nov 1993 02:28
71 lines
4619 bytes
In article <2aq8qo$3d7@info.curtin.edu.au>, Martin Hill
<M.B.Hill@info.curtin.edu.au> posts some figures for video capture
rates using the built-in hardware on a 660AV.

> Capture to RAM,  no compression,         160 x 120  = 25.35 fps
>
> Capture to RAM,  no compression,         320 x 240  = 24.38 fps
>
> Capture to disk, no compression,         160 x 120  = 18.94 fps
>
> Capture to RAM,  Apple video compressor, 160 x 120  = 11.31 fps
>
> Capture to disk, Apple video compressor, 160 x 120  = 10.54 fps
>
> Capture to disk, no compression,         320 x 240  =  6.83 fps

Were these the actual capture resolutions? (i e not 384*288 or 192*144,
which would be more typical for PAL). I ask because the difference in speed
can be quite dramatic for a VideoSpigot. Perhaps this doesn't apply so much
to the AV Macs, because of the built-in hardware for scaling and so on.

For comparison with the above, I can get about 24 fps direct to disk with the
VideoSpigot in my Centris 650 at 192*144 resolution. At 384*288 resolution,
I can get about 12.5 fps. I did these tests back when I had only 8MB of RAM;
at that time, I found very little difference in performance between captures
to RAM and captures to disk.

Of course, I should point out that the AV hardware is capable of rather better
video quality than the Spigot. The Spigot seems best suited to resolutions
around 160*120 or 192*144; at anything much larger, you notice a certain amount
of jagginess.

> By the looks of things if you have heaps of RAM or heaps of Hard
> disk space, it's quite viable, but as soon as you try to compress on
> the fly you run into problems.  Mind you, the VideoSpigot
> post-compresses as well, so it's probably fairly comparable.

The Spigot's ScreenPlay utility captures video in YUV format (straight off
the card), not RGB. Do you recall in the options for capture on the AV whether
there was a choice of YUV versus RGB format as well? I was reading the
developer notes for the AV machines over the weekend, and I recall some mention
of support for YUV, which might be faster than RGB capture.

> A major reason you'd buy an AV for video work is that it's video
> still digitising is very good and it's video overlaying and output
> with convolution is superb.  QT movie capture will be superb if/when
> (!!) someone writes a QT compressor that utilizes the DSP.  In that
> scenario, without extra h/w, you'd be getting a tremendous boost to
> digitising and playback performance.  Mind you, the fact that the
> PowerPCs are supposedly not going to include a DSP onboard would
> seem to reduce the attraction of producing products to take
> advantage of the aforesaid chip.

As I see it, the first generation of PowerPCs is going to appeal mainly to
those with a need for raw number-crunching power. I read a report which said
that Apple's efforts to get developer support for native PowerPC applications
has been having mixed results. I'm skeptical that they're going to meet their
target of selling a million machines in the first year.

As far as I'm concerned, the AV machines are definitely the sexiest bits of
hardware Apple has come up with to date. No question, this is the architecture
to support today! And hopefully there'll be a generation of "PowerAV" machines,
which combine the AV architecture with a PowerPC CPU--now _that_ would be
something...

Lawrence D'Oliveiro                       fone: +64-7-856-2889
Computer Services Dept                     fax: +64-7-838-4066
University of Waikato            electric mail: ldo@waikato.ac.nz
Hamilton, New Zealand    37^ 47' 26" S, 175^ 19' 7" E, GMT+13:00
No one ever got cut up into tiny pieces and fed to the alligators for
buying Microsoft.
Technical Details
Message-ID: <1993Nov1.152825.21529@waikato.ac.nz>
Path: arc1-ca.pugleaf.net!gmd.de!xlink.net!sol.ctr.columbia.edu!howland.reston.ans.net!wupost!waikato!waikato.ac.nz!ldo
References: <2aq8qo$3d7@info.curtin.edu.au>

Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: f569d,4b907726f65aef3,start
X-Google-Attributes: gidf569d,public
X-Google-Thread: 10f675,4b907726f65aef3,start
X-Google-Attributes: gid10f675,public
X-Google-Thread: 11002a,4b907726f65aef3,start
X-Google-Attributes: gid11002a,public
X-Google-ArrivalTime: 1993-10-31 19:18:35 PST
Path: arc1-ca.pugleaf.net!gmd.de!xlink.net!sol.ctr.columbia.edu!howland.reston.ans.net!wupost!waikato!waikato.ac.nz!ldo
From: ldo@waikato.ac.nz (Lawrence D'Oliveiro, Waikato University)
Newsgroups: comp.multimedia,aus.mac,comp.sys.mac.hardware
Subject: Re: Video Capture framerate results on 660AV
Message-ID: <1993Nov1.152825.21529@waikato.ac.nz>
Date: 1 Nov 93 15:28:25 +1300
References: <2aq8qo$3d7@info.curtin.edu.au>
Distribution: world
Organization: University of Waikato, Hamilton, New Zealand
Lines: 71