NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/comp/sys/mac/hardware/24879

Text File | 1992-12-24 | 9.3 KB | 220 lines

Newsgroups: comp.sys.mac.hardware
Path: sparky!uunet!gatech!concert!sas!mozart.unx.sas.com!saswlb
From: [e-mail] (Bill Brideson)
Subject: Summary: 24-bit Accelerated 1024x768 Graphics Cards (Long!)
Originator: [e-mail]
Sender: [e-mail] (Noter of Newsworthy Events)
Message-ID: <[e-mail]>
Date: Fri, 25 Dec 1992 00:29:06 GMT
Nntp-Posting-Host: spruce.unx.sas.com
Organization: SAS Institute Inc.
Lines: 207


   Thanks!!!  I'm overwhelmed!  I suppose some of you wondered why I didn't post
a summary until now.  I appologize for the delay, but it took me awhile to put
all the info into a "summary" form.  I got mail from these fine people:

Peter Glaskowsky          [e-mail]
                          [e-mail]
Derrick C. Mancini        [e-mail]
Gavin  Kistner            [e-mail]
Bret McCarty              [e-mail]
Martin Charest            [e-mail]
Kevin D. Baranski-Walker  [e-mail]
Kelvin Delbarre           [e-mail]
Doseok Kim                [e-mail]
Paul Campbell             taniwha![e-mail]


   In the summary below, I haven't attributed individual authors.  I hope this
doesn't affend any of them, but it seemed the most expedient way to resolve the
problem of multiple sources for single pieces of information and it allowed me
to recombine statements (keeping the original meaning intact, I hope) more
easily.  Of course, this information is based on peoples' experiences filtered
through *several* mind<->text translations.  Therefore, don't put more stock
into it than it is worth; *always* verify your understanding of a board's
characteristics with its manufacturer.  This summary comes with no warranty of
*any* kind!  Appologies in advance for any errors; certainly any mistakes are
inadvertent and no offense of any kind is intended.



Color Calibration:

   A couple of respondents confirmed my suspicion that this capability is
required for certain professions, but the rest of us don't generally require
color calibration that's as precise as what the various companies' calibrators
can accomplish.

   Another respondent reminded me that Apple is going to be releasing more
color-matching functionality in the operating system in QuickDraw GX and in
some other bits and pieces, so he encouraged me to ask the various manufacturers
about their plans for compatibility.  I haven't done that yet, but I will.



Other boards that I had not considered:

   Two respondents had considered and rejected the Futura MX board from E-
machines.  It supports resolutions up to 1024x768 and allows you to switch
between any two of those.  It has hardware pan and zoom, a 5-year warranty,
and a money-back guarantee.

   It was not thought to be particularly speedy; one supposition was that it
only accelerates scrolling.

   One person who bought this board said that to change resolutions, one must
not only restart the computer but also open the computer and change a switch on
the Futura board.  They seemed to find this to be a nuisance.

   Several other respondents pointed out that the Radius PrecisionColor 24XP
supports monitors up to 16 inches (832x624) and costs less than $500.  For this
investigation I need to consider monitor sizes up to 19 inches, but a 24XP and
a 16-inch monitor could be a very cost-effective way to go.



General stuff:


   One respondent mentioned that he was surprised how important being able to
change resolutions on-the-fly has become to him.  Several others said that this
ability is very useful.

   Likewise, those who have gotten used to "virtual desktop" really like it.
The SuperMac and RasterOps boards, when not using 24-bit mode, use the "extra"
VRAM to create a "virtual desktop".  The amount of VRAM available for the
virtual desktop depends on what bit depth you're using (the virtual desktop is
bigger when you're using 1-bit depth than it is when you're using 4-bit depth).



SuperMac Spectrum/24 Series III:

   (BTW, I heard from two developers at SuperMac, but not from anyone who
claimed to be affiliated with any of the other manufacturers.)

   Power consumption is 18.75 watts.

   Warranty period is 5 years.

   FCC certification is Class A.

   Acceleration is provided by an ASIC.  Acceleration is only active in 8- and
24-bit modes.  It can be turned on and off on the fly.

   This board does *not* support 16 bits per pixel (it *does* support 1, 2, 4,
8, and 24 bits per pixel).

   Resolution can be changed on-the-fly.

   This board supports hardware pan and zoom and "virtual desktop".

   People seem to be using it under 7.0.1 tuned without problems.

   This seems to be thought to be the fastest of the four boards.  MacUser
magazine did some video board comparisons in their May 1992 issue, and overall
they rated the Spectrum/24 Series III neck-in-neck with (possibly a shade faster
than) the Radius PrecisionColor 24X (the PrecisionColor 24XK was not included in
that test).  The Spectrum/24 Series III was the slowest of the three SuperMac
boards they tested; the Thunder/24 took first place and the Spectrum/24 PDQ Plus
took second place.  (Those two are out of this investigation's budget, however.)

   I was told that MacWorld and MacWeek did similar comparisons to MacUser and
reached similar conclusions, but I was unable to find those articles.

   Someone said they only got one piece of software with the board but others
said specifically that two programs are supplied.  At any rate, the display
management software (cdev) *is* provided; the disparity seems to be whether the
SuperMatch software comes with the board.

   The SuperMatch software is the same software distributed with the SuperMatch
color calibrator (which includes a suction cup doohickey that you put on the
screen and is sold separately).  It lets you compensate for the basic color
characteristics of many monitors.  (The May 1992 MacUser article has a really
nice table that lists which monitors are compatible with which video boards,
though they don't attempt to investigate which monitors are compatible with
which other companies' *calibrators*.)

   One respondent (who otherwise loved the Series III) said, "SuperMatch doesn't
do much for me.  It's a cdev which is supposed to match your display colors to
your scanned colors to your printed colors, not by tweaking, but by specifying
the device.  I don't have a scanner, so I don't know how well it works on that
end, but it really doesn't work at all with the Phaser IIPxi printer I have.
(The printer's blues are inevitably much darker than the screen.)  The only cdev
which comes close to correcting this is the one tailor-made to the printer (came
with it)."



Radius PrecisionColor 24XK:

   Power consumption is 1.75A at 5V plus 50mA at 12V plus 10ma at -12V, for
11.22 watts total.

   Warranty period is 1 year.

   FCC Certification is Class B, which I understand to be more stringent than
Class A (a Class B device is permitted to emit less electromagnetic radiation
than a Class A device).  This is the only board under consideration that
complies with Class B; the rest are all Class A.

   Acceleration is provided by an ASIC.  Acceleration is active for 8-, 16-, and
24-bit depths.  (1- and 4-bit depths are supported, but without acceleration.)
Acceleration can be turned on or off on-the-fly via a control panel.

   Resolution can be changed on-the-fly.

   One respondent said that this board does *not* have hardware pan and zoom or
"virtual desktop".

   Acceleration is active for 8-, 16-, and 24-bit depths.  1- and 4-bit depths
are supported, though without acceleration.  Acceleration is controlled via a
control panel.

   Speed-wise, this board was thought to be second to the Spectrum/24 Series
III (see MacUser article info, above).

   The "RadiusWare" suite of software was said to be "really well done" by
someone who says they need to change resolutions "fairly often" in their work.

   This is apparently the only board that will drive the Radius Color Pivot in
both portrait and landscape modes with switch-on-the-fly.



RasterOps Paintboard/Li:

   Power consumption is 3A at 5V, for 15 watts.

   Warranty period is 3 years.

   FCC certification is Class A.

   Acceleration is provided by an ASIC.  Acceleration can be turned on or off
on-the-fly.

   This board does *not* allow resolutions to be changed on-the-fly (you change
the resolution and then restart your Mac).

   This board has "virtual desktop" and hardware pan and zoom.

   I didn't get any "speed" data on this board, except that there are users who
"like" it.  In the MacUser tests, the RasterOpx 24XLi was right behind the
Radius PrecisionColor 24X, but I have no idea how those results extrapolate to
comparing the RasterOps Paintboard/Li to the Radius PrecisionColor 24XK.

   One respondent (who likes the board) reported a problem he's still investig-
ating in which running After Dark appears to cause the acceleration on this
board to interfere with the serial port (presumably by tying up the bus
somehow); also, in 16-bit mode highlighting doesn't display for this person.

   This board is compatible with RasterOps' CorrectColor Calibrator system.

--
Bill Brideson, [e-mail], 919-677-8000
SAS Institute, Inc.,  SAS Campus Drive,  Cary,  NC  27513
"There are people who think that everything one does with
a serious face is sensible." (18th cent.)