3D Graphics Card Roundup teaser...
Loyd Case
Hi, folks....
I just completed testing for the first half of Computer Gaming World's
two-part graphics card roundup. The two parts will run in the November and
December issues. In addition to the usual set of benchmarks (Winbench 96
graphics Winmark, CBench, Duke Nukem 3D and Quake), we also used Direct3D
Test (which comes with the DirectX 2.0 SDK) and the built-in frame rate
counter in Hellbender.
Cards which will appear in the November issue (shipping in late October)
include:
S3 ViRGE cards:
Diamond Stealth 3D 2000 4MB
Hercules Terminator 3D 4MB
Number Nine Reality 332/fx (2 MB)
Matrox Mystique
ATI 3D Xpression PC2TV (using the Rage/2 chip)
3D/fx cards:
Orchid Righteous 3D
Diamond Monster 3D
In addition to the 3D cards, we look at the 3 main ET6000-based cards, the
Hercules Dynamite/128, the STB Lightspeed 128 and the Videologic Grafixstar
600.
For the December issue, it looks like we'll have:
3D Blaster PCI
Several ViRGE/VX cards
ATI 3D Pro Turbo
Videologic PowerVR 3D accelerator
And maybe (keep your fingers crossed) one or two Permedia cards.
Best regards,
Loyd Case
Contributing Editor, Technology
Computer Gaming World
Green K A
Loyd Case (lo...@proaxis.com) wrote:
:
: I just completed testing for the first half of Computer Gaming World's
: two-part graphics card roundup. The two parts will run in the November and
: December issues.
.... teaser material deleted ....
Damn ... that sure is teaser material all right !!!
I want to know the results/specs now !!!!!!
Come'on post a quick peek at the results ....
I promise that only I'll look at it .. and no-one else will ..
.. 8-)
Damn ... I can't wait for the issue to be released ..
But they take so damn long to get to Australia !!!!!
How are u getting your hands on the cards now ???
Are they really ready to be released .. all did you
review pre-production cards ?????
Any insights would be great !!!
Thanks.
Kevin Green.
John Hamilton
In article <01bb9ed2$1827c180$9c0744c6@daddy>
"Loyd Case" <lo...@proaxis.com> wrote:
> I just completed testing for the first half of Computer Gaming World's
> two-part graphics card roundup. The two parts will run in the November and
> December issues. In addition to the usual set of benchmarks (Winbench 96
> graphics Winmark, CBench, Duke Nukem 3D and Quake), we also used Direct3D
> Test (which comes with the DirectX 2.0 SDK) and the built-in frame rate
> counter in Hellbender.
Way to go, Loyd! I get very anxious when it comes time for my CGW to arrive
each month.
This subject is absolutely THE hot subject of gaming right now, and if you
through MMX into the mix sometime in mid-97, graphics and multimedia hardware
and software absolutely determine the near future of gaming on personal computers.
Thanks for a great 'zine!
john
-------------------------------------------------------------------------------------------------
Pursuit Ballistics, Inc.
jo...@pball.com
If the Blues don't kill you, Brother,
They'll make you a mighty, mighty man.
Walter Benzing
Hi,
Loyd Case wrote:
>Hi, folks....
>I just completed testing for the first half of Computer Gaming World's
>two-part graphics card roundup.Cards which will appear in the November issue
>(shipping in late October)
>include:
>S3 ViRGE cards:
>Diamond Stealth 3D 2000 4MB
>Hercules Terminator 3D 4MB
>Number Nine Reality 332/fx (2 MB)
>Matrox Mystique
>ATI 3D Xpression PC2TV (using the Rage/2 chip)
ReHi,
I don't want to spoil the mood of those guys thinking of buying
a 3D upgrade for their computer, but some of these cards were also covered
in the october issue of the PC-Player (a game mag here in Germany)
and the result was a disaster.
To sum it up:
-none of the packed in dos games are faster than it's standard equivalent,
some looking better though.
-only two video cards actually did hardware 3d speed up with directX/W95:
ATI 3D Xpression (2 MB)
Elsa Victory 3D (S3 Virge) with 4 MB Ram
as there were other 2 MB S3 cards in the test (including a 2 MB Victory) ,
the tester guessed that hardware speedup is 4 MB only with S3/Virge
-Both cards that did do hardware 3D speed up with directX were not able
to run the directX test game (hellbender) correctly in this mode =>
switch back to software only
-the fastest directX "3D-card" according to their test was a 2D card
(Elsa Winner 1000) and this includes both matrox cards (millenium and
mystique) but all of them were mostly the same, (obviously because
none of the cards actually did HARDWARE speedup)
-the tests were done with drivers from the hardware companies AND the
drivers delivered with directX by MS
Quite obviously they were so "impressed" by this performance, they found
it necessary to make a frontpage editorial out of it.
Bye,
Walter Benzing
ben...@iegi01.etec.uni-karlsruhe.de
P.S.:
Just to cheer those guys up that got a "3D-card", it was also said that all
cards were fairly good at dos-speed, with the mystique being a tick faster than
the millenium.
P.P.S:
I'm not affected with PC-Player other than this being my favorite game mag
because it is more critical than others.
Loyd Case
My experience was that the S3 ViRGE cards actually did okay in the final
shipping version of Hellbender... not great, mind you, but okay. Also,
with Descent 2, the ViRGE version runs in 16-bit color, with filtered
textures, looks better and will run at roughly the same frame rate (plus or
minus a couple) as the unaccelerated version. To me, this is a performance
improvement.
Also, all the cards ran the shipping version of Hellbender with
acceleration on.
None of the multipurpose cards I've seen so far do great 3D... but if
you're slot-limited and on a budget, it's like getting limited 3D for free.
Now, the 3Dfx cards, on the other hand, are, as Darth Vader might say,
"most impressive".
Next month we'll cover Rendition, PowerVR, ViRGE/VX and maybe, just maybe,
a Permedia card or two.
Cheers,
Loyd
Walter Benzing <ben...@iegi01.etec.uni-karlsruhe.de> wrote in article
<513vrb$q...@nz12.rz.uni-karlsruhe.de>...
GEORGE ARUGAY MONTEMAYOR
: I just completed testing for the first half of Computer Gaming World's
: two-part graphics card roundup. The two parts will run in the November and
: December issues. In addition to the usual set of benchmarks (Winbench 96
: graphics Winmark, CBench, Duke Nukem 3D and Quake), we also used Direct3D
: Test (which comes with the DirectX 2.0 SDK) and the built-in frame rate
: counter in Hellbender.
[stuff deleted]
Very interesting. But you know what will make the next two magazines more
interesting? Video clips of games of all the 3D cards you tested
featured in that issue. Of course, this would be available only on the
CD+magazine bundle.
Screen shots and specs make a man drool, but its the .mov/.avi (in a
window size of at least 320x240, not some small 160x120) that will make a
man foam. :)
-george
Richard Fecteau - Reflection X
Walter Benzing wrote:
>
> Hi,
>
> Loyd Case wrote:
>
> >Hi, folks....
>
> >I just completed testing for the first half of Computer Gaming World's
> >(shipping in late October)
> >include:
>
> >S3 ViRGE cards:
>
> >Diamond Stealth 3D 2000 4MB
> >Hercules Terminator 3D 4MB
> >Number Nine Reality 332/fx (2 MB)
>
>
> >ATI 3D Xpression PC2TV (using the Rage/2 chip)
>
> ReHi,
>
> I don't want to spoil the mood of those guys thinking of buying
> a 3D upgrade for their computer, but some of these cards were also covered
> in the october issue of the PC-Player (a game mag here in Germany)
> and the result was a disaster.
> To sum it up:
>
> -none of the packed in dos games are faster than it's standard equivalent,
> some looking better though.
> -only two video cards actually did hardware 3d speed up with directX/W95:
>
> ATI 3D Xpression (2 MB)
> Elsa Victory 3D (S3 Virge) with 4 MB Ram
> P.S.:
> Just to cheer those guys up that got a "3D-card", it was also said that all
> cards were fairly good at dos-speed, with the mystique being a tick faster than
> the millenium.
>
> P.P.S:
> I'm not affected with PC-Player other than this being my favorite game mag
> because it is more critical than others.
No surprise here. We'll have to wait for Verite or Voodoo-based cards to
get a real 3D speedup. My money is waiting...
Rich
ALX
In article <01bb9ed2$1827c180$9c0744c6@daddy>, "Loyd Case" <lo...@proaxis.com>
wrote:
> I just completed testing for the first half of Computer Gaming World's
> two-part graphics card roundup. The two parts will run in the November and
> December issues. In addition to the usual set of benchmarks (Winbench 96
> graphics Winmark, CBench, Duke Nukem 3D and Quake), we also used Direct3D
> Test (which comes with the DirectX 2.0 SDK) and the built-in frame rate
> counter in Hellbender.
Thanks for the notice. I'll really be looking forward to seeing
the results on the various boards, especially the so-called
2nd generation boards (PowerVR, Voodoo, Rage II).
BTW, the PowerVR and Mystique don't support bilinear filtering.
Did you perform the DX2 SDK tests having filtering enabled? I'd
like to see just how much of an impact filtering has on the
performance of a board that doesn't support it in hardware.
->ALX<-
Loyd Case
I turned all the features on in Hellbender, though there was no explicit
filtering setting. Direct 3D sometimes tries to be too smart, although I
would occasionally get a message. For example, when I ran the Mystique, I
got a message from Hellbender stating "this accelerator does not support
fog. You can play the game, but fog effects will be turned off". That was
really too bad, because it probably skewed the results somewhat.
On the flip side, we ran D3D test at two different reality settings: one
with bilinear, perspective correction, gouraud shading and other features
on, the other with flat shading, bilinear off, etc.
Best regards,
Loyd Case
ALX <unk...@account.com> wrote in article
<513s5d$g...@news01.deltanet.com>...
ALX
In article <513056$l...@seagoon.newcastle.edu.au>, c910...@alinga.newcastle.edu.
au (Green K A) wrote:
> How are u getting your hands on the cards now ???
> Are they really ready to be released .. all did you
> review pre-production cards ?????
Many of the cards are probably ready to ship if not for the
wait on software. Orchid's in this exact position. Their
cards are ready, but the software (in particular Quake),
is holding it up.
->ALX<-
Dhiren Shah
gmo...@sfsu.edu (GEORGE ARUGAY MONTEMAYOR) wrote:
>Loyd Case (lo...@proaxis.com) wrote:
>: Hi, folks....
>: I just completed testing for the first half of Computer Gaming World's
>: two-part graphics card roundup. The two parts will run in the November and
>: December issues. In addition to the usual set of benchmarks (Winbench 96
>: graphics Winmark, CBench, Duke Nukem 3D and Quake), we also used Direct3D
>: Test (which comes with the DirectX 2.0 SDK) and the built-in frame rate
>: counter in Hellbender.
>[stuff deleted]
>Very interesting. But you know what will make the next two magazines more
>interesting? Video clips of games of all the 3D cards you tested
>featured in that issue. Of course, this would be available only on the
>CD+magazine bundle.
>Screen shots and specs make a man drool, but its the .mov/.avi (in a
>window size of at least 320x240, not some small 160x120) that will make a
>man foam. :)
>-george
I agree...we really need to see how smooth and how good these
different cards look...not those crappy videos they throw onto cds
these days. Do us a favor and leave off the America Online and
include a couple of videos of the 3d cards. At least include full
640x480x24bit still shots of the 3d games.
Jaggernaut
ben...@iegi01.etec.uni-karlsruhe.de (Walter Benzing) wrote:
> ATI 3D Xpression (2 MB)
> Elsa Victory 3D (S3 Virge) with 4 MB Ram
> as there were other 2 MB S3 cards in the test (including a 2 MB Victory) ,
> the tester guessed that hardware speedup is 4 MB only with S3/Virge
BS, he guessed wrong. Hardware acceleration in direct3d works fine on
the 2mb Diamond Stealth 3D 2000. Obviously these tests where run
before the correct drivers where out. As for the speed, well...keep
'em coming cos they aint there yet. Although you have to realise that
when running say Monster Truck on a 3D card it's running in 16bit with
filtering and texture smoothing at about the same speed as a 2d card
in 8bit with no special effects.
--
Jagg
Vancouver, Canada
*********************************************
* WB Handle:Term <Terminator> *
* 401st RCAF <Rams> *
* http://bc1.com/users/prophet/ramsquad.html*
*********************************************
Dave Glue
Loyd Case <lo...@proaxis.com> wrote in article
<01bb9f61$a8b68a80$cc8da3ce@daddy>...
> I turned all the features on in Hellbender, though there was no explicit
> filtering setting. Direct 3D sometimes tries to be too smart, although I
> would occasionally get a message. For example, when I ran the Mystique,
I
> got a message from Hellbender stating "this accelerator does not support
> fog. You can play the game, but fog effects will be turned off". That
was
> really too bad, because it probably skewed the results somewhat.
But nothing out of the ordinary, I don't believe the Mystique does support
Fog- it has "screen door" transparency, which is just using spaced pixels-
ugh. Dunno, Matrox seems to be having trouble keeping up in the 3D
accelerator market- no filtering, no fog/transparency, mediocre
performance- I don't see how the Mystique is going to have that much of a
chance against a 4-meg $200 PCI Blaster.
Loyd Case
After the Millenium, which was a pretty hot card in its day (and still
virtually unbeatable in high color 2D), I had real hopes for the Mystique.
At the very least, they should have included bilinear filtering. Oh, well.
As ever,
Loyd Case
Dave Glue <dav...@interlog.com> wrote in article
<01bb9ff0$c1fa62c0$dc98d4c7@Pdaveacg>...
Ken DiStefano
Loyd Case wrote:
>
> Hi, folks....
>
> I just completed testing for the first half of Computer Gaming World's
> two-part graphics card roundup. The two parts will run in the November and
> December issues. In addition to the usual set of benchmarks (Winbench 96
> graphics Winmark, CBench, Duke Nukem 3D and Quake), we also used Direct3D
> Test (which comes with the DirectX 2.0 SDK) and the built-in frame rate
> counter in Hellbender.
>
> include:
>
> S3 ViRGE cards:
>
> Diamond Stealth 3D 2000 4MB
> Hercules Terminator 3D 4MB
> Number Nine Reality 332/fx (2 MB)
>
> Matrox Mystique
>
> ATI 3D Xpression PC2TV (using the Rage/2 chip)
>
>
> Orchid Righteous 3D
> Diamond Monster 3D
>
> In addition to the 3D cards, we look at the 3 main ET6000-based cards, the
> Hercules Dynamite/128, the STB Lightspeed 128 and the Videologic Grafixstar
> 600.
>
> For the December issue, it looks like we'll have:
>
> 3D Blaster PCI
> Several ViRGE/VX cards
> ATI 3D Pro Turbo
> Videologic PowerVR 3D accelerator
> And maybe (keep your fingers crossed) one or two Permedia cards.
>
> Best regards,
>
> Loyd Case
> Contributing Editor, Technology
> Computer Gaming World
I don't need your test results. It's easy to figure out that any card
based on 3dfx's Voodoo chipset will perform the best when it comes to 3D
game acceleration.
Ken
Jeremy
>"Loyd Case" <lo...@proaxis.com> wrote:
<big snip>
Did you by any chance try out a Diamond Stealth 3D (or any Virge card)
coupled to a Diamond Monster 3D (or any Voodoo card) ? If so did you
notice any problems when running Direct 3D games ? e.g. did the game
correctly use the fastest 3D accelerator (the Voodoo and not the
Virge) ? Any other problems ? Orchid claim that the R3D and the
Stealth 3D are compatible, but that doesn't neccesarily mean that the
two run optimally together.
If you didn't do this, perhaps you could try it out before the
December issue and publish the results, or better still post them
here?
I have a Stealth 3D 2000 and I am planning on getting a Voodoo based
card also, but not if there are too many problems. I still have my old
Stealth 64 Video 2Mb VRAM, so if the Stealth 3D clashes in some way
with the Voodoo based card I'm tempted to trash the Stealth 3D and use
my Stealth 64 with the Voodoo.
Thanks
Jeremy
John Hamilton
In article <32363E...@lehigh.edu>
Ken DiStefano <kl...@lehigh.edu> wrote:
[in reference to Mr. Case's post]
> I don't need your test results. It's easy to figure out that any card
> based on 3dfx's Voodoo chipset will perform the best when it comes to 3D
> game acceleration.
Sure, and we know that each implementation using the Voodoo will be
optimal and well engineered, without any bugs, and not troublesome to
use. Just like we know every new microprocessor from Intel will be
perfect. Not. ;-)
Greg Cisko
In article <01bb9ed2$1827c180$9c0744c6@daddy>, "Loyd Case" <lo...@proaxis.com> writes:
>Hi, folks....
>
>I just completed testing for the first half of Computer Gaming World's
>two-part graphics card roundup. The two parts will run in the November and
>December issues. In addition to the usual set of benchmarks (Winbench 96
>graphics Winmark, CBench, Duke Nukem 3D and Quake), we also used Direct3D
>Test (which comes with the DirectX 2.0 SDK) and the built-in frame rate
>counter in Hellbender.
Thanks for the heads up I'm glad my subscription is paid up :-) What I would
like to know, is if the 3D cards will improve the performance of normal games?
Or does the game need to be re-optomized for 3D cards? I am mostly curious
about whether the 3D cards will improve the performance compared to a fast PCI
card. ie... If I swap to a 3D card will all games improve, or just games written
for 3D cards? I assumed the latter, but some make it sound like the former. Can
you shed some light?
Loyd Case
I haven't tried this combo yet. However, I had an extensive conversation
with the product manager from Orchid, and they have tested with a variety
of different cards, including the Millenium and S3-based cards, with very
good luck. I don't see anything special about the Monster 3D that will
prevent it from running with one of those cards.
Having said that, your idea is a good one, and I'll check it out for the
December issue. In theory, Direct3D shouldn't be a problem, since it will
autodetect the highest capability card and use it. In theory...:)
However, I can see why someone would want a combo like this to take
advantage of games written specifically for one chip or another.
Best regards,
Loyd Case
Loyd Case
Several points:
Those folks who can't afford two cards, or are unwilling to spend $300 for
an add-on will need the information.
While the 3Dfx chips are quite nice, people looking for a more general
purpose solution want to know a bit more about Rendition and Permedia.
Finally, I know several people who have used up all their slots; they can
replace a card, but not add another one.
Best regards,
Loyd Case
Ken DiStefano <kl...@lehigh.edu> wrote in article
<32363E...@lehigh.edu>...
> Loyd Case wrote:
>
> I don't need your test results. It's easy to figure out that any card
> based on 3dfx's Voodoo chipset will perform the best when it comes to 3D
> game acceleration.
>
> Ken
>
Jaggernaut
jez...@frimley.demon.co.uk (Jeremy) wrote:
>Did you by any chance try out a Diamond Stealth 3D (or any Virge card)
>coupled to a Diamond Monster 3D (or any Voodoo card) ? If so did you
>notice any problems when running Direct 3D games ? e.g. did the game
>correctly use the fastest 3D accelerator (the Voodoo and not the
>Virge) ? Any other problems ? Orchid claim that the R3D and the
>Stealth 3D are compatible, but that doesn't neccesarily mean that the
>two run optimally together.
I'm very interested in this combo too. I want to keep the Diamond 3d
2000 but not if they have *any* conflicts.
[Marc Warden]
Hi ALX.
While it is true that software is holding up the release of the Righteous
3D, it is only a rumor that it is Quake.
The Orchid rep who mentioned Quake in a note, a copy which was posted on
the newsgroup, had a question mark after Quake, intending to indicate that
Quake was possibly being considered for bundling with the board, nothing
more.
I can neither confirm or deny the rumor. I would only like to point out
that until the software titles are "locked down" and the agreements signed,
nothing is certain. I think only those titles for which agreements have
been signed have been announced as bundled titles.
Sincerely,
Marc Warden, Orchid Technology (ma...@micronics.com)
Bruce Williams
On 10 Sep 1996 06:00:06 GMT, c910...@alinga.newcastle.edu.au (Green K
A) wrote:
> Damn ... I can't wait for the issue to be released ..
> But they take so damn long to get to Australia !!!!!
Perhaps they'll post the results to their web site after the magazine
goes out. Then you might be able to get the results in advance of the
magazine's arrival in AU.
-Bruce
Bob Johnston
In article <01bb9ed2$1827c180$9c0744c6@daddy>, "Loyd Case" <lo...@proaxis.com> wrote:
>In addition to the 3D cards, we look at the 3 main ET6000-based cards, the
>Hercules Dynamite/128, the STB Lightspeed 128 and the Videologic Grafixstar
>600.
>
>Loyd Case
>Contributing Editor, Technology
>Computer Gaming World
I wonder how does a great 2D DirectDraw board like the ET6000 perform's under
software RAMP and RGB HEL mdoes compared to other boards that have "real D3D
HAL" modes.
Bob J
Bruce Williams
On 10 Sep 1996 12:03:16 GMT, jo...@pball.com (John Hamilton) wrote:
>Way to go, Loyd! I get very anxious when it comes time for my CGW to arrive
>each month.
>
...
>
>Thanks for a great 'zine!
I have to second this; I've been spending money on computer game
magazines for a good many years, but the one I keep coming back to is
CGW. As a concrete sign of this, its the only game magazine that I
subscribe to, because I trust them to deliver quality articles _every
month_.
And then, because I move around a lot and forget to send my change of
address forms, I tend to buy the issues at the newsstand as well.
Thanks for providing such great service to your readers, Lloyd!
Please feel free to share this praise with your fellow workers at CGW.
-Bruce
Ross E. Becker
ma...@micronics.com@, [Marc, Warden] wrote:
>
> While it is true that software is holding up the release of the Righteous
> 3D, it is only a rumor that it is Quake.
>
> The Orchid rep who mentioned Quake in a note, a copy which was posted on
> the newsgroup, had a question mark after Quake, intending to indicate that
> Quake was possibly being considered for bundling with the board, nothing
> more.
>
> I can neither confirm or deny the rumor. I would only like to point out
> that until the software titles are "locked down" and the agreements signed,
> nothing is certain. I think only those titles for which agreements have
> been signed have been announced as bundled titles.
>
> Sincerely,
>
> Marc Warden, Orchid Technology (ma...@micronics.com)
I would think, that at supposedly T-3 weeks to ship & counting, you
would
actually have your agreements signed, & know what you were or were not
going to ship with the card. Is this the case? Does the "end of
September"
release date look realistic at this point for the Righteous 3D? Also,
can you tell me where I will actually be able to purchase the card when
it is available? I have hunted around, and *most* companies put a
product
in their database as a "soon to be available" product, so they can start
telling customers about it, and after talking with numerous mail order
companies, CompUSA, MicroCenter, Egghead, etc. I have yet to find a
single company that even listed your product in their database. Even
going through computer shopper, I found only a half-dozen mail-order
companies which still list the Orchid farenheight card, and none of them
know of the Righteous product. I am just worried that this does not seem
like the product is actually going to ship at the end of this month, or
even like there will be reasonable distribution....
--Ross
--
===============================================================================
= Ross Becker * Lead Programmer/Analyst *
bec...@cis.ohio-state.edu =
= 793 Dreese Lab * User Environment
* =
= 2015 Neil Ave. * IICF Staff
* =
= (614) 292-7325 * Ohio State University
* =
===============================================================================
Ken DiStefano
John Hamilton wrote:
>
> In article <32363E...@lehigh.edu>
> > based on 3dfx's Voodoo chipset will perform the best when it comes to 3D
> > game acceleration.
>
> optimal and well engineered, without any bugs, and not troublesome to
> use. Just like we know every new microprocessor from Intel will be
> perfect. Not. ;-)
>
> john
>
There are only 2 cards that I know of and by the time this issue comes
out I'll have read which one is better right here. My point was, I
don't need some magazine telling me that Monster 3D kicks the Mystique
into the ground when anyone can plainly see that.
-Ken
enzo
Eric Pinnell wrote:
>
> In <01bb9f99$22856200$878da3ce@daddy>, "Loyd Case" <lo...@proaxis.com> writes:
> >After the Millenium, which was a pretty hot card in its day (and still
> >virtually unbeatable in high color 2D), I had real hopes for the Mystique.
> >At the very least, they should have included bilinear filtering. Oh, well.
> >
> >As ever,
> >
> >Loyd Case
> >
>
> Try the Lockheed Martin Real3D chipset. Hot stuff from what I've read.
The problem with Lockheed is that each year they say "Release in one
year from now each year. Apparently the current word is that Real3d
will not be available until December '97. Last year they said November
'96 - and in Jan '95 they said December '95.
I love the Real3d technology, but it seems that it will merely be
a mid-pack runner at a very high price point when it finally does
come out.
ALX
In article <517vhc$6...@amdint2.amd.com>, br...@dvorak.amd.com (Brent Burkholder)
wrote:
> Just as a warning for all of you folks out there considering Orchid,
> be advised that they have a history of not providing new drivers for
> hardware (soundwave 32 still doesn't have win 95 drivers.) After having
> been stung by them once before I don't think I'd ever buy one of their
> products again, no matter how good it looks.
This has been said before, but I suppose it bears repeating.
3DFX is producing the drivers. I suppose the situation is
similar to S3. Individual OEMs may produce their own flavor
or variant, but 3DFX Voodoo card owners can use 3DFX written
drivers.
->ALX<-
Stephen Wilkinson
In article <5176r9$j...@client2.news.psi.net>, bo...@voicenet.com (Bob Johnston) wrote:
[snip]
>I wonder how does a great 2D DirectDraw board like the ET6000 perform's under
>software RAMP and RGB HEL mdoes compared to other boards that have "real D3D
>HAL" modes.
This would be a benchmark of your CPU rather than your GFX board in
that your processor must perform the calculations needed to shade,
texture and draw your polygons rather than having the spiffy 3D
chip(s) on your GFX board do it instead. When there is no 3D hardware
on your GFX board, D3D will emulate the requested functionality in
software.
Cheers,
Stephen Wilkinson "Programming is like pinball.
Software Engineer the reward for doing it well
Interactive Creations Inc. is the opportunity to do it
wi...@airmail.net again." (anon - Wizards Bane)
RNDM in Warbirds
Eric Pinnell
In <01bb9f99$22856200$878da3ce@daddy>, "Loyd Case" <lo...@proaxis.com> writes:
>After the Millenium, which was a pretty hot card in its day (and still
>virtually unbeatable in high color 2D), I had real hopes for the Mystique.
>At the very least, they should have included bilinear filtering. Oh, well.
>
>As ever,
>
>Loyd Case
>
Loyd,
Try the Lockheed Martin Real3D chipset. Hot stuff from what I've read.
Eric Pinnell
Loyd Case
We're working hard on it, but consumer level boards are a ways out, from
what they tell us.
Cheers,
Loyd
Brent Burkholder
In article <01bba001$a6c10660$9d8da3ce@daddy>,
Loyd Case <lo...@proaxis.com> wrote:
>I haven't tried this combo yet. However, I had an extensive conversation
>with the product manager from Orchid, and they have tested with a variety
>of different cards, including the Millenium and S3-based cards, with very
>good luck.
Just as a warning for all of you folks out there considering Orchid,
be advised that they have a history of not providing new drivers for
hardware (soundwave 32 still doesn't have win 95 drivers.) After having
been stung by them once before I don't think I'd ever buy one of their
products again, no matter how good it looks.
--
--------------------------- ------------------------------------------------
Brent Burkholder < < "To coerce God is to coerce reality and answer
Brent.Bu...@amd.com > > your own prayers." - Diamond Mask, Julian May
-------------------------- -------------------------------------------------
Rick Collin
What about the Diamond Fire GL or other Glint chip boards? I know
they're expensive, but how do they perform with 3D accel? Anybody
have any experience or should I look to the CAD newsgroups?
Tom Ryan
Hi Loyd,
Did you use the profiling utiltiy in SciTech Display Doctor 5.3? It is
a very good indication of the raw bandwidth available in any particuar
graphics card.
Also, we may be doing a bit of a press tour soon in the Bay Area
showing some cool technology from us and our customers. Are you at the
editorial offices much or can we arrange a visit with you as well?
Let me know
-Tom Ryan, SciTech Software
"Loyd Case" <lo...@proaxis.com> wrote:
>Hi, folks....
>I just completed testing for the first half of Computer Gaming World's
>two-part graphics card roundup. The two parts will run in the November and
>December issues. In addition to the usual set of benchmarks (Winbench 96
>graphics Winmark, CBench, Duke Nukem 3D and Quake), we also used Direct3D
>Test (which comes with the DirectX 2.0 SDK) and the built-in frame rate
>counter in Hellbender.
>Cards which will appear in the November issue (shipping in late October)
>include:
>S3 ViRGE cards:
>Diamond Stealth 3D 2000 4MB
>Hercules Terminator 3D 4MB
>Number Nine Reality 332/fx (2 MB)
>Matrox Mystique
>ATI 3D Xpression PC2TV (using the Rage/2 chip)
>3D/fx cards:
>Orchid Righteous 3D
>Diamond Monster 3D
>In addition to the 3D cards, we look at the 3 main ET6000-based cards, the
>Hercules Dynamite/128, the STB Lightspeed 128 and the Videologic Grafixstar
>600.
>For the December issue, it looks like we'll have:
>3D Blaster PCI
>Several ViRGE/VX cards
>ATI 3D Pro Turbo
>Videologic PowerVR 3D accelerator
>And maybe (keep your fingers crossed) one or two Permedia cards.
>Best regards,
>Loyd Case
>Contributing Editor, Technology
>Computer Gaming World
+-----------------------------------------------------------------------+
| SciTech Software - Building Truly Plug'n'Play Software -- Today! |
+-----------------------------------------------------------------------+
| Tom Ryan, Director of Marketing | Ph/FAX: 916-894-8400/916-894-9069 |
| SciTech Software, Inc. | Email : To...@scitechsoft.com |
| 505 Wall Street | ftp : ftp.scitechsoft.com |
| Chico, CA 95928 | www : http://www.scitechsoft.com |
+-----------------------------------------------------------------------+
Pat
BEWARE of SciTech.
A lot of people have reported problems with V5.3.
This is especailly true for users of #9 Motion 771 w/Hawkeye drivers.
There is a fairly full folder of complaints on AOL with very
little in the way of support -- mostly snotty replies or lame
suggestions.
LawMAN
> I haven't tried this combo yet. However, I had an extensive conversation
> with the product manager from Orchid, and they have tested with a variety
> of different cards, including the Millenium and S3-based cards, with very
> good luck. I don't see anything special about the Monster 3D that will
> prevent it from running with one of those cards.
>
> Having said that, your idea is a good one, and I'll check it out for the
> December issue. In theory, Direct3D shouldn't be a problem, since it will
> autodetect the highest capability card and use it. In theory...:)
> However, I can see why someone would want a combo like this to take
> advantage of games written specifically for one chip or another.
>
I hope that when you review the PowerVR in the December issue, you could
specifically ask VideoLogic/NEC whether they are supporting bilinear
filtering in their 3D cards. If not, are they going to? ALX is right that
without bilinear filtering, it doth sucks...
Ciao.
--
.oOOOOOo.
LawMAN (lawrence....@nortel.com) \\ //
"When death beckons, silence covers us like a dark shroud." \\ //
Nortel Transport and Broadband Networks, Montreal, Canada. \o/
X
Bradley Dominik
Rick,
look at the June issue of PC Magazine. They looked at OpenGL accelerators,
and they rated the FireGL as one of the worst.
Brad
bsd...@psu.edu
Rick Collin <rick....@mail.wdn.com> wrote in article
<517t0h$3...@news2.cais.com>...
David Lau-Kee
ds...@monmouth.com (Dhiren Shah) wrote:
|gmo...@sfsu.edu (GEORGE ARUGAY MONTEMAYOR) wrote:
...
|>Screen shots and specs make a man drool, but its the .mov/.avi (in a
|>window size of at least 320x240, not some small 160x120) that will make a
|>man foam. :)
|
|>-george
|
|I agree...we really need to see how smooth and how good these
|different cards look...not those crappy videos they throw onto cds
|these days. Do us a favor and leave off the America Online and
|include a couple of videos of the 3d cards. At least include full
|640x480x24bit still shots of the 3d games.
We're releasing a game called Scorched Planet in October/November,
which supports software, Matrox Mystique, Verite boards and Voodoo
boards.
If there's interest, I'll put some screenshots from each of the
'versions' on our web site, so you can compare how it looks on
each of the chips mentioned. I'll also put some 2 level demos
up in a week or two, if people want.
movs and avis may not help too much because of the playback issues.
David
Scott Mathers
Pat wrote:
>
> BEWARE of SciTech.
>
> A lot of people have reported problems with V5.3.
> This is especailly true for users of #9 Motion 771 w/Hawkeye drivers.
> There is a fairly full folder of complaints on AOL with very
> little in the way of support -- mostly snotty replies or lame
> suggestions.
>
Yeah, I'd be really upset if I gave them my money. Version 5.2 made
my card go insane. I e-mailed them, basically saying I was interested
in buying it, not until they fixed it for my 771. Well 5.3 came out,
now it will turn my win 95 to a funky translucent shade of yellow.
Not so cool. Next card will have built-in VESA 2.0 support. Then
I can say screw those damn drivers.
What good 2D cards have VESA 2.0 support built-in?
desslock
The Matrox Millenium ain't too shabby.
The Matrox Mystique will also have VESA 2.0 built in, and unproven 3d
capabilities.
--
Trips
Scott Mathers wrote:
>
> Pat wrote:
> >
> > BEWARE of SciTech.
> >
> > A lot of people have reported problems with V5.3.
> > This is especailly true for users of #9 Motion 771 w/Hawkeye drivers.
> > There is a fairly full folder of complaints on AOL with very
> > little in the way of support -- mostly snotty replies or lame
> > suggestions.
> >
>
> Yeah, I'd be really upset if I gave them my money. Version 5.2 made
> my card go insane. I e-mailed them, basically saying I was interested
> in buying it, not until they fixed it for my 771. Well 5.3 came out,
> now it will turn my win 95 to a funky translucent shade of yellow.
> Not so cool. Next card will have built-in VESA 2.0 support. Then
> I can say screw those damn drivers.
>
> What good 2D cards have VESA 2.0 support built-in?
The STB Lightspeed 128 has VESA 2.0 built in, and it screams. It's fully
Win 95 and DirectX compatible right out of the box, and it beat every
card I tested it against in Chris' Bench, Quake timerefresh, Warbirds
Framerate, and Air Warrior frame rate.
Trips
Jean
> > In addition to the 3D cards, we look at the 3 main ET6000-based cards, the
> > Hercules Dynamite/128, the STB Lightspeed 128 and the Videologic Grafixstar
> > 600.
> >
>
have a P166 and the Xpression 3D. Guess what I did the exact same test.
I said to myself , geez these cards will make my Xpression 3D look like
crap but suprise I get higher framerate in all the tests (using vesa
driver from ATI)!!!! I couldn't believe it myself.... I didn't find an
answer yet.
Matt Thompson-Moltzen
(snip)
> What good 2D cards have VESA 2.0 support built-in?
I just bought myself a Hercules Dynamite 128/Video and strongly
recommend it. This mother rocks in DOS and WIN95. VESA 2.0 is built in
to the ROM. AH-64D Longbow just loves it!
Viking.
Glenn Dowling
In article <519gpk$h...@cruella.criterion.canon.co.uk>, lau...@csl.com says...
>
>We're releasing a game called Scorched Planet in October/November,
>which supports software, Matrox Mystique, Verite boards and Voodoo
>boards.
>
>If there's interest, I'll put some screenshots from each of the
>'versions' on our web site, so you can compare how it looks on
>each of the chips mentioned. I'll also put some 2 level demos
>up in a week or two, if people want.
Screenshots would be good
Framerates would be better
Giving us the URL would be the ultimate!!
<GaD>
Royce Liao
Scott Mathers (smat...@vivanet.com) wrote:
: Yeah, I'd be really upset if I gave them my money. Version 5.2 made
: my card go insane. I e-mailed them, basically saying I was interested
: in buying it, not until they fixed it for my 771. Well 5.3 came out,
: now it will turn my win 95 to a funky translucent shade of yellow.
: Not so cool. Next card will have built-in VESA 2.0 support. Then
: I can say screw those damn drivers.
Try out Dietmar's S3VBE20 freeware VBE 2.0 core. You can download it
from the S3 www-site (http://www.s3.com)
: What good 2D cards have VESA 2.0 support built-in?
Matrox MGA Millenium
Scott Pedersen
David Lau-Kee wrote:
>
> ds...@monmouth.com (Dhiren Shah) wrote:
>
> |gmo...@sfsu.edu (GEORGE ARUGAY MONTEMAYOR) wrote:
> ...
> |>Screen shots and specs make a man drool, but its the .mov/.avi (in a
> |>window size of at least 320x240, not some small 160x120) that will make a
> |>man foam. :)
> |
> |>-george
> |
> |I agree...we really need to see how smooth and how good these
> |different cards look...not those crappy videos they throw onto cds
> |these days. Do us a favor and leave off the America Online and
> |include a couple of videos of the 3d cards. At least include full
> |640x480x24bit still shots of the 3d games.
>
> which supports software, Matrox Mystique, Verite boards and Voodoo
> boards.
>
> If there's interest, I'll put some screenshots from each of the
> 'versions' on our web site, so you can compare how it looks on
> each of the chips mentioned. I'll also put some 2 level demos
> up in a week or two, if people want.
> David
Yep please! Actually i was checken' that game out (i think) in issue 36
of Edge, looks pretty darn impressive! David, if possible could you
also please post some specs on the performance of the game through the
various cards e.g. frame rates with what functions.(If figures are
availible that is)
PS: how will the game run on a Pentium 100 with 32 meg of Ram, and a
voodoo based card, in SVGA?
Thanks in advance,
Scott Pedersen
Robert Jenks
On Tue, 10 Sep 1996 15:49:00 -0400, Richard Fecteau - Reflection X
<rich.f...@analog.com> wrote:
>Walter Benzing wrote:
>>
>> Hi,
>>
>> Loyd Case wrote:
>>
>> >Hi, folks....
>>
>> >I just completed testing for the first half of Computer Gaming World's
>> >two-part graphics card roundup.Cards which will appear in the November issue
<< major snippage - gotta save bandwidth...>>
>> Just to cheer those guys up that got a "3D-card", it was also said that all
>> cards were fairly good at dos-speed, with the mystique being a tick faster than
>> the millenium.
>>
>> P.P.S:
>> I'm not affected with PC-Player other than this being my favorite game mag
>> because it is more critical than others.
>
>No surprise here. We'll have to wait for Verite or Voodoo-based cards to
>get a real 3D speedup. My money is waiting...
>
>Rich
I, too, am waiting and planning to get a 3DFx based card. Trouble is,
I need to upgrade two systems from VLB to PCI motherboards so I will
be forced to get two video cards. I have thought about getting a
mystique and then later adding a 3DFx daughtercard to give it some
real 3D power. My questions, will the 3DFx card handle all the 3D
rendering or will the mystique intercept those functions that it can
perform and only pass through functions that it does not support.
Perhaps I should get a pure 2D card like a Diamond stealth 64 3000 and
supplement it with a 3DFx add-on. Any opinions....
Robert Jenks <rje...@ldl.net>
Adam Billyard
LawMAN wrote:
>
>
> I hope that when you review the PowerVR in the December issue, you could
> specifically ask VideoLogic/NEC whether they are supporting bilinear
> filtering in their 3D cards. If not, are they going to? ALX is right that
> without bilinear filtering, it doth sucks...
>
It really isn't as simple as that. The Matrox Mystique seems to have
got a lot of stick in this newsgroup for not having bilinear filtering
as if this is the most important feature of a card. We've extensively
tested pretty much every 3D card that's out there or that has been
announced using hand-coded drivers for our games.
One of our games, Scorched Planet, supports 3dfx, mystique and verite
and on each card we play to their respective strengths. The Mystique
might not have bilinear filtering but it does have the second fastest
sustained fill rate we've seen from *any* card (first being the 3dfx).
It also has palettized textures (like the playstation supports) which
means you can get very much more detailed textures on the board at any
one time - texture upload/swapping is a significant hit on overall
performance - a point that is often ignored.
My point is that its easy to be seduced by the marketing bollocks (which
is often propagated by the media) that suggests that such and such a
feature is vital when in fact it doesn't stop having great games
running. Fill rate is the feature that most cards spectacularly fail to
deliver - S3 and ATI being the notable examples where software
rasterisation on a fast Pentium is quicker. And yet I believe both
offer filtering - looking nice and running at a miserable framerate is
unacceptable to games players.
The fact that MtMadness runs like a dog in software and runs like a dog
with hardware - even using the 3dfx its pretty sad to see - is as much a
testament to the inefficiencies of D3D than anything else. We found
Scorched Planet runs 30% faster using RenderWare device drivers than
going thru D3D. I guess we have to just hope D3D does eventually
deliver some performance...
Adam.
--
-------------
Adam Billyard,
Criterion Software Limited,
Westbury Court, Buryfields, Guildford, Surrey, GU25AZ, UK.
Tel: +44 (0) 1483 406203 Fax: +44 (0) 1483 406211
M.B.
On Thu, 12 Sep 1996 19:29:26 -0400, desslock <dess...@interlog.com>
wrote:
>Scott Mathers wrote:
>>
>> Pat wrote:
>> >
>> > BEWARE of SciTech.
>> >
>> > A lot of people have reported problems with V5.3.
>> > This is especailly true for users of #9 Motion 771 w/Hawkeye drivers.
>> > There is a fairly full folder of complaints on AOL with very
>> > little in the way of support -- mostly snotty replies or lame
>> > suggestions.
>> >
>>
>> Yeah, I'd be really upset if I gave them my money. Version 5.2 made
>> my card go insane. I e-mailed them, basically saying I was interested
>> in buying it, not until they fixed it for my 771. Well 5.3 came out,
>> now it will turn my win 95 to a funky translucent shade of yellow.
>> Not so cool. Next card will have built-in VESA 2.0 support. Then
>> I can say screw those damn drivers.
>>
>> What good 2D cards have VESA 2.0 support built-in?
>
>The Matrox Millenium ain't too shabby.
>The Matrox Mystique will also have VESA 2.0 built in, and unproven 3d
>capabilities.
>--
My #9 FX reality 332 w/ ViRGE 3D has built in vesa 2.0
Dave Potts
Robert Jenks wrote:
>
> I, too, am waiting and planning to get a 3DFx based card. Trouble is,
> I need to upgrade two systems from VLB to PCI motherboards so I will
> be forced to get two video cards. I have thought about getting a
> mystique and then later adding a 3DFx daughtercard to give it some
> real 3D power. My questions, will the 3DFx card handle all the 3D
> rendering or will the mystique intercept those functions that it can
> perform and only pass through functions that it does not support.
> Perhaps I should get a pure 2D card like a Diamond stealth 64 3000 and
> supplement it with a 3DFx add-on. Any opinions....
>
> Robert Jenks <rje...@ldl.net>
There are two different types of 3D video cards comming out. Those that
require a previous 2D card and the all in one cards. As of now the all
in one cards either perform well for 3D or well for 2D not both. I'm
sure that this will eventually be corrected but I too couldn't wait till
that happened. If you want a real powerhouse purchase one of the better
2D cards now and then when the good stand alone 3D cards come out get
one of them. There is a connector on the 2D cards that support pass
thru, this means the 3D cards will handle all the 3D work and pass the
2D work to the 2D card. This is good for a game that uses both 3D and
2D effects suchs as Zbuffering and Gaud (sp?) shading. Since allinone
cards only have one RAM you have to split this among the 3D work and
the 2D work, but two seperate cards have their own RAMS and the 3D
RAM will be overlayed (zbuffered) with the 2D. It give better performance
but you pay for it. And it takes 2 PCI slots not just one. Eventually
the All in one cards will be just as good but till then the seperate
cards will be better.
Dave
--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
-- David Potts Views and Beliefs are my own, and do not --
-- Lucent Technologies represent Lucent Technologies in any way --
-- (Bell Labs Innovations) shape or form. --
-- dav...@lucent.com --
--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
~Ghost Rider~
: > We're releasing a game called Scorched Planet in October/November,
: > which supports software, Matrox Mystique, Verite boards and Voodoo
: > boards.
: >
: > If there's interest, I'll put some screenshots from each of the
: > 'versions' on our web site, so you can compare how it looks on
: > each of the chips mentioned. I'll also put some 2 level demos
: > up in a week or two, if people want.
Any support for ViRGE ?
BTW wat's your website ?
--
=======================Ghost=Rider============================ \ == / =====
Tan Khim Boon => tank...@iscs.nus.sg _____-/\-_____
National University Of Singapore \ / \_\/_/
Finger : tank...@fingerhost.iscs.nus.sg____\___/O\___/____
==========NUS=DISCS==========================\_\\_//_/=====================
rj
>be advised that they have a history of not providing new drivers for
>hardware (soundwave 32 still doesn't have win 95 drivers.) After having
>been stung by them once before I don't think I'd ever buy one of their
>products again, no matter how good it looks.
3DFX has already stated they will be writing all the drivers for their
chipset, so that really should not be a concern.
rj
--
sig? I don't need no stinkin sig
Patrick Scott
Jaggernaut <ja...@concentric.net> wrote in article
<5157p8$l...@herald.concentric.net>...
> 'em coming cos they aint there yet. Although you have to realise that
> when running say Monster Truck on a 3D card it's running in 16bit with
> filtering and texture smoothing at about the same speed as a 2d card
> in 8bit with no special effects.
I think they missed that point entirely.... Good of you to bring it up.
Once again, why news groups outperform mags!
--
Patrick Scott
Raconteur Extraordinaire
ALX
In article <516u49$3...@saturn.exodus.net>, ma...@micronics.com@ [Marc Warden]
wrote:
> Hi ALX.
>
> While it is true that software is holding up the release of the Righteous
> 3D, it is only a rumor that it is Quake.
>
> The Orchid rep who mentioned Quake in a note, a copy which was posted on
> the newsgroup, had a question mark after Quake, intending to indicate that
> Quake was possibly being considered for bundling with the board, nothing
> more.
Hey Marc,
The e-mail message from the Orchid rep that I posted to this
newsgroup (with his permission) was one in a two part message.
A follow-up e-mail that I received indicated clearly that a
Quake port was currently being worked on with an expected date
of mid-October (with the intention of sending it to all early
Orchid R3D buyers), thus removing the question mark from the
previous e-mail. Of course, this was about a month ago, so
things may have changed since then.
->ALX<-
Loyd Case
My understanding of how Direct3D works is that it will use the highest
speed card in your system. As for games which have been ported using the
proprietary libraries, if a game has been written specifically for the
Mystique, it won't use the 3Dfx card and vice-versa.
I'll try to test a dual-3D card setup in a couple of weeks and report
back... too many deadlines to try it now.
As ever,
Loyd Case
John Chih-Wei Tang
Jean <10212...@compuserve.com> wrote in article
<32389F...@compuserve.com>...
> Pc Gamer did test all these cards in their october issue. I happen to
> have a P166 and the Xpression 3D. Guess what I did the exact same test.
> I said to myself , geez these cards will make my Xpression 3D look like
> crap but suprise I get higher framerate in all the tests (using vesa
> driver from ATI)!!!! I couldn't believe it myself.... I didn't find an
> answer yet.
If you read the insert about how they conducted the tests, you'll figure
out that PC Gamer's tests are invalid. There were so many variances in
their tests -- in Duke3D, they supposedly took "the average" frame rate.
That's a pretty tough thing to do, eh? I'd suspect the results wouldn't be
consistent unless they walked along the exact same path facing the exact
same directions for each and every test.. not easy to do without
automation. And then in Quake, instead of doing TIMEREFRESH at the
beginning of the game, they opted to move to a different location. Again,
lots of variance here.. So I wouldn't put any stock in their October
evaluation.. those guys really ought to learn a thing or two about
controlled tests.
John
ALX
> It really isn't as simple as that. The Matrox Mystique seems to have
> got a lot of stick in this newsgroup for not having bilinear filtering
> as if this is the most important feature of a card.
Aside from the issue of the Mystique not offering any filtering
(which IMO is an important feature), it also uses a crude
screen-door like (four level?) effect to simulate transparency.
Performance, as you state, is an important feature. But I'd
think an improvement in image quality would be as important
a factor in purchasing a 3D card. The lack of filtering and
use of screen-door patterns doesn't really seem all that
appealing, IMO.
->ALX<-
lwhitney
Well, sorry but I think it is as simple as that.
Bilinear filtering is a must have feature, not marketing hype. Just
look at the difference. This is what makes mario64 look so cool. Also
check out the example on the rendition web site.
Sure, fill-rate is important, but high quality texturing is the
check-mark item I want to see first before I start comparing the cards
at all.
Kind of like gaming at 640x480 - used to be a luxury, now it's becoming
an expected level of quality.
If you are a gaming enthusiast I suggest you steer clear of the
Mystique.
Lee
Adam Billyard wrote:
>
> LawMAN wrote:
> >
> >
> > I hope that when you review the PowerVR in the December issue, you could
> > specifically ask VideoLogic/NEC whether they are supporting bilinear
> > filtering in their 3D cards. If not, are they going to? ALX is right that
> > without bilinear filtering, it doth sucks...
> >
>
> It really isn't as simple as that. The Matrox Mystique seems to have
> got a lot of stick in this newsgroup for not having bilinear filtering
> as if this is the most important feature of a card. We've extensively
> tested pretty much every 3D card that's out there or that has been
> announced using hand-coded drivers for our games.
...
The Mystique
> might not have bilinear filtering but it does have the second fastest
> sustained fill rate we've seen from *any* card (first being the 3dfx).
...
> My point is that its easy to be seduced by the marketing bollocks (which
> is often propagated by the media) that suggests that such and such a
> feature is vital when in fact it doesn't stop having great games
> running.
...
> Adam.
David Frost
> We're releasing a game called Scorched Planet in October/November,
> which supports software, Matrox Mystique, Verite boards and Voodoo
> boards.
>
> If there's interest, I'll put some screenshots from each of the
> 'versions' on our web site, so you can compare how it looks on
> each of the chips mentioned. I'll also put some 2 level demos
> up in a week or two, if people want.
>
Yes please! Screenshots would be great and a playable demo would
be even better!
Regards
David Frost
New Zealand
LawMAN
In article <323754...@lehigh.edu>, kl...@lehigh.edu says...
>
>John Hamilton wrote:
>>
>> In article <32363E...@lehigh.edu>
>> Ken DiStefano <kl...@lehigh.edu> wrote:
>>
>> [in reference to Mr. Case's post]
>>
>> > I don't need your test results. It's easy to figure out that any card
>> > based on 3dfx's Voodoo chipset will perform the best when it comes to 3D
>> > game acceleration.
>>
>> Sure, and we know that each implementation using the Voodoo will be
>> optimal and well engineered, without any bugs, and not troublesome to
>> use. Just like we know every new microprocessor from Intel will be
>> perfect. Not. ;-)
>>
>> john
>>
>
>There are only 2 cards that I know of and by the time this issue comes
>out I'll have read which one is better right here. My point was, I
>don't need some magazine telling me that Monster 3D kicks the Mystique
>into the ground when anyone can plainly see that.
>
>-Ken
Who cares what you think? Not everyone is totally loaded like you to have 2
graphic adapters in their systems. So even if every 'Mr.Know_It_All' knows that
the Voodoo chipset is the best quality, NOT everyone can afford it, can they?
Therefore the magazine will show to future users of 3D the best
price/performance choice of a graphic adapter.
Anyone can plainly see that EXCEPT evidently you with an IQ of below ZERO !!!
A conclusion is simply the place where you got tired of thinking.
--
.oOOOOOo.
LawMAN (lawre...@nortel.com) \\ //
Harry
Just a coment from an previous owner of a Diamond 3d . It's games also
"looked" much nicer than the non enhanced versions but gameplay was
abismal. I have heard comments made about Matrox's decision to not
include certain 3D functions in the new Mystique was to help improve
performance.
IMHO, nice looking pictures will only get you so far through a game.
Eventually its the gameplay performance that keeps you coming back.
So I'd be willing to give up a little graphical detail (to a point)
for better game play .
This comment is in no way an endorsement of the Mystique and if money
were no object I would definately go with a Voodoo/2d card combo but a
lot of us here have to make the dreaded price/performance decision .
Boomer
lawrence....@nortel.com (LawMAN) wrote:
>In article <323754...@lehigh.edu>, kl...@lehigh.edu says...
>>
>>John Hamilton wrote:
>>>
>>> In article <32363E...@lehigh.edu>
>>> Ken DiStefano <kl...@lehigh.edu> wrote:
>>>
>>> [in reference to Mr. Case's post]
>>>
>>> > I don't need your test results. It's easy to figure out that any card
>>> > based on 3dfx's Voodoo chipset will perform the best when it comes to 3D
>>> > game acceleration.
Anybody got any updated info re: voodoo chipset? Any availability
info, web pages, etc?
ALX
In article <323a55d8...@news.cris.com>, hst...@concentric.net (Harry)
wrote:
> IMHO, nice looking pictures will only get you so far through a game.
> Eventually its the gameplay performance that keeps you coming back.
> So I'd be willing to give up a little graphical detail (to a point)
> for better game play .
So would I, but the Mystique IMO, simply gives up too much
graphical detail.
> This comment is in no way an endorsement of the Mystique and if money
> were no object I would definately go with a Voodoo/2d card combo but a
> lot of us here have to make the dreaded price/performance decision .
For those already satisfied with their 2D board, the
price/performance ratio of adding a Voodoo is certainly better
than a Mystique.
I dare say a 2D+3DFX combo's price/performance doesn't
differ much from the Mystique's. If you're to believe
Loyd Case's recent comparisons of the various boards, the
3DFX is easily 3 to 4x faster than the Mystique. A
2D+3DFX combo would cost perhaps $450, with the 2 meg Mystique
alone at $150. In this scenario, the price/performance
between the two would be equal (assuming the 3DFX is faster
by a factor of 3).
Of course, if you place far more weight on price than
performance, the comparison above is invalid, but in
terms of straight price/performance ratio values, the
two are pretty much equal.
->ALX<-
LawMAN
In article <U6AOykw8...@cet.com>, r...@cet.com says...
Only one problem, what is 3DFx track record for stable drivers ???
That's the one and only question, Watson.
Frank Milera
In article <3238C7...@cris.com>, tr...@cris.com says...
>
>Scott Mathers wrote:
>>
>> Pat wrote:
>> >
>> > BEWARE of SciTech.
>> >
>> > A lot of people have reported problems with V5.3.
>> > This is especailly true for users of #9 Motion 771 w/Hawkeye drivers.
>> > There is a fairly full folder of complaints on AOL with very
>> > little in the way of support -- mostly snotty replies or lame
>> > suggestions.
Just for the record, I use SDD 5.3 in DOS only (Win95 DOS, that is), and have
experienced no problems with my Stealth 64 DRAM PCI
Steve Berlinsky
Boomer wrote:
>
> Anybody got any updated info re: voodoo chipset? Any availability
> info, web pages, etc?
At the risk of ending up on the big sh*t list in the ICI skies the only
logical answer is:
TWO WEEKS!
<G> Actually I believe Loyd C did mention something about it being out
in a couple of weeks.
There is a web page (search under Orchid). I don't have the address
handy because of a total reload of WIN95 due to a mb upgrade.
Ape
Chad E. Kirk
I agree. Display Doctor is a waste of money. It would not work on my PC,
which has an
integrated Vesa local bus 32 bit, Cirrus Logic CL-5424 video with 1 meg
video dram.
When the Win 95 GUI would load, my screen would blank, then if I hit
ctrl+alt+del and
tab twice to shut down the computer, I could then see the, Windows is now
shutting
down and It is now safe to turn off your computer screens. So, I just
waited untill my
"free trial" was over than I uninstalled it. Oh, btw, you must edit the
registry to get all of the
keys that SDD 5.3 leaves in there, it leaves three keys for some reason.
later
--
Chad E. Kirk
zi...@eastky.com
Henry Tran
>>> > A lot of people have reported problems with V5.3.
>>> > This is especailly true for users of #9 Motion 771 w/Hawkeye drivers.
>>> > There is a fairly full folder of complaints on AOL with very
>>> > little in the way of support -- mostly snotty replies or lame
>>> > suggestions.
>
>Just for the record, I use SDD 5.3 in DOS only (Win95 DOS, that is), and have
>experienced no problems with my Stealth 64 DRAM PCI
Just wanted to know if Scitech 5.2 is really worth it? Does it improve
your FPS or what? Thanks for any help.
-Henry
Frank Lachance
Go ahead.. but what is your web page?
David Lau-Kee <lau...@csl.com> wrote in article
>
> We're releasing a game called Scorched Planet in October/November,
> which supports software, Matrox Mystique, Verite boards and Voodoo
> boards.
>
> If there's interest, I'll put some screenshots from each of the
> 'versions' on our web site, so you can compare how it looks on
> each of the chips mentioned. I'll also put some 2 level demos
> up in a week or two, if people want.
>
> movs and avis may not help too much because of the playback issues.
>
> David
>
>
>
David Webb
Chad E. Kirk wrote:
>
> I agree. Display Doctor is a waste of money. It would not work on my PC,
> which has an
> integrated Vesa local bus 32 bit, Cirrus Logic CL-5424 video with 1 meg
> video dram.
>
> waited untill my
> "free trial" was over than I uninstalled it. Oh, btw, you must edit the
> registry to get all of the
> keys that SDD 5.3 leaves in there, it leaves three keys for some reason.
the centering would go out of wack and the whole screen would move a
few inches to the right or left. When I tried to uninstall it, the
uninstall would not work (my 21 days were not up). Luckilly I
monitored its installation with cleansweap and got rid of the thing.
I guess I should go look at my registry now...
--
===================================================================
David Webb doc...@ix.netcom.com
http://www.geocities.com/heartland/2916/
Education - that which discloses to the wise and disguises
from the foolish their lack of understanding (Ambrose Bierce)
Trips
Corsair wrote:
>
> In article <3238C7...@cris.com> Trips <tr...@cris.com> writes:
>
> >The STB Lightspeed 128 has VESA 2.0 built in, and it screams. It's fully
> >Win 95 and DirectX compatible right out of the box, and it beat every
> >card I tested it against in Chris' Bench, Quake timerefresh, Warbirds
> >Framerate, and Air Warrior frame rate.
>
> How much did you pay for it Trips?
>
> Corsair
It was $199 at CompUSA...
Trips
Corsair
>The STB Lightspeed 128 has VESA 2.0 built in, and it screams. It's fully
>Win 95 and DirectX compatible right out of the box, and it beat every
>card I tested it against in Chris' Bench, Quake timerefresh, Warbirds
>Framerate, and Air Warrior frame rate.
How much did you pay for it Trips?
Corsair
__________________________________________________________
Web CAG of The Unofficial "Jolly Rogers" Site
http://www-home.calumet.yorku.ca/mdonalds/www/home.htm
__________________________________________________________
CAG of the "VF-84 Jolly Rogers" Simulation Squadron
__________________________________________________________
"I want to be remembered with the skull-&-crossbones..."
- Okinawa: 1945
__________________________________________________________
Tony Marshall
ja...@concentric.net (Jaggernaut) wrote:
>ben...@iegi01.etec.uni-karlsruhe.de (Walter Benzing) wrote:
>
>
>
>> ATI 3D Xpression (2 MB)
>> Elsa Victory 3D (S3 Virge) with 4 MB Ram
>
>> as there were other 2 MB S3 cards in the test (including a 2 MB Victory) ,
>> the tester guessed that hardware speedup is 4 MB only with S3/Virge
>
>BS, he guessed wrong. Hardware acceleration in direct3d works fine on
>the 2mb Diamond Stealth 3D 2000. Obviously these tests where run
>before the correct drivers where out. As for the speed, well...keep
>'em coming cos they aint there yet. Although you have to realise that
>when running say Monster Truck on a 3D card it's running in 16bit with
>filtering and texture smoothing at about the same speed as a 2d card
>in 8bit with no special effects.
>
That's no good though. They need to get the framerate nice and high
instead of adding more colours/higher res etc. At least, that's where
my priority lies... If there's time left over, then fine, add some
extra features, but just using the power of the card to increase the
detail and not the framerate is useless in my book!
Tony Marshall | <tangerine dream.yello.gordon giltrap
Carlisle, England | neuronium.depechemode.psb's.oasis.jmj
to...@zigzag.u-net.com| ashra.stranglers.vangelis.floyd.goa >
----------------------|--------------------------------------
My Mystery Tune Page http://www.personal.u-net.com/~zigzag
Tony Marshall
ds...@monmouth.com (Dhiren Shah) wrote:
>>Very interesting. But you know what will make the next two magazines more
>>interesting? Video clips of games of all the 3D cards you tested
>>featured in that issue. Of course, this would be available only on the
>>CD+magazine bundle.
>
>>Screen shots and specs make a man drool, but its the .mov/.avi (in a
>>window size of at least 320x240, not some small 160x120) that will make a
>>man foam. :)
>
>>-george
>
>I agree...we really need to see how smooth and how good these
>different cards look...not those crappy videos they throw onto cds
>these days. Do us a favor and leave off the America Online and
>include a couple of videos of the 3d cards. At least include full
>640x480x24bit still shots of the 3d games.
>
Why not covermount a VHS video tape? I've seen quite a few UK mags
with this, and it'd be far better than a crappy old .AVI or whatever
(which always looks awful)
[Marc Warden]
In <51bcjc$g...@news01.deltanet.com>, unk...@account.com (ALX) writes:
>Hey Marc,
>
>The e-mail message from the Orchid rep that I posted to this
>newsgroup (with his permission) was one in a two part message.
>A follow-up e-mail that I received indicated clearly that a
>Quake port was currently being worked on with an expected date
>of mid-October (with the intention of sending it to all early
>Orchid R3D buyers), thus removing the question mark from the
>previous e-mail. Of course, this was about a month ago, so
>things may have changed since then.
Hi ALX.
Learn something new everyday.
Sincerely,
Marc Warden, Orchid Technology (ma...@micronics.com)
CJ Martin
Trips <tr...@cris.com> wrote in article <3238C7...@cris.com>...
> The STB Lightspeed 128 has VESA 2.0 built in, and it screams. It's fully
> Win 95 and DirectX compatible right out of the box, and it beat every
> card I tested it against in Chris' Bench, Quake timerefresh, Warbirds
> Framerate, and Air Warrior frame rate.
>
> Trips
For some odd reason, EF2000 does not recognize the built in VESA 2.0
support, forcing me to load DD 5.3 (DOS version only, thank you!).
Janes Longbow correctly recognized the VESA 2.0 support, and runs like a
champ. Duke 3D also runs very well, even in full screen 800 x 600 mode.
WarBirds runs about 5-10 fps faster than my Matrox equipped home machine
(but the home rig is a 133, vice the 166 at work).
CJ
Scott
I had to de-install 5.3 since quake didn't like it. Crashed everytime. 5.2
and earlier work just fine though.
I have an S3trio32 card with 1 meg memory.
--
Scott Tyson
Whip...@quiknet.com
Kali Sig:Squidly
Quake Sig: CC_Squidly (Clan Cthulhu)
"If I turn all my lights out.. I won't have to clean my house"
Flotsam and Jetsam
Frank Milera <Fra...@ix.netcom.com> wrote in article
<51fvaf$e...@dfw-ixnews5.ix.netcom.com>...
> In article <3238C7...@cris.com>, tr...@cris.com says...
> >
> >Scott Mathers wrote:
> >>
> >> Pat wrote:
> >> >
> >> > BEWARE of SciTech.
> >> >
Ben Hall
> Just wanted to know if Scitech 5.2 is really worth it? Does it
improve
> your FPS or what? Thanks for any help.
YES!! Absolutely! It didn't on my ATI Video card, but it more than doubled
the performance of my new S3 Trio64 card (used by Dieamond Stealth64 a lot)
It is certainly worth downloading. If it doen't work all you've lost is a
little time.. It is certainly cheaper than buying a new video card.
Ben
Mike Marini
Trips wrote:
> > >Win 95 and DirectX compatible right out of the box, and it beat every
> > >card I tested it against in Chris' Bench, Quake timerefresh, Warbirds
> > >Framerate, and Air Warrior frame rate.
> >
> >
> > Corsair
>
>
> Trips
Hey Trips, the last issue of PC Gamer did a comparison of the STB board
with one from Hercules, both with the same Tseng chipset.
According to them the Hercules beat the STB by a small, but significant
margin (using the Quake & Duke3D frame counters)
Did you come across this other board?
Chry
Samuel Ng
LawMAN wrote:
> >
> >3DFX has already stated they will be writing all the drivers for their
> >chipset, so that really should not be a concern.
> >
>
> Only one problem, what is 3DFx track record for stable drivers ???
> That's the one and only question, Watson.
>
They do not have a track record because it is their 1st PC product.
The Voodoo chipset was developed to power arcade machines which has
been and still is 3dfx's mainstay.
Sam
Geoff Yates
I'd like to agree with everybody on this thread so far.
Thanks for the tip on the registry.
ps. just for the record my PC is equipped with a VESA LB Cirrius Logic
5429.
Jaggernaut
to...@zigzag.u-net.com (Tony Marshall) wrote:
>That's no good though. They need to get the framerate nice and high
>instead of adding more colours/higher res etc. At least, that's where
>my priority lies... If there's time left over, then fine, add some
>extra features, but just using the power of the card to increase the
>detail and not the framerate is useless in my book!
>
>Tony Marshall | <tangerine dream.yello.gordon giltrap
>Carlisle, England | neuronium.depechemode.psb's.oasis.jmj
>to...@zigzag.u-net.com| ashra.stranglers.vangelis.floyd.goa >
>----------------------|--------------------------------------
>My Mystery Tune Page http://www.personal.u-net.com/~zigzag
Completely agree with you. I just want to see games with graphic
detail similar to EF2000 running faster but instead they just add even
more graphics details to bog things down even more. On my Stealth 3D
2000 running the Descent II 3D port on a p133 with 32mb of ram the
game runs quite smooth until things start exploding. That's when it
get's choppy once again. This to me looks like the explosions are not
optimized for the 3d card so perhaps the initial poor performance of
the Stealth 3D can be partly attributed to lack of efficient
programming. I read that the initial release of the Hellbender demo
ran poorly but after optimization it now runs quite smooth. I intend
on getting a 3dfx card later anyway but the Stealth's 3D capabilities
are not exactly crap as some people would have us believe.
--
Jagg
Vancouver, Canada
***************************
WB Handle:Term <Terminator>
401st RCAF <Rams>
http://bc1.com/users/prophet/ramsquad.html
Steven C. Den Beste
Jaggernaut wrote:
>
> "Ben Hall" <b...@icity.intcity.com> wrote:
>
> >YES!! Absolutely! It didn't on my ATI Video card, but it more than doubled
> >the performance of my new S3 Trio64 card (used by Dieamond Stealth64 a lot)
> > It is certainly worth downloading. If it doen't work all you've lost is a
> >little time.. It is certainly cheaper than buying a new video card.
>
> >Ben
>
> I'm really dubious of your claim. The most increase I would expect to
> see in performance would be 10%.. If that even.
>
My performance increase with Duke Nukem 3D was about 50%; I have a
"Stealth 64 Video VRAM PCI" connected to a P133.
The improvement was astounding, and I think most of it comes from the
abysmal quality of Diamond's BIOS. Diamond's software has always been a
weak spot.
Smoke Crack and Worship Satan
In article <323A06...@netcom.ca>, lwhitney <lwhi...@netcom.ca> wrote:
> Well, sorry but I think it is as simple as that.
>
> Bilinear filtering is a must have feature, not marketing hype. Just
> look at the difference. This is what makes mario64 look so cool. Also
> check out the example on the rendition web site.
>
> Sure, fill-rate is important, but high quality texturing is the
> check-mark item I want to see first before I start comparing the cards
> at all.
>
> Kind of like gaming at 640x480 - used to be a luxury, now it's becoming
> an expected level of quality.
>
> If you are a gaming enthusiast I suggest you steer clear of the
> Mystique.
>
That's a good point, but delivering *FAST* textured, perspective correct
polygons is more important to me. Filtering is icing on the cake.
I fail to see the spectacular difference between something that gets
blurry up close and something that gets pixelated up close. Big frigging
improvement! Like I said, speed is more important.
Also, I've played Mario 64 and I think it's a terribly overhyped game.
I do think it makes excellent use of 3D, though. I just can't stand
that cutesy Mario shit. I'd like to see Mario running around with
a meat axe, clubbing those adorable enemies to death instead.
Smoke Crack and Worship Satan
In article <51d7ak$s...@tandem.CAM.ORG>, lawrence....@nortel.com (LawMAN)
wrote:
> In article <323754...@lehigh.edu>, kl...@lehigh.edu says...
> >
> >John Hamilton wrote:
> >>
> >> In article <32363E...@lehigh.edu>
> >> Ken DiStefano <kl...@lehigh.edu> wrote:
> >>
> >> [in reference to Mr. Case's post]
> >>
> >> > I don't need your test results. It's easy to figure out that any card
> >> > based on 3dfx's Voodoo chipset will perform the best when it comes to 3D
> >> > game acceleration.
> >>
> >> optimal and well engineered, without any bugs, and not troublesome to
> >> use. Just like we know every new microprocessor from Intel will be
> >> perfect. Not. ;-)
> >>
> >> john
> >>
> >
> >There are only 2 cards that I know of and by the time this issue comes
> >out I'll have read which one is better right here. My point was, I
> >don't need some magazine telling me that Monster 3D kicks the Mystique
> >into the ground when anyone can plainly see that.
> >
> >-Ken
>
> Who cares what you think? Not everyone is totally loaded like you to have 2
> graphic adapters in their systems. So even if every 'Mr.Know_It_All' knows
> that
> the Voodoo chipset is the best quality, NOT everyone can afford it, can they?
> Therefore the magazine will show to future users of 3D the best
> price/performance choice of a graphic adapter.
>
> Anyone can plainly see that EXCEPT evidently you with an IQ of below ZERO !!!
>
> A conclusion is simply the place where you got tired of thinking.
>
Well, from what I've been able to synthesize...
First, the Voodoo chipset is by *FAR* the fastest on the market.
Absolutely no doubt about this from any source I've read. No question.
Second, why is replacing your existing video card a good thing? If you
have a shitty Cirrus Logic card, then it makes sense. However, why would
I want to replace my tried and true Matrox Millenium with a 2D/3D card
that may have below-average 2D performance even though the 3D acceleration
is good?
So it seems to me spending $300 for a dedicated 3D card makes sense.
After all, what are you going to do with your current video card when
you replace it? Use it as a coaster?
What I really want to know is how fast the Rendition chipset is in
standard 2D windows benchmarks. I know it delivers faster-than-software
rendering at 16-bit color with filtering on, which is my #1 criteria
for purchasing a 3D card, but I don't want crappy 2D performance to
go with that.
At any rate, $199 for the Rendition with 4MB EDO memory is a bargain,
and it looks like a good choice regardless. Let's hope people rally
around Direct3D so we can actually pick the card we want.
P.S. GO DOWNLOAD THE HYPERBLADE DEMO! It's *really* good. And it
uses Direct3D.
Jaggernaut
"Ben Hall" <b...@icity.intcity.com> wrote:
>YES!! Absolutely! It didn't on my ATI Video card, but it more than doubled
>the performance of my new S3 Trio64 card (used by Dieamond Stealth64 a lot)
> It is certainly worth downloading. If it doen't work all you've lost is a
>little time.. It is certainly cheaper than buying a new video card.
>Ben
More than doubled????!!! Can you post some benchmarks to prove this.
I'm really dubious of your claim. The most increase I would expect to
see in performance would be 10%.. If that even.
--
Tommy McClain
Smoke Crack and Worship Satan <nom...@nomail.com> wrote in article
<51l9cv$d...@nntp.crl.com>...
> What I really want to know is how fast the Rendition chipset is in
> standard 2D windows benchmarks. I know it delivers faster-than-software
> rendering at 16-bit color with filtering on, which is my #1 criteria
> for purchasing a 3D card, but I don't want crappy 2D performance to
> go with that.
We've seen some Rendition cards card and they seem to have the fastest SVGA
support out there. We're hoping to post more stuff soon.
> At any rate, $199 for the Rendition with 4MB EDO memory is a bargain,
> and it looks like a good choice regardless. Let's hope people rally
> around Direct3D so we can actually pick the card we want.
>
> P.S. GO DOWNLOAD THE HYPERBLADE DEMO! It's *really* good. And it
> uses Direct3D.
I've been told the demo doesn't supprot hardware acceleration. They removed
texture mapping. I haven't downloaded because if this, but if this is
incorrect I'd gladly download it.
Tommy McClain
Dimension 3D
http://www.Dimension3D.com/
Dave Glue
Smoke Crack and Worship Satan <nom...@nomail.com> wrote in article
<51l9lf$d...@nntp.crl.com>...
> That's a good point, but delivering *FAST* textured, perspective correct
> polygons is more important to me. Filtering is icing on the cake.
>
> I fail to see the spectacular difference between something that gets
> blurry up close and something that gets pixelated up close. Big frigging
> improvement! Like I said, speed is more important.
Sure, if you're not pulling 25+ fps, then speed is the main consideration.
But if you can't see that bi-linear filtering is a big improvement over
point sampled, then I wonder if you've actually ever seen it.
You've played Mario64, I take it? Bi-linear filtering _IS_ the N64. It
would look _horrible_ without it. It is the THE BIGGEST improvement in
image quality for PC games in a long time. It's the reason VF2, Sega Rally
etc look so good.
I believe the latest EDGE had a shot of Rave Racer on the Videologic. Look
at that, then look at the shots of Whiplash on the 3DFX. Can you tell me
that point-sampled textures don't look like *crap*!? Whiplash is a far
less graphically complex game than Rave Racer, but the blocky textures in
the PowerVR just ruin it- it looks nothing like the arcade. Bi-linear
filtering is *huge*. It's hardly just "blurry" instead of "blocky"- the
textures are "blurred" somewhat, just as anti-aliasing blurs the edges of
objects to make them appear higher resolution than they are. It's a
tremendous leap in image quality, it can cover up for a lot of graphical
mistakes and low-res textures. I really can't believe you would make such
a statement, big-pixellated textures was *your* main complaint about games
being hyped for being in SVGA when your viewpoint is close-to-medium
distance, everything looks like a mess of pixels! The Shambler textures in
Quake look really good close up, don't they?
> Also, I've played Mario 64 and I think it's a terribly overhyped game.
> I do think it makes excellent use of 3D, though. I just can't stand
> that cutesy Mario shit. I'd like to see Mario running around with
> a meat axe, clubbing those adorable enemies to death instead.
Me too. But no one would be impressed with Mario on a technical level if
it didn't have filtering- heck, all N64 games would look like shite without
its filtering capabilities.
Jaggernaut
"Steven C. Den Beste" <denb...@world.std.com> wrote:
>> More than doubled????!!! Can you post some benchmarks to prove this.
>> I'm really dubious of your claim. The most increase I would expect to
>> see in performance would be 10%.. If that even.
>>
>My performance increase with Duke Nukem 3D was about 50%; I have a
>"Stealth 64 Video VRAM PCI" connected to a P133.
>The improvement was astounding, and I think most of it comes from the
>abysmal quality of Diamond's BIOS. Diamond's software has always been a
>weak spot.
Well, I have a Diamond Stealth 3D 2000 and I have seen negligible
performance increase with either SDD5.2 or 5.3. I guess Diamond
finally got it right.
guyver
> > >>
> > >> [in reference to Mr. Case's post]
> > >>
> > >> > I don't need your test results. It's easy to figure out that any card
> > >> > based on 3dfx's Voodoo chipset will perform the best when it comes to 3D
> > >> > game acceleration.
> > >>
> go with that.
>
> At any rate, $199 for the Rendition with 4MB EDO memory is a bargain,
> and it looks like a good choice regardless. Let's hope people rally
> around Direct3D so we can actually pick the card we want.
>
> P.S. GO DOWNLOAD THE HYPERBLADE DEMO! It's *really* good. And it
> uses Direct3D.
I tried a couple of 3d video cards with the same game and in term of frame
rate the 3dfx is the fastest.
--
the opinion expressed in that posting are mine only and
doesn't represent the opinion of my employers.
David Lau-Kee
David Frost <fro...@ihug.co.nz> wrote:
...
|Yes please! Screenshots would be great and a playable demo would
|be even better!
Hi-res screenshots are available to download from:
http://www.csl.com/studios/spshots.htm
I've put up in-game screenshots from Scorched Planet. 3 shots from
each of the Matrox Mystique, 3DFX Voodoo, Rendition Verite and native
RenderWare software versions. Performance wise, on a 133 in 640x480,
the voodoo is fastest at around 40fps, the Mystique comes
in at around 25-30fps and the verite at about 15-20fps.
In terms of '3D features' used by Scorched Planet, the Verite has
fogging enabled and the Voodoo has bilinear filtering and
fogging enabled.
The software version (RenderWare) has 8-bit color, true
perspective-correct texturing, fogging, is native Win95 and
native DOS, and comes in at 25-30fps on a 133 in 320x200.
Note: the hardware versions use RenderWare device drivers (not
Direct3D), and the native Win95 versions use RenderWare to
render and DirectDraw to blit.
For the people who asked about other chips & generic D3D support:
Scorched Planet only supports 3D chips which offer real performance
improvements over straight RenderWare software; and our test D3D
version was 50-60% slower in s/w than native RenderWare and 30% slower
in D3D-driven hardware than RenderWare-driven hardware, so we're
not planning to release a D3D version.
I'll put playable demos for each version up on the same site
in two weeks.
E&OE, caveat emptor, your mileage may vary, etc.
Enjoy!
David
Trips
No, but now I'll have to go find one... thanks for the info.
Trips
Alex Fernandez
On Wed, 18 Sep 1996 13:09:21 GMT, lau...@csl.com (David Lau-Kee) wrote:
>David Frost <fro...@ihug.co.nz> wrote:
>...
>[snip]
>
>I've put up in-game screenshots from Scorched Planet. 3 shots from
>each of the Matrox Mystique, 3DFX Voodoo, Rendition Verite and native
>RenderWare software versions. Performance wise, on a 133 in 640x480,
>the voodoo is fastest at around 40fps, the Mystique comes
>in at around 25-30fps and the verite at about 15-20fps.
>
>In terms of '3D features' used by Scorched Planet, the Verite has
>fogging enabled and the Voodoo has bilinear filtering and
>fogging enabled.
Why is it that the developers didn't use more of the 3D features of those
chips? Seems to me that using the Verite just for fog effects is just a
wasted effort, IMHO.
>
>[snip]
>
>David
>
>
===========================================================
Alex Fernandez
MTS
Lucent Technologies
No. Andover, MA
email:
alexfe...@lucent.com
tel: (508)960-6510
===========================================================
sean
Hmm, it seems like there is a lot of comments but what we all need is
number and a compiled list of features in order for us to decide which
is best. Secondly, are the new 3D boards that are coming out designed
to replace your video card or designed to go with it. I'm thinking of
getting a good 2D card (Hercules Dynamite 128) and waiting until the
dust settles on the 3D card technology before I buy any 3D hardware to
couple to my video card. L8R...Sean
Dave Glue
David Lau-Kee <lau...@csl.com> wrote in article
<51os59$b...@cruella.criterion.canon.co.uk>...
> RenderWare software versions. Performance wise, on a 133 in 640x480,
> the voodoo is fastest at around 40fps, the Mystique comes
> in at around 25-30fps and the verite at about 15-20fps.
So basically, if you don't have a Voodoo, don't bother. Sorry, but a game
that barely deviates graphically from Terminal Velocity, running at about
the same frame rate as Quake on the Verite is not impressive. And I'm not
eager to play games in VGA on my P166. I've played the Mystique version,
and although "smooth" some serious draw-in was evident. I don't know why
the engine taxes the Verite to such a degree, this is the first I've heard
of the Mystique being faster than the Verite in 3D (albeit without
filtering). And good lord, you're not even using filtering on the Verite-
why bother at all?
Dave Glue
Robbie Diehl <lep...@thepoint.net> wrote in article
<323F8E...@thepoint.net>...
> > Bilinear filtering is a must have feature, not marketing hype. Just
> > look at the difference. This is what makes mario64 look so cool. Also
> > check out the example on the rendition web site.
> >
> > Sure, fill-rate is important, but high quality texturing is the
> > check-mark item I want to see first before I start comparing the cards
> > at all.
>
> Then again, Mario 64 uses "tri-linear" interpolation. I reckon that
> looks even better than bi-linear interpolation.
Other that being a buzzword, what does that mean?
First off, Mario is in 256*224, and on a TV- which has some filtering of
its own due to a blurry picture. Secondly, all Tri-linear filtering is, is
seperate mip-maps swapped in and the levels filtered between them. It
takes a significantly larger memory hit and bandwidth to do with little
benefit. With the low resolution, TV and lack of cartridge space, I would
be amazed if Mario 64 used any mip-map swapping at all.
ALX
In article <51os59$b...@cruella.criterion.canon.co.uk>, lau...@csl.com (David
Lau-Kee) wrote:
> I've put up in-game screenshots from Scorched Planet. 3 shots from
> each of the Matrox Mystique, 3DFX Voodoo, Rendition Verite and native
> RenderWare software versions. Performance wise, on a 133 in 640x480,
> the voodoo is fastest at around 40fps, the Mystique comes
> in at around 25-30fps and the verite at about 15-20fps.
Thanks for the excellent screen shots. The difference between
the various boards are quite stunning, in particular the
Rendition. I expected the Rendition Verite to use bilinear
filtering, but the screen shots clearly show no filtering at
all. The performance is a bit disappointing as well (15-20fps
despite the use of only point sampling?).
> In terms of '3D features' used by Scorched Planet, the Verite has
> fogging enabled and the Voodoo has bilinear filtering and
> fogging enabled.
So the Voodoo is the only board to have fogging+bilinear
filtering? Would Verite users have the option of enabling
filtering, despite the poor frame rates?
> and 30% slower in D3D-driven hardware than RenderWare-driven
> hardware, so we're not planning to release a D3D version.
Ouch, that's quite an overhead.
->ALX<-
Robbie Diehl
> Bilinear filtering is a must have feature, not marketing hype. Just
> look at the difference. This is what makes mario64 look so cool. Also
> check out the example on the rendition web site.
>
> Sure, fill-rate is important, but high quality texturing is the
> check-mark item I want to see first before I start comparing the cards
> at all.
Then again, Mario 64 uses "tri-linear" interpolation. I reckon that
looks even better than bi-linear interpolation. Either way, any
interpolation seems better than none at all for the PC.
Ethna Brown
I have an S3 Trio64V+ card, which is supposedly supported by Display
'Doctor' 5.3, but I am trying to run Deathgate in SVGA mode using
univbe version 5.3, but it won't work. Has anyone any info on this
problem.
Pat <patb...@ix.netcom.com> wrote:
>BEWARE of SciTech.
>A lot of people have reported problems with V5.3.
>This is especailly true for users of #9 Motion 771 w/Hawkeye drivers.
>There is a fairly full folder of complaints on AOL with very
>little in the way of support -- mostly snotty replies or lame
>suggestions.
>Tom Ryan wrote:
>>
>> Hi Loyd,
>>
>> Did you use the profiling utiltiy in SciTech Display Doctor 5.3? It is
>> a very good indication of the raw bandwidth available in any particuar
>> graphics card.
>>
>> Also, we may be doing a bit of a press tour soon in the Bay Area
>> showing some cool technology from us and our customers. Are you at the
>> editorial offices much or can we arrange a visit with you as well?
>>
>> Let me know
>>
>> -Tom Ryan, SciTech Software
>> "Loyd Case" <lo...@proaxis.com> wrote:
>>
>> >Hi, folks....
>>
>> >I just completed testing for the first half of Computer Gaming World's
>> >two-part graphics card roundup. The two parts will run in the November and
>> >December issues. In addition to the usual set of benchmarks (Winbench 96
>> >graphics Winmark, CBench, Duke Nukem 3D and Quake), we also used Direct3D
>> >Test (which comes with the DirectX 2.0 SDK) and the built-in frame rate
>> >counter in Hellbender.
>>
>> >Cards which will appear in the November issue (shipping in late October)
>> >include:
>>
>> >S3 ViRGE cards:
>>
>> >Diamond Stealth 3D 2000 4MB
>> >Hercules Terminator 3D 4MB
>> >Number Nine Reality 332/fx (2 MB)
>>
>> >Matrox Mystique
>>
>> >ATI 3D Xpression PC2TV (using the Rage/2 chip)
>>
>> >3D/fx cards:
>>
>> >Orchid Righteous 3D
>> >Diamond Monster 3D
>>
>> >In addition to the 3D cards, we look at the 3 main ET6000-based cards, the
>> >Hercules Dynamite/128, the STB Lightspeed 128 and the Videologic Grafixstar
>> >600.
>>
>> >For the December issue, it looks like we'll have:
>>
>> >3D Blaster PCI
>> >Several ViRGE/VX cards
>> >ATI 3D Pro Turbo
>> >Videologic PowerVR 3D accelerator
>> >And maybe (keep your fingers crossed) one or two Permedia cards.
>>
>> >Best regards,
>>
>> >Loyd Case
>> >Contributing Editor, Technology
>> >Computer Gaming World
>>
>> +-----------------------------------------------------------------------+
>> | SciTech Software - Building Truly Plug'n'Play Software -- Today! |
>> +-----------------------------------------------------------------------+
>> | Tom Ryan, Director of Marketing | Ph/FAX: 916-894-8400/916-894-9069 |
>> | SciTech Software, Inc. | Email : To...@scitechsoft.com |
>> | 505 Wall Street | ftp : ftp.scitechsoft.com |
>> | Chico, CA 95928 | www : http://www.scitechsoft.com |
>> +-----------------------------------------------------------------------+