Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Read and cry existing voodoo 2 owners (24-bit colour)

110 views
Skip to first unread message
Message has been deleted

tans...@pilot.msu.edu

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to

you must have deep pockets

mike


Nicolas Bryant

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to

24-bit colour? That seems unlikely. More probably they are just taking
advantage of the fact that the 16-bit frame buffer is converted to 24-bits at
the gamma correction stage before it hits the DAC. I'd be pretty surprised if
they've magically managed to jump to a 24-bit frame buffer.

Jeff Buchanan

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to

The Quantum board does nothing more than 2 CL VD2 in SLI.
The only "advantage" is it takes up zero PCI slots. This is offset by
the disadvantage of taking up 1 AGP slot. 24 bit color? If it uses the
same FBI and RAMDAC from the reference design, then they'll be displaying
in 16 bit also since 3DFX uses the 16 bit ICS RAMDAC.

Kai Anders Wold wrote:


Peter Scott

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to

This is just marketing hype! And we've heard enough marketing hype in this
NG before! In fact it's wrong; Voodoo 2 cannot do 3d in a window, can it?


>"TV-out", 3D in a window, dual monitor support (is this relevant to
gamers?)


Robert M Lowrey

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to

Peter Scott wrote in message <6g0mq1$pae$1...@www.3dfx.com>...


>This is just marketing hype! And we've heard enough marketing hype in this
>NG before! In fact it's wrong; Voodoo 2 cannot do 3d in a window, can it?
>


Well, see http://www.ece.ucdavis.edu/~dohnal/winglide/winglide.html and you
can do 3D in a window *now* with Voodoo1 and Voodoo2 (with reduced
performance)

John P. Van Smith

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to


Peter Scott <psc...@spambeckman.com> wrote in article


<6g0mq1$pae$1...@www.3dfx.com>...
> This is just marketing hype! And we've heard enough marketing hype in
this
> NG before! In fact it's wrong; Voodoo 2 cannot do 3d in a window, can it?
>
>

> >"TV-out", 3D in a window, dual monitor support (is this relevant to
> gamers?)

I have seen TV cards that work through a pass through that do 2d in a
window. They 2d card draws a certain color square and the tv card pastes
the image on that color. The old Philips cards did this...

John

Ed Krauss

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to Jeff Buchanan

It's PCI, not AGP!

Jeff Buchanan wrote:

--
Ed Krauss
e...@shore.net

Jeff Buchanan

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to


Ed Krauss wrote:

> It's PCI, not AGP!
>

I stand corrected, PCI it is. Ok, so for your extra $101, you gain an extra
PCI
slot. Perfrormance and features should be the same.


Usul

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to

John P. Van Smith <pon...@gate.net> wrote in article
> I have seen TV cards that work through a pass through that do 2d in a
> window. They 2d card draws a certain color square and the tv card pastes
> the image on that color. The old Philips cards did this...
>
> John


That seems interesting, but I don't think that is the issue here... When
you say that your card can do 3d in a window, that doesn't just mean that
you can display the game screen in a window with some passthrough/TV-out
hack. It means that your board supports all the necessary functions (for
example window overlays), to work seamlessly in a "windowed environment".
Example of this is any professional 3d application which combines 3d
graphics acceleration with 2d. Besides 3d graphics, you have toolbars,
text, dialog boxes and windows etc all on the same screen, all tied in
together... Voodoo1&2 don't support this, Voodoo Rush does, Banshee will.


ImajikFX

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to

Umm, its actually the other way arround i think. 24bit colour before it
hits the Frame Buffer (going from 16 to 24 would be folly).

Regards
Syama Mishra

Nicolas Bryant <nick....@dial.pipex.com> wrote in article
<3523DB10...@dial.pipex.com>...

ImajikFX

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to

The early GLiNt cards worked like this, they were seperate PCI cards that
hooked up to your 2D card and did 3D in a window by just overwriting the
space of the 3D Window (worked with 3D Max) so i don't think its a hack,
and it should support all the functions if this is the case.

Regards
Syama Mishra

Usul <muad...@NOSPAM.hotmail.com> wrote in article
<01bd5e92$573fd840$736c15d0@dark_frigidaire>...

Michael T. Scott

unread,
Apr 2, 1998, 3:00:00 AM4/2/98
to


Jeff Buchanan wrote:

It's actually more like an extra $200. Their 16 MB single-board SLI card is
$599--I see 5 mail-order sources offering the 8 MB CL 3D Blaster V2 for under
$200 US (a total of 15 sources offering it for under $220). What you get for
your extra $200+ is TV-out and a free PCI slot. I feel strangely unmoved 8^).

-- mts

NOTE: to respond to me via e-mail, either remove the "NOT" in my address, or
click the link below:

mailto:michae...@axon.axnet.com

Danny Sanchez

unread,
Apr 3, 1998, 3:00:00 AM4/3/98
to

Try again... what you get is single board SLI for the company that
pionneered SLI with Voodoo technology. I believe over the long haul
you'll also get more reliable performance with the single board versus
pushing 90MHz over a ribbon cable. Also Creative and Diamond are
relying on software fixes for SLI from 3Dfx. Quantum3D has already
dealt with the issues with the Obsidian 100SB. If you want to use 2
slots, then use 2 slots. Plus, I'm sure you will be able to overclock
the single board more than 2 boards in SLI mode (wait a minute, Creative
doesn't let you overclock from their property page do they). Just my
.02 cents

Danny Sanchez

unread,
Apr 3, 1998, 3:00:00 AM4/3/98
to Peter Scott

Actually is was misquoted and has since been corrected.

Danny Sanchez

unread,
Apr 3, 1998, 3:00:00 AM4/3/98
to Nicolas Bryant

Your comments are correct. This was misquoted and has since been
corrected I believe on VE.

WIlliam Ball

unread,
Apr 3, 1998, 3:00:00 AM4/3/98
to

Have you tried Winglide mentioned above by Robert? It displays Quake II in a overlay
window on my desktop without and TV in/out hack. Only problem is that it is very slow and
dark (no gamma support).


--
Bill

Usul wrote in message <01bd5e92$573fd840$736c15d0@dark_frigidaire>...

Mike Oliva

unread,
Apr 3, 1998, 3:00:00 AM4/3/98
to

Quantum's boards have an expansion slot which you can place a 2D
daughtercard in, so it may be possible to do Voodoo 2 in a window
with the daughtercard installed and set as your primary display, but
I'm not totally certain on this.

Peter Scott wrote in message <6g0mq1$pae$1...@www.3dfx.com>...

>This is just marketing hype! And we've heard enough marketing hype in this

>NG before! In fact it's wrong; Voodoo 2 cannot do 3d in a window, can it?


Usul

unread,
Apr 4, 1998, 3:00:00 AM4/4/98
to

Yes, I've seen it and it's pretty cool. I assume that the sofware is
copying Voodoo2 frame buffer through PCI bus and blitting it in a window.
That still qualifies as a "hack" in my opinion ( a cool hack though).
Especially since it's slow enough to be unusable.
If there was a direct connection between Voodoo2's frame buffer and the one
of 2D card, it could probably be as fast or almost as fast as full screen
Voodoo2. I think that Voodoo Rush and it's 2D companion share the same
frame buffer, which is even better.

That brings another question that I was curious about since the Rush made
it's first appearance. What makes Rush so different from Voodoo 1&2 that it
can do this? Why wasn't it possible to add a 2D chip on V2 board and have
it work in the same way as Rush? I suspect this question will remain
unanswered.

WIlliam Ball <bill...@email.msn.com> wrote in article
<6g3adv$e1g$1...@www.3dfx.com>...

Nicolas Bryant

unread,
Apr 4, 1998, 3:00:00 AM4/4/98
to

Voodoo 2 renders internally at 32-bit RGBA which is dithered down to 16 bits
for storage in the frame buffer. However, gamma correction is added to the
signal *after* it has been read from the frame buffer, and therefore what
actually hits the DAC is, technically, a 24-bit signal.

But quite honestly, who cares? :-) Danny Sanchez has now confirmed my
suspicions that the X-16 and related cards render in the same bit depth as
other voodoo 2 cards.

> Umm, its actually the other way arround i think. 24bit colour before it
> hits the Frame Buffer (going from 16 to 24 would be folly).
>
> Regards
> Syama Mishra
>
> Nicolas Bryant <nick....@dial.pipex.com> wrote in article
> <3523DB10...@dial.pipex.com>...

Hardball

unread,
Apr 4, 1998, 3:00:00 AM4/4/98
to

On Fri, 03 Apr 1998 08:36:20 -0800, Danny Sanchez
<dan...@quantum3d.com> wrote:

>Plus, I'm sure you will be able to overclock
>the single board more than 2 boards in SLI mode (wait a minute, Creative
>doesn't let you overclock from their property page do they). Just my
>.02 cents

Actually, somebody wrote a tiny app that lets you change the clock
speed without running regedit and searching for the key. It sits in
your systray.

WIlliam Ball

unread,
Apr 4, 1998, 3:00:00 AM4/4/98
to

Usul wrote in message <01bd5f79$edd235e0$736c15d0@Dark_Frigidaire>...

>Yes, I've seen it and it's pretty cool. I assume that the sofware is
>copying Voodoo2 frame buffer through PCI bus and blitting it in a window.
>That still qualifies as a "hack" in my opinion ( a cool hack though).
>Especially since it's slow enough to be unusable.

It's slow - maybe 30% of the V2 fullscreen. On a fast system it's playable.


Bill

dsp

unread,
Apr 6, 1998, 3:00:00 AM4/6/98
to

CAll 3dFx if you like.. I did. The 24 Bit color? Yeah I've got that
too on my FireGL 1000Pro. You see, not 24 bit color from the voodoo2
side of the chipset, 24 bitcolor from the 2daccellerator. When you
activate the voodoo2 side (using directx, glide, rederware, etc., the
voodoo2 takes over.

The voodoo2 Chipset won't handle 24 bit color in this config. Read
about the schips themselves. That's like trying to get a Pentium or
Pentium2 to do 64bit math.

On Thu, 2 Apr 1998 18:02:49 +0200, "Kai Anders Wold"
<trin...@online.no> wrote:

>From
>http://www.voodoooextreme.com :
>
>Building on its success in the high-end 3D graphics market, Quantum3D, Inc.
>has announced Obsidian2 , its next generation of realtime 3D graphics
>accelerator-marking the company's entry into the consumer market. Utilizing
>the Voodoo2 chipset from 3Dfx Interactive, Obsidian2 offers high-end
>features and realistic, smooth 3D performance at consumer price points.
>One of the first boards in the new line is the Obsidian2 S12 AGP-the first
>Voodoo2-based AGP 3D graphic accelerator.
>"With the Obsidian line we set the standard for 3D realtime graphics
>performance, including innovating two TMU single-board scan-line
>interleaving last year with our 100SB," said Ross Q. Smith, senior vice
>president and founder of Quantum3D.
>"Now with Obsidian2, we're taking the gloves off and entering the consumer
>market.
>In keeping with the high-end power of Obsidian, Obsidian2 offers gamers
>frame rates their Quake opponents could only dream of-at consumer prices!"
>Obsidian2 includes unique features such as (simultaneous RGB and NTSC/PAL?)


>"TV-out", 3D in a window, dual monitor support (is this relevant to gamers?)

>and digital video out. The boards also deliver up to XX frames-per-second in
>performance and true 24-bit color. Obsidian2 comes in two flavors, the
>Obsidian2S with AGP or PCI bus, and 12MB of total memory. Based on
>Quantum3D's pioneering the single-board two texture memory unit scan
>line-interleaving (SLI), the Obsidian2X series board, the X16, comes with a
>total of 16MB of on-board memory. The Obsidian2 consumer series starts at an
>estimated retail price $329, and comes with a gaming bundle including
>XXXXXX.
>"Hardware savvy customers have been anxiously awaiting the release of the
>Voodoo2 chipset. With the lower-end manufacturers fighting each other with 8
>and 12MB single processor designs, there hasn't been a clear standout in the
>Voodoo2 war-until now," said Kelt Reeves, president of Falcon Northwest.
>"The release of Quantum3D's 16MB, dual processor Obsidian2 will bring a new
>standard of performance to the market. We're proud to include the Obsidian2
>in our high-end systems, giving our customers the highest level of graphics
>performance ever offered outside of military simulators."
>The Obsidian2S series includes 4MB of pixel memory, and two 4MB texture
>memory units for a total of 12MB of on-board memory. Obsidian2S also
>supports SVGA monitor resolutions of up to (800x600, with z-buffering
>enabled and 24-bit color). The Obsidian2 X16 has 4MB of pixel memory and two
>6MB texture memory units. The X16 supports XGA monitor resolutions of up to
>(1024x768, with z-buffering enabled and 24-bit color). Key functions include
>edge anti-aliasing (Diamond has it too), single-pass trilinear or bilinear
>texture filtering for smoother visuals, per pixel LOD mip mapping, alpha
>blending and perspective correction? The Obsidian2 pass-through design is
>compatible with almost all 2D accelerators-and installation is easy.
>Pricing for the new Obsidian2 line starts at $329 for the Obsidian2 S12 PCI.
>The S12 with AGP bus is priced at $399. The ultra-fast X16, utilizing single
>board SLI, 16MB of memory and TV-out is $599. The consumer boards, along
>with Quantum3D's Obsidian200SB, also announced today, will be available
>directly from Quantum3D through their new Web Store and through official
>resellers starting next month.
>Quantum3D is also offering its existing customers a special rebate offer.
>With the purchase of a new Obsidian2 S12 PCI board, and proof-of-purchase
>and return of their original Obsidian product, customers can get rebates
>worth up to $329-or a free board. Obsidian 100SB-4440 owners can get the new
>Obsidian2 S12, PCI version, for free when they ship back their original
>Obsidian board. Those customers owning an Obsidian 100SB-4400 can earn $200
>back on their Obsidian S12 PCI purchase, and owners of the Obsidian 50-4440
>can get a $100 rebate.
>
>
>


Clayton Macleod

unread,
Apr 6, 1998, 3:00:00 AM4/6/98
to

Hell, a 386slc can do 128bit math
--
remove SPAM from my email to reply by email or I won't get it ;)

dsp <d...@newsguy.com> wrote in article
<352899af...@enews.newsguy.com>...

0 new messages