rec.autos.simulators

I have also not experienced any of the menu problems people have reported
with their TNT.  And also I noticed the fps doesn't change much from 800
resolution to 1280 resolution.  There's only a +5 to +8 fps increase at the
lower resolution.

I am definitely buying GPL now.  Way to go Papy!

GTIracer

Two things,

1. I run GPL with one voodoo II  (8mb) at 1024x768 res and it looks
great,all graphics on at 30fps(in car)more with two in SLI.

2.The colours on my voodoo don`t look washed out like people keep saying?( I
don`t want that f1rs arcade look),it looks great,very realistic.

The in car fps are what really count and with two voodoo II`s that is better
than TNT,so why pay for a TNT when GPL in SLI is better.
After all,you won`t see much difference from 1024 to 1280 and if the fps is
worse why bother with TNT?

The colors on my Voodo look like the contrast has been turned way up.  ( Is
this what is refered to as " Washed out " ?)

I cannot stay on the road with the inside car view,  I HAVE to use the
arcade view.  From there, the frame rates are simular ( 34> ).

Why pay more for one card when you can pay the same for three?  Well,
usually, this would sound like a good thing.  Except that PCI slots are at a
priemiem these days, so why use an additional 2 slots when I can use just
one AGP slot and get comparible frame rates and deeper color?  Made sense to
me.

Question : If you were just purchasing a new system, and you had to choose
between buying one card that did an excellent job of both 2 & 3D, or
purchasing 1 card for 2d, and another 2 for 3d, which way would you go?  You
will be graded for honesty...  :-)

BTW: In order to "test" GPL on my new system, I yanked the CL VD2 out of my
old P200 to see how it worked.  Impressive.  So I ordered a STB Black Magic
to sit alongside my STB TNT, to have the best of both worlds.  Then Papy
released the OGL patch for GPL.  I almost wish they hadn't did that.  Now I
have to come up with a reason to install the Black Magic when it arrives.
Decisions, decisions...

Again, that's a matter of opinion as to what's "better".  And us hard-core
types rarely make decisions based on economy or "value". If we were greatly
concerned about that, we'd run away from flying and driving sims and take up
chess and cards.  :-)

For right now, I think the V2/TNT combo's just fine.  But ask me again this
time next week.  :-)

Jerry Morelock

Try 1280x1024.. something voodoo SLI can never do.

They do.. try a side by side comparison and it will really look like it is.
Kinda like seeing the world through dirty glasses, then finally being able
to see without glasses all together.

Because with one TNT you get an awesome 2D card... an awesome 3D card.. and
an awesome AGP card.  With Voodoo 2 SLI you get.. alot of boards in your
computer generating too much heat for us overclockers... you get a big dent
in your wallet (as much as 400 bucks for 2 voodoos).. and you get the best
GLIDE only support.  But that's just one API...

You'll never know the difference no matter how many voodoo 2's you put in
your 'puter.  You see a major difference.. if you didn't, Papyrus wouldn't
have wasted their time with the OpenGL patch for the TNT cards..  THEY see a
difference, that's why they made the patch for us TNT owners.. so we could
see the difference too.

Be lucky that their aren't 2 AGP ports in my 'puter, otherwise Nvidia may
just set up TNT cards in SLI mode!!!  now THAT WOULD ROCK!!!!

- Eric

Here is what they responded:

<<<<snip>>>>

I'll respond to your concerns point by point, but will say you seemed to
completely misunderstood the benchmarks.

The numbers are high for the Voodoo2 because it was tested with 2 cards in
SLI mode. Plus, a Pentium II 450 was used.

As for 3Dfx coming to us to do the tests, it's not unfair. 3Dfx is a client
of ours and initially just wanted us to bless their own testing, but we
decided that it would have been best to do the testing ourselves. In fact, I
did the testing myself and can assure that no tweaking was done. By the way,
NVIDIA and others are our clients too, and they could ask us to do the same
testing for them, or doing something else.

As for the Glide API comparison, I totally agree and that's why we only
published the results on the detailed pages and not the main page. Since
3Dfx commissioned the test they decided they wanted to show Glide scores. I
can understand why they would want to ,but I agree that comparing Glide to
Direct3D is not really fair. That's why we also tested the Direct3D version
and posted the scores those next to each other. Check out the benchmark
pages on their site to see what they have to say about Glide being used in
GameGauge benchmark testing.

As for us being unfair in our benchmark testing, I totally disagree. We made
sure that each cards' latest drivers and bios were used for each card. I
downloaded them the day of the test myself. Also, as stated on our Benchmark
Platform and Testing Procedures page I partitioned, formatted, and installed
Win98 and the benchmark apps myself. After that I made an image of the drive
and restored that image after each board was installed into the machine. I
don't 3Dfx would have went through all this trouble if they did it
themselves. So I think we made the testing much more fair.

Voodoo2 SLI not using DirectX? Seems you're very confused about the
GameGauge testing programs. They all require DirectX 5 to run. Only one of
them actually supports the Glide API(which we already discussed above). I
used the latest DirectX 5 drivers off 3Dfx' own site to run the test. And
then I ran the tests again using their new DirectX 6 drivers. So the Voodoo2
SLI was using DirectX.

As for a TNT being faster than a Voodoo2 SLI, that's just not possible. Now
I could see a TNT being faster than one Voodoo2 card, but not 2 Voodoo2
cards in SLI. Even NVIDIA has stated that they're not faster than Vodoo2
SLI.

You don't what a GameGauge benchmark is? Now, I think I know why you're
having such hard problems understanding the testing. GameGauge benchmark
testing is a suite of games that Computer *** World came up with. They
thought that testing actual game performance is much better than relying on
benchmarks like ZDBOp's 3D WinBench and with to regards to current games I
would have to agree. So anyway, CGW got special versions of Turok, Forsaken,
Incoming and F22 from the game developers that would output an average frame
rate after the application was run. Then used the demos of GL Quake and
Quake 2 and ran timedemo on them. All the scores outputted are in
frame-per-second (FPS) and you total all the scores to get the final
GameGauge score. I was under the impression everybody knew what GameGauge
was and that the scores were in FPS. Evidently that's not case, but I do
think you're in the minority or I would have received a lot more email.

<<<<snip>>>>

The point is.. at 1024x768 resolution.. a voodoo 2 SLI is faster than a TNT
card.. BUT... these are tested on games that are orginally programmed FOR
the voodoo cards.. and any resolution higher than that leaves voodoo 2 SLI
out!!!   They were not originally intended for a TNT card... games that are
coming out now run better than even voodoo 2 SLI cards.. such as SHOGO (at
least for me) and most likely Trespasser as well.  Because they are built
using DirectX 6 which runs awesome on a TNT card.

Basically.. I'm getting tired of seeing tests done on games built around
GLIDE APIs.   Who in the hell started using Quake/ Forsaken/ Turok as test
bases anyways???? Why not do a test on FlightSim 98... or SHOGO?? Oh gee..
maybe because a friggin Voodoo 2 in DUAL SLI mode will STILL suck eggs
against anything with DirectX 6?

QUIT TESTING GLIDE GAMES AGAINST TNT CARDS!!!! IT'S NOT ACCURATE!!!

hear.. lemme test my 4x4 in the desert against your ferrari in the desert...
gee.. looks like the ferrari didn't make it.  the 4x4 is a faster vehicle
than a ferrari.  The desert tests prove it.   YEAH... RIGHT!!  That's what
GameGauge is doing.

get you through times of no woman better

Nafi Coker

<SNIP>

Ok, but it IS ENTIRELY ACCURATE.  If a game has BOTH Glide and Direct 3D
support, then the point is what card runs it better, Voodoo or something
else?  Why cripple the Voodoo cards by forcing them to use D3D?  You're
missing THE advantage of the Voodoo cards, that they have a native
programming capability and can do things better than Direct 3D can on the
same card.  On games that have only D3D support, then some other card may
very well run it better.  But if it's like, for example, F1RS that has both
D3D and Glide modes, with a Voodoo 2 I'm gonna pick Glide.  And with that
one Glide is better than the D3D version, so a test on that game would be
ENTIRELY ACCURATE to use the Glide mode on Voodoo 2 and D3D for everything
else, you're finding out which card runs the game better, the API used isn't
the point.  Forcing the Voodoo 2 to use the D3D version in that test would
be crippling it unfairly, since their is a way to get better performance on
that card.  Flightsim 98 only uses D3D, so the benchmark will also tell you
which card runs it better.  And if it's Jane's F15, then it's entirely
accurate to note that there just isn't a D3D version, so the Glide mode is
all there is.

The thing is, your other analogy (other than trying to belittle the Voodoo
cards while making the TNT the end-all be-all) is also demonstrating the the
validity of the GameGauge and refuting your argument.  If you take the 4x4
and the Ferrari to the desert, and take both to the racetrack, you'll notice
that the Ferrari outperforms the 4x4 at the track, BUT the 4x4 can still run
at the racetrack.  This is D3D, TNT runs it but so does the Voodoo 2 (ditto
with Open GL).  But take both to the desert, and the Ferrari is unable to
perform at all.  This is Glide.  If I want to play a Glide only game, then
the "Ferrari" is really nice and pretty, but totally useless.  Grasp this
concept and you'll figure out why people own Voodoo 2's, and aren't
impressed by a bunch of benchmarks saying card XYZ is 10% faster etc. etc.

What you want is benchmarks rigging the tests so that the advantages of the
Voodoo 2 aren't used, not a fair fight.

Ken

As to your game choices, I quite agree.  If I can't fly it or drive it,
forget it!