rec.autos.simulators

Dave

I've got so many games I still have to check on the 8500, N4 being one of
them. I'll install N4 today and let you know.

maybe you could run the short little OpenGL test in N4 I asked for a while
ago in the Nvidia NG?

***********************************
Gentlemen,

all the info regarding GF/Radeon is confusing, but I am really interested in
finding out what's better for me.

Could I politely ask an owner of a Radeon8500, preferrably on an AthlonXP
1800+ or similar system, to run the following test under OpenGL for me:

Nascar 4 (by Papyrus) run in OpenGL
Resolution 1280 x 960 (16 bit colours)
All graphics detail and quality settings ON and set to maximum, except
- Specular Highlights OFF
- Mirror viewing distance only 50% (but front 100%)
16 sounds heard
3D sound and EAX both ON
4xFSAA (please _not_ Smoothvision)
Anisotrophy = 4 (32 tap)
Multitexture forced ON
VSync forced OFF

If you wish to overclock your system, please do so, but tell me what speeds
you had it overclocked to.

Could I ask you to please go to Watkins Glen, the pits, hit the brake as
soon as your car gets placed on the track, and see what fps you get there
while standing still?
Then, drive out of the pits, keep a speed of around 80mph, and see the fps
you get when you enter the first
righthander (the first of the Esses).
Then continue to the inner loop and see
what fps you have at the 500 yards sign (on the left of the track), then
continue through the outer loop and check which is the lowest fps you get
after leaving the outer loop (happens within 200 yards from the outer loop),
and then again tell me which fps you get as you enter and leave T11 (the
final righthander)?

I don't want to be impertinent by asking this, I know I'm asking quite a
lot. But I'd be very grateful if someone could do this test run. I'm sure
there'd be others interested, too. From all the messages it's really
difficult to find out which of the two cards (GF3Ti500 and Radeon8500) is
the better card for me.
*****************************

TIA

Achim

OK, I ran the test per your instructions, but I didn't take note of every
place that you asked. Instead I have just taken note of the minimum and
maximum frame rates. I didn't overclcock my cpu (1.33ghz) either because
that requires me to change a dipswitch on the mb to change the clock
multiplier to 11x, I don't oc the FSB so forget about me doing that. I can
think of three things that make this test invalid, I'm using WinXP and the
frame rate can vary compared to the same game on Win98, I'm using immature
video drivers compared to Nvidia's unified drivers that have been tweaked
over quite a few revisions, and finally, I had Force Feedback on which
affects the frame rate. Personally, I think your settings are unrealistic
for any video card, the FSAA level you selected is for resolutions 1024x768
and lower, when you change the settings in the ATI control panel it even
tells you the max res you should use that setting at. FSAA kills the frame
rate, as does anisotropic filtering. Smoothvision is not an on/off thing, a
4X Quality setting instead of 4x Performance setting is still Smoothvision.
Smoothvision is just the ATI trademark for their version of FSAA. I used 4X
Quality, but could have even gone up to 6X if I had wanted. As for
anisotropic filtering, it is not selected by 4tap etc. There are two
possible settings for this in the ATI Opengl drivers, high and highest
quality, I used highest. I used all the settings in Nascar4 that you
requested, although I don't see why you would want to use 50% distance in
mirrors or 100% in front. I find for myself that 60% in front and 15-20%
mirrors is plenty. With your settings I got from 35 to 43 fps. With the
settings that I use to play the game I got 35 to 85 fps. But, that is with
vsync on and my monitor refresh is set to 85 so that is the max possible. I
did have it off once and was at Daytona and saw just over 90fps with no
other cars on the track. I'm not too concerned with max frame rate myself
and find minimum frame rate more important. If you can achieve a good
minimum  frame rate while at the worst possible load then that is more
important to experiencing a smooth drive. I don't think the settings you
aksed me to test would give a stasfactory experience when in the middle of
the pack at WG. The settings I use do. I expect you would problably get a
better frame rate with a GF3 Ti 500 using Quincunx instead of quality
settings, at least until ATI improves their drivers, if they ever do.

thanks a lot for doing the test, and for honestly reporting your results.

Please be assured that I didn't make up these settings just for fun, the
settings I asked you to use are precisely the ones I use. I also use FF, but
forgot to mention it.

The reason I asked you to check fps at certain track locations was just to
be able to compare a little more precisely. Your min/max does that too, but
I I felt two or three more measurements can't hurt as they give a better
impression of the average fps around the track.

As for WinXP, it wasn't like this with older Nvidia drivers, but the newer
ones have the same fps in both WinME and W2K on my system. Strange but true.
Although I keep having the suspicion that under W2K, the graphics quality is
slightly inferior (just like the sound quality) despite perfectly identical
settings everywhere. Hence, your XP test is precisely the information I
wanted, thank you for that!

I find that together with Anisotrophy, 4xFSAA does slightly improve the
image quality even at 1280x960 when approaching lines (white lines, top edge
of armco etc.) at certain angles. That's why I use it. It isn't much, but I
really like to have max quality.

The Quincunx FSAA is visually inferior to 4xFSAA, which is why I prefer the
latter over the former.

GF offers 3, and I use the medium setting, which is 32 tap, actually.

Just the settings I use. I could have repeated the test with any settings
you preferred to use as well, but thought it saves you (or whoever would run
the test for me) work and time if I specify the settings, because this way,
you/he don't have to type out all your/his settings in the NG message.

Same here. Minimum and average fps are far more important than max fps, and
this is one of the reasons why I asked you to take several fps samples along
an entire lap on that track.

They're the settings I use both online and offline, honestly.

Certainly, but as I said, it's inferior in quality to 4xFSAA.

Thanks a lot for running the test for me. I really appreciate you doing it!

Achim

Ok, no prob. I did start out taking various readings at the locations, but
got confused as to just exactly where some of them where. I should say that
minimium frame rate was just brief and the average was closer to the maximum
frame rate. I'm a believer of frame rate is king still and quite often turn
off features that I know kill the fill rate of a video card. I play a lot of
flight sims and the frame rates can really get low during a fur ball so find
it best to turn off FSAA etc. It might look good and run smooth, but as soon
as you get in that furball all the FSAA in the world means nothing and
keeping up the frame rate is more important.

Yes, especially in driving sims. Bad framerates make the car harder to
control.

Achim

 still and quite often turn