rec.autos.simulators

F1RS -- 3D card

Roy Harringto

F1RS -- 3D card

by Roy Harringto » Thu, 26 Feb 1998 04:00:00


> Yeah the other guys are right.  Any card with  "3dfx" (ie. utilizes the
> voodoo chipset) is an excellent choice.

>How does this game run with the Screamin' 3d card?

Roy
John Walla

F1RS -- 3D card

by John Walla » Thu, 26 Feb 1998 04:00:00

On Tue, 24 Feb 1998 15:46:08 -0600, "Greg Cisko"


>I would watch that and remember how people in the NG
>said F1RS looked cartoonish, and thought "ya right". Then I found those
>people were doing D3D not Glide.

Not me - I still think it looks somewhat bright and garish, and I've
run it in Glide from day one.

To clarify, I do LIKE the graphics, but feel they could be better if
more muted and realistic.

Cheers!
John

David Schi

F1RS -- 3D card

by David Schi » Sat, 28 Feb 1998 04:00:00



Thought I'd give the results I got with a TSW wheel and pedals.  I
decided to measure the rpm so that I would at least understand what
the settings do on my system.

With accelerator sensitivity at 0% I get the same as what you stated,
essentially it is an on/off switch that has a transition near full
depression.

Win95   RPM vs Throttle depression
                Sensitivity
% Depression    50%     75%     100%
0               1600    1600    1600
25              1700    3200    4500
50              2800    5300    8000
75               3600   8600    11000
100             16500   16500   16500

The 100% Sensitivity setting is almost a straight line while the 50%
setting is almost linear up to 75% depression (at 3600 rpm) then rises
very rapidly from 75 to 100 depression at 16500 rpm.

Windows NT4 with SP3 and NT4DX5 patch
        RPM vs Throttle depression
                Sensitivity
% Depression    50%     75%     100%
0               2700    5100    8500
25              3500    7500    11000
50              5000    10500   13000
75              10500   13000   14500
100             16500   16500   16500

Here, with accelerator sensitivity at 100%, in neutral, the motor
idles at 8500 RPM.  When the brake is fully depressed, the rpm goes
down to 1600.  So the full movement of the accelerator is from 8500 to
the 16500 and thus is more sensitive (allows greater precision).
However, idling at 8500 is not satisfactory, since there is little
deacceleration when the throttle is let off.

If these numbers are plotted in a graph, the 75% sensitivity is almost
a straight line.

At 50% the line is very curved with the center much lower.  The slope
is much less  from 0 to 50% depression than from 50% to 100%, making
higher throttle adjustment more difficult than lower.

The 100% is slightly curved with the center higher.  The slope is
slightly more from 0 to 50% depression than from 50% to 100%, making
lower throttle adjustment slightly more difficult than higher.

The effect of the NT drivers that causes a higher idle rpm than the
1600 that I get in Win95, I think causes a reduced de-acceleration
when lifting off of the throttle, whereas in Win95, the transition
between acceleration and deacceleration is instantaneous like an
on/off switch.  Since it is very difficult to drive with a rpm offset
of even 5100, I'm forced to use the 50% sensitivity setting with a
2100 rpm idle.

Of course these results are subject to some measurement error, since I
just used a piece of cardboard placed along the accelerator pedal and
marked off 25% increments.  Also, other drivers, joystick ports, and
input devices would most likey yield different results.  Perhaps the
best course of action is to quantify the curves of each combination of
drivers and hardware and find one that offers the most linearity at
the point of acceleration most critical.

Dave.

Tilman Bo

F1RS -- 3D card

by Tilman Bo » Mon, 02 Mar 1998 04:00:00


John Wallace wrote on Wed, 25 Feb 1998 13:11:25 GMT:

  To  all  the  people  complaining about the `bright and cheery'
graphics: I initially thought it was  actually  pretty  dark  and
gloomy  (that's an understatement: the tunnel in Monaco was pitch
black, and the long straights in Spa  and  Hockenheim  shaded  by
trees felt like I was driving through the surrounding woods, lit-
erally) running in 85Hz. When I switched to 60Hz to get some more
performance on the grid, I was shocked to see how bright and low-
contrast everything was.  I tried setting the Gamma value both in
DOS by setting the relevant environment variables and in Windows'
display  properties  (this is a Pure 3D), but neither had any ef-
fect, as opposed to, say, Tomb Raider 2.  I  am  now  running  at
75Hz,  which  is  almost perfect as far as brightness goes, and I
can't perceive any real performance drop WRT 60Hz. (Just for  the
record,  setting the refresh rate works just fine *both* w/ envi-
ronment variables and from display properties. In  case  of  con-
flict, the environment variables seem to take precedence, and you
can't set anything other than some fixed values -- 60, 75, 85 and
120.)

  One thought though just hits me: Just  took  a  brief  look  at
sst1init.dll, and it seems to set many of  the  same  environment
variables  I'm  using  from  DOS. And since people have said that
F1RS will not start without this library in the system directory,
it  would seem it overrides any Gamma settings that may have been
done before. I mean how does this look:

sst1InitGammaRGB(): Setting GammaRGB = (%.2f,%.2f,%.2f)

OTOH, for some quantities, it's using the values  in  the  corre-
sponding  environment variables explicitly, while not even touch-
ing some others. There are references to sst_screenrefresh, but I
can't make any sense of them.

Anyway. I guess I'm rambling. Back to the point:

  Could  people  with this complaint try different screen refresh
rates and see if it makes any difference  for  them?  Also,  does
anyone have anything more concrete, like an impression of perfor-
mance hits at higher refresh rates, *for this particular game*? I
know the Quake-heads swear by 60Hz, but so far, none of the games
*I* have tried have shown any significant payoff for  sacrificing
my  eyes...  I'll  be sticking to 75Hz for now, as that is pretty
much where my eyes stop noticing screen flicker.  (Lucky  me,  my
cousin still complains at 80Hz, poor git.)

  And  on a related note: Does anyone know if the sst_grxclk set-
ting takes precedence over that in display properties,  like  the
refresh  rate?  I haven't bothered to take the chip's temperature
so far for comparison, and haven't noticed much of a  performance
boost from 50 to 55MHz.

  If  all this is old hat (I didn't see anything on it though), I
apologize for the wasted BW.

--
Cheers, Tilman
                          -=--=*=--=-
The axiom `Honest men have nothing to fear from the police' is
currently under review by the Axioms Appeals Board.
                                Terry Pratchett, _Men at Arms_

Tilman Bo

F1RS -- 3D card

by Tilman Bo » Mon, 02 Mar 1998 04:00:00


Kevin wrote on Tue, 24 Feb 1998 20:26:48 -0800:

  I think you need to see it in context. The Mystique is a pretty
old card.  At the time it came out, the tradeoffs they  made  IMO
were the right ones. Back then, a voodoo card was *much* more ex-
pensive, so the Mystique didn't really compete in the same class.
For  the  lower  price, you gotta sacrifice *something*, and they
made a conscious decision to leave out features,  and  in  return
make those they did implement fast. Which they were, at the time,
and for the price. If someone didn't agree  with  that  decision,
no-one  forced them to buy that card, Matrox never made a big se-
cret of the rationale behind it.

  Granted, the Mystique hasn't aged well, overtaken  by  tumbling
prices  for  cards  that did more, better, faster. Originally, it
was one of many possible options how to build a 3D card,  and  as
it  turned  out, it wasn't the best one. But historically, it was
still a sound decision -- hindsight is always 20/20, ya know.

But I agree, I wouldn't want to play F1RS without my Voodoo.

--
Cheers, Tilman
                          -=--=*=--=-
The axiom `Honest men have nothing to fear from the police' is
currently under review by the Axioms Appeals Board.
                                Terry Pratchett, _Men at Arms_

Ian Fir

F1RS -- 3D card

by Ian Fir » Mon, 02 Mar 1998 04:00:00



The Voodoo chipset does not support Gamma adjustments above 75Hz.

--
Regards,
Ian Firth
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Diversions Software - Game Development - http://www.divsoft.com
     Home of Prairie Dog Hunt PRO '97 - DS Sporting Clays

Richard Walk

F1RS -- 3D card

by Richard Walk » Tue, 03 Mar 1998 04:00:00



>  If  all this is old hat (I didn't see anything on it though), I
>apologize for the wasted BW.

What I think is your actual problem has been addressed before, but you
raise some interesting points, nevertheless <g>

To get Glide programs to recognise your gamma settings at the higher
refresh rates, you need to include the command

SST_VIDEO_24BPP=1

along with your gamma environment variables. btw - I use set
SST_GAMMA=1.7 at 85Hz refresh.

As for performance differences, I've never noticed any difference. Is
there any reason why there should be one? (I don't know either way).
Assuming that there isn't anything that makes it inherently slower
though, I would have thought that a higher refresh rate would actually
mean (on average) less time to wait between a frame being prepared and
the video card being able to send it to the screen - hence very slightly
better performance.

I know that using

FX_GLIDE_SWAP_INTERVAL=0
SST_SWAP_EN_WAIT_ON_VSYNC=0

you can force the card to draw the frames immediately, but I always seem
to see a small degredation in performance. Useful settings for
benchmarking though ;-)

Cheers,
Richard

Tilman Bo

F1RS -- 3D card

by Tilman Bo » Tue, 03 Mar 1998 04:00:00


Ian Firth wrote on Sun, 1 Mar 1998 22:11:33 -0700:



> >   If  all this is old hat (I didn't see anything on it though), I
> > apologize for the wasted BW.

> The Voodoo chipset does not support Gamma adjustments above 75Hz.

(I know that.)

  Sorry,  I didn't make this completely clear: Setting Gamma dif-
ferently, both by means of environment variables and by means  of
the  display  properties control panel, doesn't make a difference
*at any refresh rate*.  The brightness *is* different between re-
fresh  rates,  but  even at 60 or 75Hz it doesn't depend on Gamma
settings, at least for me, and at least so far.

--
Cheers, Tilman
                          -=--=*=--=-
The axiom `Honest men have nothing to fear from the police' is
currently under review by the Axioms Appeals Board.
                                Terry Pratchett, _Men at Arms_

Tilman Bo

F1RS -- 3D card

by Tilman Bo » Tue, 03 Mar 1998 04:00:00


Richard Walker wrote on Mon, 02 Mar 1998 07:51:25 GMT:

  Sorry, as I said, didn't see anything. Should've prolly checked
Deja News, which I'l be doing now.

I'll give that a try, thx. I thought it only enabled 24bpp mode.

  But to reiterate this one more time: I knew already that by de-
fault Gamma settings will be ignored at 85Hz, but it didn't  make
a difference at lower refresh rates, either. (As opposed to other
games, where I double-checked and it did make a difference.)

Neither have I.

  Neither do I. ;-) (I actually doubt it, but who knows...) I had
only seen on some Quake pages that I consulted for general Voodoo
info  that apparently in that particular gamers' community people
explicitly used 60Hz, presumably  for  performance  reasons  (why
else  would  one  bother).  It  almost seemed to me to be a given
*amongst Quake players* that this is necessary to  get  the  last
tenth of a frame per second out of it. (They're crazy folk, those
Quake-heads... No offense, but they're possessed with fps.) so  I
thought there might be something to it.

  Then  again,  the page that I remember off-hand saying so isn't
very reliable. They had some highly dubious info, and some of  it
was even outright wrong.

  This line of reasoning does have some merit, but the actual ef-
fect would to my mind depend on the ratio  between  refresh  rate
and  frame  rate,  the  otimum  being an integer ratio, don't you
think?

  Yeah,  I  also  get  conflicting  results  using these. The way
things are turning out, it looks like I'll need to  put  separate
items  in  my  boot  menu  even for each Windows game... I have a
whole submenu dedicated to games already, but so far it was  only
to acommodate for DOS games' varying needs.

Thanks for the reply, Richard.

--
Cheers, Tilman
                          -=--=*=--=-
The axiom `Honest men have nothing to fear from the police' is
currently under review by the Axioms Appeals Board.
                                Terry Pratchett, _Men at Arms_

Ian Fir

F1RS -- 3D card

by Ian Fir » Wed, 04 Mar 1998 04:00:00



I see.

Whenever I launch Quake2, I have to enter the Video settings, and hit
return once (without changing anything), to get gamma to adjust.  Maybe a
similar problem.

--
Regards,
Ian Firth
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Diversions Software - Game Development - http://www.divsoft.com
     Home of Prairie Dog Hunt PRO '97 - DS Sporting Clays

Richard Walk

F1RS -- 3D card

by Richard Walk » Wed, 04 Mar 1998 04:00:00



>> To get Glide programs to recognise your gamma settings at the higher
>> refresh rates, you need to include the command

>> SST_VIDEO_24BPP=1

>I'll give that a try, thx. I thought it only enabled 24bpp mode.

I don't know why it makes the gamma settings work, but it does - at least
for my Orchid R3D.

All I can say is that it *does* work for my machine / installation. I
know this because, like you, I found it horribly dark at 85Hz but could
adjust the gamma settings to get the same brightness as at 60Hz.

Maybe because one of the best players said that he used it. It's amazing
how such things can spread around a competitive community and quickly
become accepted as truth ;-)

Personally, I would never run anything at 60Hz these days - way too much
eye strain :(

Yep - but you need a game that uses constant frame rates or that ratio is
going to go all over the place. There are some merits in a GP2-like fixed
framerate model but it also has some significant disadvanatages for a
sim.

You don't need to do that! For Glide games like F1RS you can do
everything via a .bat file <g> I have created a F1RS.BAT which contains
the following:

set FX_GLIDE_NO_SPLASH=1
set SST_FASTPCIRD=1
REM set FX_GLIDE_SWAP_INTERVAL=0
REM set SST_SWAP_EN_WAIT_ON_VSYNC=0
set SST_FASTMEM=1
set SST_GAMMA=1.7
set SST_GRXCLK=55
set SST_VIDEO_24BPP=1
set SST_SCREENREFRESH=85
call C:\WINDOWS\UbiSoft\UBISETUP.exe -play F1RS

This sets the environment variables independently of what is in your
autoexec.bat or where ever. This way you can easily have separate
settings for each Glide game. They also get re-initialised everytime you
run it so it makes it easy to experiment with different settings - no
need to reboot ;-)

Cheers,
Richard


rec.autos.simulators is a usenet newsgroup formed in December, 1993. As this group was always unmoderated there may be some spam or off topic articles included. Some links do point back to racesimcentral.net as we could not validate the original address. Please report any pages that you believe warrant deletion from this archive (include the link in your email). RaceSimCentral.net is in no way responsible and does not endorse any of the content herein.