rec.autos.simulators

by **Rodney Arnd** » Tue, 21 Jan 2003 02:34:22

I seemed to be asked quite a bit about the frame rate hit or performance
quality of N2003. Hopefully I can answer most of those questions with this
one post so I don't have to reply to 100 emails <g>.

I've been asked this many times already & it seems as if it's a popular
topic, so hopefully I'll be able to answer most every ones questions with
the following...

Obviously a lot depends on the track & what's being drawn around you. There
are a bunch of new graphic options that can really kill your frame rates.
But this doesn't make it a bad thing. There is no system on the market today
that will allow you to run with every graphic setting maxed & that will
produce acceptable frame rates. Everyone is going to have to make a decision
on what eye candy you prefer to have on versus how well you want your frame
rates. Maybe in a few years as computer & graphic cards improve, we'll be
able to turn on more graphic options. But nobody should expect to be able to
max all the graphics and expect to see frame rates of 80 90 100 or whatever
your goal is.

What's nice is the options are there. Over the years as we upgrade &
computers improve, we'll eventually have that ability to improve our
graphics which will improve the sim. I'm not saying it's bad now... not at
all. I've got my settings pretty much the same as I've had them set for
N2002 & the graphics still look better than before. I'm also getting pretty
much the same frame rates as before.

Here's what I'm currently running...

Alienware system P4 1.5 400mhz FSB 640MB memory with a geforce 4TI 4600.

Sim settings...

1600X1200X16 resolution OGL

42 cars drawn ahead 3 behind. Drawing distance 100% I've got all shadows
turned of except car shadows on the ground. I also have reflection on
structures turned off(For me personally I'd rather see the cars in front of
me than shadows or reflections on the track) There has to be some give &
take.. at least till I get my new Alienware system in 2 years <g>.

I have most of the other options set on, or on a medium quality level
including mirror detail.

At Atlanta sitting by myself at the end of pit road my frame rates are in
the low 80's from the***pit view. At various points around the track my
rates will go into the 100's (with no other cars drawn on the track)
Starting scratch in a race with 42 cars drawn ahead my frame rates will drop
to the mid 30's. Still very raceable, & they claim the human eye can't
detect anything quicker than 35 any ways.

Now at a smaller track where more has to be drawn my frame rates will not
start out as high. Also if your at a track with lot's of sun with solar
effects enabled you'll notice a hit as well. Every system will be different
& drivers will have different preferences. It's a matter of experimenting
with performance versus graphical enhancements.

I'm pleased with my performance. Yeah sure I'd like to be able to see all
those shadows & reflections, but technology isn't at that point yet. It's
nice to know that the papy has given us the ability to choose these options
and there will come a day when we'll be able to turn all that eye candy on
:)

--
www.sascar.com

by **Mitch_** » Tue, 21 Jan 2003 03:04:59

I think the phrase "human eye cant see over 35fps" is an urban legend.
Maybe it's correct for a motion picture with motion blur but not a video
game. Until the last few years I'd never had the hardware to get much over
35fps. Now that times have changed I can get 100+fps at times and I can
*REALLY* see a difference between 35fps and 100fps. The sweet spot for me
begins about 40fps.

Thanks for the info Rodney lookin forward to NR2k3 :-)

Mitch

by **Tony Kelle** » Tue, 21 Jan 2003 04:59:50

Probably depends somewhat on the individual too. I can notice a difference
between 35, and 45 fps, but from 45 on up I really see no visual difference.

TK

by **Joao Gonzale** » Mon, 20 Jan 2003 05:57:12

On the subject of frame rates: a friend told me there's no point in having a
FR that's better than your monitor's refresh rate, eg: my monitor does 85 Hz
so a frame rate over 85 fps doesn't make a difference because the monitor is
not drawing the frames fast enough.

It make sense. Any comments???

Regards,

Jo?o Gonzalez

> Probably depends somewhat on the individual too. I can notice a difference
> between 35, and 45 fps, but from 45 on up I really see no visual
difference.

> TK

> > I think the phrase "human eye cant see over 35fps" is an urban legend.
> > Maybe it's correct for a motion picture with motion blur but not a video
> > game. Until the last few years I'd never had the hardware to get much
> over
> > 35fps. Now that times have changed I can get 100+fps at times and I can
> > *REALLY* see a difference between 35fps and 100fps.

by **Dave Henri** » Tue, 21 Jan 2003 06:19:02

> On the subject of frame rates: a friend told me there's no point in
> having a FR that's better than your monitor's refresh rate, eg: my
> monitor does 85 Hz so a frame rate over 85 fps doesn't make a
> difference because the monitor is not drawing the frames fast enough.

> It make sense. Any comments???

> Regards,

> Jo?o Gonzalez

>> Probably depends somewhat on the individual too. I can notice a
>> difference between 35, and 45 fps, but from 45 on up I really see no
>> visual
> difference.

>> TK

>> > I think the phrase "human eye cant see over 35fps" is an urban
>> > legend. Maybe it's correct for a motion picture with motion blur
>> > but not a video game. Until the last few years I'd never had the
>> > hardware to get much
>> over
>> > 35fps. Now that times have changed I can get 100+fps at times and
>> > I can *REALLY* see a difference between 35fps and 100fps.

I don't know how it applies to the Papyrus sims, but other titles, like
the EA F1 series, definately see a handling improvement as the framerate
climbs. Two advantages to high framerates, are the physics seem be sampled
more and if your system is loafing getting high framerates, then the chance
of the fps bogging down is reduced as well.
Since both titles, Papyrus' N2k2 and ISI's F1 2k2 have frame rates that
top out near 144-150 fps. The closer you can get to the max framerate, the
better the handling will remain.
dave henrie

by **Olav Malmi** » Tue, 21 Jan 2003 07:29:27

> Probably depends somewhat on the individual too. I can notice a difference
> between 35, and 45 fps, but from 45 on up I really see no visual difference.

There is also the issue of varying framerates during play. I don't
think many of us are annoyed by GPL's 36 since it's rock solid. A
framerate that varies between 40 and 70 is a lot worse IMHO.

And as Mitch said, that humaen eye 30 fps thing is indeed a urban
legend as long as no motion blur is used. I wonder when this will be
properly implemented into sims...

> TK

> > I think the phrase "human eye cant see over 35fps" is an urban legend.
> > Maybe it's correct for a motion picture with motion blur but not a video
> > game. Until the last few years I'd never had the hardware to get much
> over
> > 35fps. Now that times have changed I can get 100+fps at times and I can
> > *REALLY* see a difference between 35fps and 100fps.

--
Olav Malmin
remove .spam when replying

by **MadDAW** » Tue, 21 Jan 2003 20:54:46

I know that in the Ratbag games the physics sample rates are tied to the
frame rate, but the over all sample rate isn't suppose to change. Now don't
ask me why they picked 57 but thats the magic number for the physics sample
rate. The way they have it set up is based on frame rate. So if your running
right at 57 fps you would have one sample per frame, which would be ideal.
If your running at 28.5 fps you would have 2 samples per frame, 114 fps then
2 samples you would 114, and so on. Now the issue comes in when you don't
have that steady frame rate. Without a steady frame rate you end up getting
vary rates of physics updates, and can come up with a few more or few less
updates than the 57 mark.

MadDAWG

by **Larr** » Thu, 23 Jan 2003 03:08:03

Has anyone determined what hurts the most and what hurts the least, and what
comes in the middle ?

Larry

by **frederickso** » Thu, 23 Jan 2003 03:14:17

That second group of shadow options are the real killers, but they sure do
look cool....

by **Dave Henri** » Sat, 25 Jan 2003 23:10:43

I saw this explanation earlier and it still confuses me. I understand
the 1 sample per frame at 57fps. If I read it right, if you get 114fps you
get 2 samples per frame. But the lower framerate #'s don't make sense. at
28.5fps do you mean you get ONE sample per every TWO frames?
dave henrie

by **MadDAW** » Sun, 26 Jan 2003 01:06:55

At 114 you would have two frames per sample and at 28.5 frames you have just
the opposite, two samples per frame. My bet is that the way actual frame
rates jump around they probably never hit the exact 57 samples. One second
it may be 56 and the next it might be 58, maybe even greater deviations..
Which if that's the case it could explain the effect of frame rates on the
driving. Makes you wonder about the frame rate caps in the earlier Papy
games don't it?

MadDAWG

by **Haqsa** » Sun, 26 Jan 2003 08:47:16

If it's like other games that do asynchronous physics, the graphics frames
are interpolated between the physics frames. They are never going to be
exactly in sync, they are always interpolated. The closer together the
physics frames are (higher frequency physics) the less interpolation error,
among other things. But frame rate will not affect physics, that's the
whole point of doing it that way. What it affects is the perception of the
physics. Lower frame rate affects your ability to perceive, predict, and
react to motion, and therefore increases the chance of driver error.

by **Eldre** » Tue, 28 Jan 2003 01:20:52

>A
>framerate that varies between 40 and 70 is a lot worse IMHO.

Been there, done that - with N2002. :-(

Eldred
--
Homepage - http://www.umich.edu/~epickett
GPLRank:-0.381
N2002 Rank:+17.59

Never argue with an idiot. He brings you down to his level, then beats you
with experience...
Remove SPAM-OFF to reply.

rec.autos.simulators

N2003 frame rates

N2003 frame rates

N2003 frame rates

N2003 frame rates

N2003 frame rates

N2003 frame rates

N2003 frame rates

N2003 frame rates

N2003 frame rates

N2003 frame rates

N2003 frame rates

N2003 frame rates

N2003 frame rates

N2003 frame rates