Re: F1CE2000: small request

I think you're off-base with your number-based approach to analyzing the relative merits of F1CS2K -- the values may be the same, but the cars FEEL much different than they did in F1 2000. My guess is that the error that existed in F1 2000's graphics engine was contributing to the "slow" feel, and that these values are correct for the properly- optimized graphics engine in F1CS2K. Virtually everyone who has F1CS2K thinks that the cars feel and handle better than in F1 2000 (including a growing throng of GP3 converts), so crunch your numbers all you want, but meanwhile the rest of us will be enjoying modern F1 nirvana. ;-) -- JB In article <[email hidden]>, Gregor Veble <[email hidden]> wrote:

Quoted text (13 lines)

> Leviticus wrote: > > > > inertia=(2000.00,2500.00,1000.00) > > engine_inertia=(0.190) > > stiffness=(40.0) > > > > Thanks for the prompt response! > > Too bad that the values are just as I feared they would be, meaning > unchanged from the F1 2000 version. The values for inertia are way too > high (they are in the standard SI units), while the tire stiffness > appears to be too low, both of these factors leading to slow response

of

Quoted text (10 lines)

> the vehicle. This makes the cars easier to control (gives them more > 'feel'), but is hardly representative of an F1 vehicle. > > Those values can be changed in F1 2000, but it then creates havoc with > the physics/AI engine. I have little doubt the same would happen in > F1CE2k, as the physics engine was,based on the above data, most likely > not altered a lot. > > -Gregor >

Sent via Deja.com http://www.deja.com/ Before you buy.

In article <[email hidden]>, [email hidden] (Andre Warringa) wrote:

Quoted text (7 lines)

> On Thu, 07 Dec 2000 09:13:47 +0100, Gregor Veble > <[email hidden]> wrote: > > >Hi all, > > > >I'm considering buying F1CE2k, but would first like to see if some > >aspects were tweaked/fixed. What I'd like to ask the owners of the

new

Quoted text (13 lines)

> >version is to go to the file 'default.vhs', if it still exists (it > >should be a plain text file), from the new version (F1CS2k), and say > >what values are stored under the inertia, engine inertia and tire > >stiffness. > > > >Thanks in advance for any replies, > > > >-Gregor > > Please buy it, I sure would like to know your opinion about the > physics. I, and most of us I think, can only judge a physics model by > how it feels and how the cars react compared to what we seem on TV. > You seem to really -know- about this stuff :)

You're considerably more diplomatic than I am, Andre, but I second your plea that Gregor should buy F1CS2K -- it DOES feel significantly different than F1 2000, IMO. VERY similar, but more "crisp" and less "heavy." For example, using the EXACT same Indy setup that I used in F1 2000 to achieve lap times in the 1:11.xx range, I've only been able to get down into the 1:14.xx range with F1CS2K, but it's not because anything feels worse -- just DIFFERENT. More edgy, perhaps, which forces me to drive a little more gingerly, as in GPL. Or GPL with wings . . . <G> -- JB Sent via Deja.com http://www.deja.com/ Before you buy.

Hi all, I'm considering buying F1CE2k, but would first like to see if some aspects were tweaked/fixed. What I'd like to ask the owners of the new version is to go to the file 'default.vhs', if it still exists (it should be a plain text file), from the new version (F1CS2k), and say what values are stored under the inertia, engine inertia and tire stiffness. Thanks in advance for any replies, -Gregor

inertia=(2000.00,2500.00,1000.00) engine_inertia=(0.190) stiffness=(40.0) On Thu, 07 Dec 2000 09:13:47 +0100, Gregor Veble <[email hidden]> wrote:

Quoted text (12 lines)

>Hi all, > >I'm considering buying F1CE2k, but would first like to see if some >aspects were tweaked/fixed. What I'd like to ask the owners of the new >version is to go to the file 'default.vhs', if it still exists (it >should be a plain text file), from the new version (F1CS2k), and say >what values are stored under the inertia, engine inertia and tire >stiffness. > >Thanks in advance for any replies, > >-Gregor

Leviticus wrote: > > inertia=(2000.00,2500.00,1000.00) > engine_inertia=(0.190) > stiffness=(40.0) > Thanks for the prompt response! Too bad that the values are just as I feared they would be, meaning unchanged from the F1 2000 version. The values for inertia are way too high (they are in the standard SI units), while the tire stiffness appears to be too low, both of these factors leading to slow response of the vehicle. This makes the cars easier to control (gives them more 'feel'), but is hardly representative of an F1 vehicle. Those values can be changed in F1 2000, but it then creates havoc with the physics/AI engine. I have little doubt the same would happen in F1CE2k, as the physics engine was,based on the above data, most likely not altered a lot. -Gregor

On Thu, 07 Dec 2000 09:13:47 +0100, Gregor Veble <[email hidden]> wrote:

Quoted text (12 lines)

>Hi all, > >I'm considering buying F1CE2k, but would first like to see if some >aspects were tweaked/fixed. What I'd like to ask the owners of the new >version is to go to the file 'default.vhs', if it still exists (it >should be a plain text file), from the new version (F1CS2k), and say >what values are stored under the inertia, engine inertia and tire >stiffness. > >Thanks in advance for any replies, > >-Gregor

Please uy it, I sure would like to know your opinion about the physics. I, and most of us I think, can only judge a physics model by how it feels and how the cars react compared to what we seem on TV. You seem to really -know- about this stuff :) Andre

On Thu, 07 Dec 2000 09:48:23 +0100, Gregor Veble <[email hidden]> wrote: >Leviticus wrote: >> >> inertia=(2000.00,2500.00,1000.00) >> engine_inertia=(0.190) >> stiffness=(40.0)... >Too bad that the values are just as I feared they would be, meaning >unchanged from the F1 2000 version. The values for inertia are way too >high (they are in the standard SI units), while the tire stiffness >appears to be too low, both of these factors leading to slow response of >the vehicle. Do you know in what unit the stiffness is given? And the intertia, that's XYZ axes for sure, but do you know the directions of these axes? I normally use OpenGL's convention of Y+ = up, Z+ is out the screen, X+ is to the right (righthanded). (I'm creating an importer for SCGT, uhm, ISI cars for my own sim, so I'd like to know how to import the .VEH best) In fact, regarding stiffness of tires, I'm using a spring rate (apart from the suspension ofcourse) to do this (which *should* vary with tire pressure). > This makes the cars easier to control (gives them more >'feel'), but is hardly representative of an F1 vehicle. What would be better values in your opinion? Ruud van Gaal, GPL Rank +53.25 MarketGraph : http://www.marketgraph.nl Pencil art : http://www.marketgraph.nl/gallery/ Car simulation: http://www.marketgraph.nl/gallery/racer/

Ruud van Gaal wrote:

Quoted text (11 lines)

> > On Thu, 07 Dec 2000 09:48:23 +0100, Gregor Veble > <[email hidden]> wrote: > > >Leviticus wrote: > >> > >> inertia=(2000.00,2500.00,1000.00) > >> engine_inertia=(0.190) > >> stiffness=(40.0) > > Do you know in what unit the stiffness is given?

No, that one I still haven't quite figured out. It should be in force per angle units, as it represents the force per slip angle in the linear regime of the tire (not the flexibility of the tire, at least not according to the usual definition), but it all depends on how the tire model is done. Still, the cars slide at way too high slip angles when cornering at the limit. Setting the value to 60 or even 80 in F1 2000 produces a much more lively and responsive experience (and introduces the full lock understeer). > And the intertia, that's XYZ axes for sure, but do you know the > directions of these axes? I normally use OpenGL's convention of Y+ = > up, Z+ is out the screen, X+ is to the right (righthanded). > (I'm creating an importer for SCGT, uhm, ISI cars for my own sim, so > I'd like to know how to import the .VEH best) > For inertia it is no problem which way the axis is oriented (+ or -), as it's a (symmetric) 2nd rank tensor. The z value is definitely through nose-tail, as my experimenting with F1 2000 clearly showed that. I believe the x-axis being left-right and y being up-down, just as your guess was.

Quoted text (10 lines)

> > In fact, regarding stiffness of tires, I'm using a spring rate (apart > from the suspension ofcourse) to do this (which *should* vary with > tire pressure). > > > This makes the cars easier to control (gives them more > >'feel'), but is hardly representative of an F1 vehicle. > > What would be better values in your opinion? >

Inertia is way too high. Let's take the z-axis for example. An F1 car is limited to 1.8m of width, and the height is close to 1m. That means that the upper limit for the inertia around the z-axis for a 600kg car shouldn't exceed 600 kg m^2 (if all the mass was concentrated about 1m from the center of gravity, which it isn't). With guessing the distribution of weigth in an F1 car, a more realistic value would be about 200 kg m^2 for the z - value. As the length of the F1 car is about 4m, the inertia values for x and y axis should be around 800 to 1000. All the estimates are done with the mass being approximately uniformly distributed over the whole car. The actual values should probably be lower still, as the mass is most likely concentrated much closer to the center of gravity. The engine inertia and the tires stiffness are harder to estimate, but clearly the engine doesn't want to stop revving once in a spin, which keeps the rear tires spinning as well, making a car do severe donuts even with throttle off in a slow-medium speed spin. I've found the best car behaviour with tire stiffness from 60-80. Halving the engine inertia makes for a more responsive experience as well. While the car becomes a real joy to drive with these settings, the problem is that the physics code for non-player cars becomes unstable with such settings (probably the timestep in calculations is shorter for the player car), and so does the AI - at least in F1 2000. I do hope that this isn't true in the new version, but it seems to me that the physics portion was largely left intact. BTW, does anyone know how to access the vehicle files in SCGT? I think increasing the tire stiffness there would make the experience even better. -Gregor

[email hidden] wrote:

Quoted text (8 lines)

> > I think you're off-base with your number-based approach to analyzing > the relative merits of F1CS2K -- the values may be the same, but the > cars FEEL much different than they did in F1 2000. My guess is that > the error that existed in F1 2000's graphics engine was contributing to > the "slow" feel, and that these values are correct for the properly- > optimized graphics engine in F1CS2K. >

Well, I think different. You wouldn't believe how long I tried to tweak F1 2000 to make it feel better and more realistic, while at the same time allowing the AI to be able to drive it. It included tweaking the cars and the AI setups. I've thrown in the towel after weeks of trying. If inertia is wrong, it is wrong. It's a physics quantity and if F1CS2k truly is a sim, it should reflect the wrong inertia in the driving feel. And if the sim cannot be tweaked the way I want it to, I'll pass on it. I do wish you all the enjoyment you can get out of it. -Gregor

So what exactly is the purpose of this thread anyway? -- [email hidden] "Gregor Veble" <[email hidden]> wrote in message news:[email hidden]... > Leviticus wrote: > > > > inertia=(2000.00,2500.00,1000.00) > > engine_inertia=(0.190) > > stiffness=(40.0) > > > > Thanks for the prompt response! > > Too bad that the values are just as I feared they would be, meaning > unchanged from the F1 2000 version. The values for inertia are way too > high (they are in the standard SI units), while the tire stiffness > appears to be too low, both of these factors leading to slow response of > the vehicle. This makes the cars easier to control (gives them more > 'feel'), but is hardly representative of an F1 vehicle. > > Those values can be changed in F1 2000, but it then creates havoc with > the physics/AI engine. I have little doubt the same would happen in > F1CE2k, as the physics engine was,based on the above data, most likely > not altered a lot. > > -Gregor

Greg Cisko wrote:

Quoted text (7 lines)

> > So what exactly is the purpose of this thread anyway? > > -- > > [email hidden] >

My own way of finding out whether it's worth purchasing F1CS2k :). -Gregor

Gregor, I've tried some of the physics tweaks for F1 2000 (do you post on the High Gear boards, by chance?), and they DID help, but so far, I'm VERY pleased with the way things feel in F1CS2K. In many sims, the inertia just doesn't "feel" right (GP3 comes to mind for me here, while GPL does feel right), but in F1CS2K, it does "feel" better to me. I'm not sure what calculations are going on in the background, or how accurate those numbers are, but it does translate nicely, and it does feel better than F1 2000. I'm actually a bit surprised that nobody has remarked on the subtle handling differences -- I think people are attributing most of the "new" feel to the graphics enhancements, and maybe that is what's creating the new feel for me, but the handling does feel different. Did I mention that the handling feels different to me, BTW? ;-) -- JB In article <[email hidden]>, Gregor Veble <[email hidden]> wrote:

Quoted text (13 lines)

> Greg Cisko wrote: > > > > So what exactly is the purpose of this thread anyway? > > > > -- > > > > [email hidden] > > > > My own way of finding out whether it's worth purchasing F1CS2k :). > > -Gregor >

Sent via Deja.com http://www.deja.com/ Before you buy.

[email hidden] wrote:

Quoted text (23 lines)

> > Gregor, > > I've tried some of the physics tweaks for F1 2000 (do you post on the > High Gear boards, by chance?), and they DID help, but so far, I'm VERY > pleased with the way things feel in F1CS2K. In many sims, the inertia > just doesn't "feel" right (GP3 comes to mind for me here, while GPL > does feel right), but in F1CS2K, it does "feel" better to me. I'm not > sure what calculations are going on in the background, or how accurate > those numbers are, but it does translate nicely, and it does feel > better than F1 2000. > > I'm actually a bit surprised that nobody has remarked on the subtle > handling differences -- I think people are attributing most of > the "new" feel to the graphics enhancements, and maybe that is what's > creating the new feel for me, but the handling does feel different. > > Did I mention that the handling feels different to me, BTW? > > ;-) > > -- JB >

Sure I do post at High Gear, under the handle Gecko (this one is from my flight simming). I am downloading the demo right now, and I'll try to tweak it as well. I'll report on it when I'm done. I do hope it works as I wish it to! -Gregor

In article <[email hidden]>, Gregor Veble <[email hidden]> wrote: > [email hidden] wrote: > > > > Gregor, > > > > I've tried some of the physics tweaks for F1 2000 (do you post onthe > > High Gear boards, by chance?), and they DID help, but so far, I'mVERY > > pleased with the way things feel in F1CS2K. In many sims, theinertia > > just doesn't "feel" right (GP3 comes to mind for me here, while GPL > > does feel right), but in F1CS2K, it does "feel" better to me. I'mnot > > sure what calculations are going on in the background, or howaccurate > > those numbers are, but it does translate nicely, and it does feel > > better than F1 2000. > > > > I'm actually a bit surprised that nobody has remarked on the subtle > > handling differences -- I think people are attributing most of > > the "new" feel to the graphics enhancements, and maybe that iswhat's

Quoted text (10 lines)

> > creating the new feel for me, but the handling does feel different. > > > > Did I mention that the handling feels different to me, BTW? > > > > ;-) > > > > -- JB > > > > Sure I do post at High Gear, under the handle Gecko (this one is from

my > flight simming). > > I am downloading the demo right now, and I'll try to tweak it as well. > I'll report on it when I'm done. I do hope it works as I wish it to! Cool! Please report back on your findings/feelings. BTW, I was using YOUR physics tweaks on F1 2000, BTW! ;-) -- JB Sent via Deja.com http://www.deja.com/ Before you buy.

On Thu, 07 Dec 2000 13:11:00 +0100, Gregor Veble <[email hidden]> wrote:

Quoted text (12 lines)

>> >> inertia=(2000.00,2500.00,1000.00) >> >> engine_inertia=(0.190) >> >> stiffness=(40.0) >> >> Do you know in what unit the stiffness is given? > >No, that one I still haven't quite figured out. It should be in force >per angle units, as it represents the force per slip angle in the linear >regime of the tire (not the flexibility of the tire, at least not >according to the usual definition), but it all depends on how the tire >model is done. Still, the cars slide at way too high slip angles when >cornering at the limit.

Would the lateral force be then: forceLat=stiffness*slipAngle (angle in degrees)? In that case, stiffness would seem low indeed. I use a curve to do lateral forces from tires, as things get so different from linear when approaching limits (besides, it's much easier that way for me). ... >For inertia it is no problem which way the axis is oriented (+ or -), as >it's a (symmetric) 2nd rank tensor. Ah yes, ofcourse. It's more for the centers I think, where the wing centers are located. > The z value is definitely through >nose-tail, as my experimenting with F1 2000 clearly showed that. I >believe the x-axis being left-right and y being up-down, just as your >guess was. Good, looks like ISI uses the OpenGL system then. Haven't had problems importing a car body (VRL file) wrt orientation, so that probably is right. ... >Inertia is way too high. Let's take the z-axis for example. An F1 car is >limited to 1.8m of width, and the height is close to 1m. That means that >the upper limit for the inertia around the z-axis for a 600kg car >shouldn't exceed 600 kg m^2 (if all the mass was concentrated about 1m >from the center of gravity, which it isn't). Ah yes, as a First-Sony inertia attempt I have inertia's that are equal to the mass of the car in XYZ directions in my sim, which seems to work relatively ok, although I'm not an actual driving wizard. Having inertia's greater than the car's mass would make it seem indeed that most of the mass is away from the center of gravity. ... >BTW, does anyone know how to access the vehicle files in SCGT? I think >increasing the tire stiffness there would make the experience even >better. Aren't all the .VEH files in some .MAS files, like VEHICLES.MAS? I believe there is a MAS file in the original distribution in which all the default cars are stored. The .VHS file is probably stored separately; haven't looked at it yet. Ruud van Gaal, GPL Rank +53.25 MarketGraph : http://www.marketgraph.nl Pencil art : http://www.marketgraph.nl/gallery/ Car simulation: http://www.marketgraph.nl/gallery/racer/

In article <[email hidden]>, Gregor Veble <[email hidden]> wrote: I'm with Gregor on this one, though it alone won't deter me from a purchase. He's correct in the math. And if the physics is wrong, but the "feel" is right, there's a kludge that makes at least me suspect. Admittedly, it might be there to overcome some other inaccuracies or inadequacies of the model/platform. Then again, maybe the subjective feel of a PC pilot isn't quite the last word in fidelity (not really a dig, afterall, how many of us do know what an $X million F1 car feels like?). Anyway, FWIW, I enjoyed the original, though I felt it to be too lethargic (an Elise I drove seemed much more responsive). Nevermind the horrid AI. I plan to take a look at the demo of #2. Greg? :-)

Quoted text (21 lines)

>[email hidden] wrote: >> >> I think you're off-base with your number-based approach to analyzing >> the relative merits of F1CS2K -- the values may be the same, but the >> cars FEEL much different than they did in F1 2000. My guess is that >> the error that existed in F1 2000's graphics engine was contributing to >> the "slow" feel, and that these values are correct for the properly- >> optimized graphics engine in F1CS2K. >> > >Well, I think different. You wouldn't believe how long I tried to tweak >F1 2000 to make it feel better and more realistic, while at the same >time allowing the AI to be able to drive it. It included tweaking the >cars and the AI setups. I've thrown in the towel after weeks of trying. > >If inertia is wrong, it is wrong. It's a physics quantity and if F1CS2k >truly is a sim, it should reflect the wrong inertia in the driving feel. >And if the sim cannot be tweaked the way I want it to, I'll pass on it. >I do wish you all the enjoyment you can get out of it. > >-Gregor

"Gregor Veble" <[email hidden]> wrote in message news:[email hidden]...

Quoted text (7 lines)

> Greg Cisko wrote: > > > > So what exactly is the purpose of this thread anyway? > > My own way of finding out whether it's worth purchasing F1CS2k :). > > -Gregor

OH! Thats why eh? I thought it was to make me feel totally inadequate, like the 4 years I spent drinking beer at the University had been wasted or something. I thought slip angle was how hi the hand got bef.....naw better not...family newsgroup and all.... dave henrie

"David Masten" <[email hidden]> wrote in message news:90peto$r7e$[email hidden]... > In article <[email hidden]>, Gregor Veble<[email hidden]> wrote: > > I'm with Gregor on this one, though it alone won't deter me from apurchase. > He's correct in the math. And if the physics is wrong, but the "feel" is > right, there's a kludge that makes at least me suspect. Admittedly, itmight > be there to overcome some other inaccuracies or inadequacies of the > model/platform. Then again, maybe the subjective feel of a PC pilot isn't > quite the last word in fidelity (not really a dig, afterall, how many ofus do > know what an $X million F1 car feels like?). Anyway, FWIW, I enjoyed the > original, though I felt it to be too lethargic (an Elise I drove seemedmuch > more responsive). Nevermind the horrid AI. I plan to take a look at thedemo > of #2. > > Greg? :-) Since you asked, I think his inertia queries are bogus and baseless. This is in all likely hood a fine game and to dismiss it because of some lame crap like that is laughable. Apparently F1 2000CS has weather (though not dynamic), better AI, improved tracks and a much better framerate than F1 2000. All up- graded in less than a year. All I can say about that id DUH :-) -- [email hidden]

Hi again, I managed to download the demo and test it a bit. As far as I see it, the physics changes haven't been all that dramatic. The cars do feel a bit different to drive, but that might have been me not playing F1 2000 in a while. Anyway, I tried to tweak the cars a bit. I did the same changes as in F1 2000, and it lead to about the same results, which is very good. The car behaviour is really good then, but the AI cars still have some problems with the mods. That is not all that much of a problem, as one can only adjust his or her own car that they like to drive towards more realistic settings, while leaving the other cars as they were, as the mods do not affect the grip nor power of the engine. Looking from the simulation point of view, much like F1 2000 before it, F1 CS 2000 is a great sim, if only one tweaks it a little. However, what unpleasantly surprised me was the force feedback implementation. See, I only got a FF wheel only recently, so I wasn't playing F1 2000 much with it before. However, switching from GP3 to F1 CS 2000 is a slight disappointment. The FF effects are there (be sure to loose the speed sensitivity option, though), but they act as a simple spring effect, with the coefficient of spring being dependent on the downforce, it seems. While when not driving on the limit this gives a rather good impression, on the limit it feels inadequate. It is all the more surprising as SCGT sports rather good FF effects. Therefore my verdict would be, if you don't have a FF wheel and are prepared to apply a little tweaking to the physics of your own car, I would say both F1 CS 2000 and GP3 are 'there' in terms of an F1 sim, and if you have a couple of friends you want to play an F1 sim with online (but not too many :) ), F1 CS 2000 would be the only choice. However, I believe FF users should look towards GP3 if they want to get most out of their hardware. I don't think you could go much wrong with any of these products, though. -Gregor [email hidden] wrote in message <90od8r$j4r$[email hidden]>...

Quoted text (12 lines)

> >Cool! Please report back on your findings/feelings. > >BTW, I was using YOUR physics tweaks on F1 2000, BTW! > >;-) > >-- JB > > >Sent via Deja.com http://www.deja.com/ >Before you buy.

You got it wrong, that's the 'slap' angle :). -Gregor Dave Henrie wrote in message ... > OH! Thats why eh? I thought it was to make me feel totally inadequate, >like the 4 years I spent drinking beer at the University had been wastedor >something. > I thought slip angle was how hi the hand got bef.....naw better >not...family >newsgroup and all.... >dave henrie >

[email hidden] wrote: > snip

Quoted text (7 lines)

> > Virtually everyone who has F1CS2K thinks that the cars feel and handle > better than in F1 2000 (including a growing throng of GP3 converts), so > crunch your numbers all you want, but meanwhile the rest of us will be > enjoying modern F1 nirvana. > > ;-)

Yepp, nirvana, that's the word I was looking for to describe F1??2k... ;) -- l8er ronny Your mouse has moved. Windows must be restarted for the change to take effect. Reboot now? |\ _,,,---,,_ I want to die like my Grandfather, ZZZzz /,`.-'`' -. ;-;;,_ in his sleep. |,4- ) )-,_. ,\ ( `'-' Not like the people in his car, '---''(_/--' `-'\_) screaming their heads off!