rec.autos.simulators

Damage Realism in GPL/N3

JG_

Damage Realism in GPL/N3

by JG_ » Thu, 10 Dec 1998 04:00:00

Hi All,

The one thing that bugs me about GPL, is the total lack of realism
when it comes to the damage model...

What I mean is, nothing else is canned, but the damage, since day one
of sims I guess, has been. In GPL I expect, with realistic damage, to
be able to bend the suspension up, if I were to hit the wall at 45degs
at 20mph. These care weren't that strong were they??

Here we are with the pinnacle of computer racing sims. and the damage
modelling seems as advanced as ICR1. Was it Coulthard at Adelaide who
crashed a Williams in the pit entry? Do this in GPL with realistic
damage and you may not even get a flat, heck people cheat at the
Lesmos at 180Mph.

Does anyone agree here? GPL's physics engine is awesome, every aspect
is modelled, tiny bit too much brake, the wheel locks, but only if XYZ
conditions are also true. I am no expert on this, but from all I have
read it is VERY realistic. However, when the car comes into contact
with anything, it's back to some very simple maths equations, and only
3 or 4 canned "types".

To me, and this is where I'll state me opinion, this does matter...
Maybe it doesn't for many, but I would like to know. I get the idea
having no visual clue that you have a flat is a pain, but I see it
more deeply as: why can't we knock the toe in, squash the front tyres
in toward the car body, like ICR2, but with greater calculation of how
far/bad that suspension gets crushed in. Have the wheel go all
wibbly-wobbly, then maybe drop off at a certain time later.

I appreciate with body damage, that in GPL it's hard to model crumple
zones. These things were pretty solid wern't they? I don't know really
so I won't go on, but this is very important I feel for N3.

Clips with the wall in N2 can have really dodgy impact sounds but F9
reveals nothing is damaged. A slightly lesser impact and you may be
surprised to hear the spotter start "looks like..." In N3, with NROS
and general online racing, I think it's really going to be necessary
for N3 to show us a damage model that allows as much simulation as the
car itself. Every little scrape will need to have an impact on the
body of a stocker, and this should play as much a part of a car's
speed in a race, as the calculations with the tyres and the road..

Thanks for listening,



-:http://www.racesimcentral.net/

DJacob

Damage Realism in GPL/N3

by DJacob » Thu, 10 Dec 1998 04:00:00

You crash?   Not me, the real fun is racing.
        Don

KPineb

Damage Realism in GPL/N3

by KPineb » Thu, 10 Dec 1998 04:00:00


writes:

I'm not a programmer, but what you're describing sounds like a lot of work to
me, and I'd rather have the sim now than at some time in the next century. ;-)

Besides, it's a RACING sim, not a crashing sim.

Regards

Pinebox

Michael E. Carve

Damage Realism in GPL/N3

by Michael E. Carve » Thu, 10 Dec 1998 04:00:00


% Hi All,

% The one thing that bugs me about GPL, is the total lack of realism
% when it comes to the damage model...

<snip>

I would assume that you are not driving in Pro Damage...

--
**************************** Michael E. Carver *************************
     Upside out, or inside down...False alarm the only game in town.

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=<[ /./.  [-  < ]>=-=-=-=-=-=-=-=-=-=-=-=-=-=

Ben Farnema

Damage Realism in GPL/N3

by Ben Farnema » Thu, 10 Dec 1998 04:00:00

You're right.  It's a simulator, it should simulate.  I can understand
wanting to put more attention to driving than crashing, but not to the
extent that the damage is inconsistant, or has a lack of effect on driving.
What about it, Papy?

>Hi All,

>The one thing that bugs me about GPL, is the total lack of realism
>when it comes to the damage model...

>What I mean is, nothing else is canned, but the damage, since day one
>of sims I guess, has been. In GPL I expect, with realistic damage, to
>be able to bend the suspension up, if I were to hit the wall at 45degs
>at 20mph. These care weren't that strong were they??

>Here we are with the pinnacle of computer racing sims. and the damage
>modelling seems as advanced as ICR1. Was it Coulthard at Adelaide who
>crashed a Williams in the pit entry? Do this in GPL with realistic
>damage and you may not even get a flat, heck people cheat at the
>Lesmos at 180Mph.

>Does anyone agree here? GPL's physics engine is awesome, every aspect
>is modelled, tiny bit too much brake, the wheel locks, but only if XYZ
>conditions are also true. I am no expert on this, but from all I have
>read it is VERY realistic. However, when the car comes into contact
>with anything, it's back to some very simple maths equations, and only
>3 or 4 canned "types".

>To me, and this is where I'll state me opinion, this does matter...
>Maybe it doesn't for many, but I would like to know. I get the idea
>having no visual clue that you have a flat is a pain, but I see it
>more deeply as: why can't we knock the toe in, squash the front tyres
>in toward the car body, like ICR2, but with greater calculation of how
>far/bad that suspension gets crushed in. Have the wheel go all
>wibbly-wobbly, then maybe drop off at a certain time later.

>I appreciate with body damage, that in GPL it's hard to model crumple
>zones. These things were pretty solid wern't they? I don't know really
>so I won't go on, but this is very important I feel for N3.

>Clips with the wall in N2 can have really dodgy impact sounds but F9
>reveals nothing is damaged. A slightly lesser impact and you may be
>surprised to hear the spotter start "looks like..." In N3, with NROS
>and general online racing, I think it's really going to be necessary
>for N3 to show us a damage model that allows as much simulation as the
>car itself. Every little scrape will need to have an impact on the
>body of a stocker, and this should play as much a part of a car's
>speed in a race, as the calculations with the tyres and the road..

>Thanks for listening,

>Jon.Guest


>-:http://www.simproject.com/ccg/

John Moor

Damage Realism in GPL/N3

by John Moor » Thu, 10 Dec 1998 04:00:00

I agree to a point, but for the most part it's ok. If I go over a curb
really hard it will knock the supsension out of whack, but I would like
some panel damage ala Viper.
JM

> Hi All,

> The one thing that bugs me about GPL, is the total lack of realism
> when it comes to the damage model...

> What I mean is, nothing else is canned, but the damage, since day one
> of sims I guess, has been. In GPL I expect, with realistic damage, to
> be able to bend the suspension up, if I were to hit the wall at 45degs
> at 20mph. These care weren't that strong were they??

> Here we are with the pinnacle of computer racing sims. and the damage
> modelling seems as advanced as ICR1. Was it Coulthard at Adelaide who
> crashed a Williams in the pit entry? Do this in GPL with realistic
> damage and you may not even get a flat, heck people cheat at the
> Lesmos at 180Mph.

> Does anyone agree here? GPL's physics engine is awesome, every aspect
> is modelled, tiny bit too much brake, the wheel locks, but only if XYZ
> conditions are also true. I am no expert on this, but from all I have
> read it is VERY realistic. However, when the car comes into contact
> with anything, it's back to some very simple maths equations, and only
> 3 or 4 canned "types".

> To me, and this is where I'll state me opinion, this does matter...
> Maybe it doesn't for many, but I would like to know. I get the idea
> having no visual clue that you have a flat is a pain, but I see it
> more deeply as: why can't we knock the toe in, squash the front tyres
> in toward the car body, like ICR2, but with greater calculation of how
> far/bad that suspension gets crushed in. Have the wheel go all
> wibbly-wobbly, then maybe drop off at a certain time later.

> I appreciate with body damage, that in GPL it's hard to model crumple
> zones. These things were pretty solid wern't they? I don't know really
> so I won't go on, but this is very important I feel for N3.

> Clips with the wall in N2 can have really dodgy impact sounds but F9
> reveals nothing is damaged. A slightly lesser impact and you may be
> surprised to hear the spotter start "looks like..." In N3, with NROS
> and general online racing, I think it's really going to be necessary
> for N3 to show us a damage model that allows as much simulation as the
> car itself. Every little scrape will need to have an impact on the
> body of a stocker, and this should play as much a part of a car's
> speed in a race, as the calculations with the tyres and the road..

> Thanks for listening,

> Jon.Guest


> -:http://www.simproject.com/ccg/

John Walla

Damage Realism in GPL/N3

by John Walla » Fri, 11 Dec 1998 04:00:00


>The one thing that bugs me about GPL, is the total lack of realism
>when it comes to the damage model...
>Does anyone agree here? GPL's physics engine is awesome, every aspect
>is modelled, tiny bit too much brake, the wheel locks, but only if XYZ
>conditions are also true.

Given unlimited processor power and unlimited development time I would
like to see realistic damage. Two things however ;

1) Coding such a thing will detract from development time (not a big
issue), but also from processor time (this is a big issue). I think we
will see realistic damage, but GPL was just a first step along the
road within the limits of what time and power allow. Where this will
go in future will be incredible.

2) I'm not Ayrton Senna and you're not Michael Schumacher. Racecars
are much stronger than your average car in general, but open wheelers
are not, and any contact will usually damage them pretty badly. Not
that many people would be able to drive a GPL car with realistic
damage for any length of time.

On top of that I'd say that GPL primarily simulates driving and
racing, and the damage is there not so much to be totally accurate but
enough to keep you honest by not ramming, grass-cutting or shifting
with your foot flat to the floor. Overall you're right, and the damage
isn't as ultra-realistic as other parts of the sim, but there are
justifiable reasons for this, at least IMO.

Cheers!
John

Mike Laske

Damage Realism in GPL/N3

by Mike Laske » Fri, 11 Dec 1998 04:00:00


>On top of that I'd say that GPL primarily simulates driving and
>racing, and the damage is there not so much to be totally accurate but
>enough to keep you honest by not ramming, grass-cutting or shifting
>with your foot flat to the floor. Overall you're right, and the damage
>isn't as ultra-realistic as other parts of the sim, but there are
>justifiable reasons for this, at least IMO.

To back up John's argument, I'd like to add that you also need to realise
the priorities when developing a game.  If you're designing a racing sim,
you want most of the development efforts to be channeled into the racing
experience itself, or if you're designing a combat flight simulation, you'd
want most of the development spent on the flight model, weapons, systems,
graphics, and AI.  What proportion of the game is spent driving and flying
vs the amount of time crashing.  The crashing is a very minor element, all
things considered and I don't know of any sim that has realistic damage
modelling.  You're right in suggesting that damage modelling isn't good
enough - I agree with you, but there are and always will be other aspects of
the title that are more important and will be developed further (unless
you're developing an impact simulator <g>).  What I would say is that every
designer and developer likes to improve on what has gone before them, so to
look on the bright side, the situation should always get better.

Mike.

Don Hancoc

Damage Realism in GPL/N3

by Don Hancoc » Fri, 11 Dec 1998 04:00:00


    Well, I agree the INTENT of your argument, but for the one point
highlighted above: damage modelling, IMO is VERY critical, because it
happens so frequently (to me, anyway <G>).  Every race driver certainly
knows that even MINOR toe-in damage or even a simple tiny piece of paper
over the grill, can certainly ruin your day.  I've watched races where the
2nd place car had to pit to take a paper cup off the grill!  I've also seen
people lose a race because a relatively minor mishap early in the race made
them slower and slower as time went on.
    I truly believe damage modelling should be HIGHLY developed. More so
than clouds (except for their temperature changing capabilities) or crowds
or even sounds!

Don Hancock
"Gunner"
ICQ 24045680

JG_

Damage Realism in GPL/N3

by JG_ » Sat, 12 Dec 1998 04:00:00



>Michael E. Carver mcarver, described by the Rev Spooner as a shining wit,
>said:

>>% The one thing that bugs me about GPL, is the total lack of realism
>>% when it comes to the damage model...

>><snip>

>>I would assume that you are not driving in Pro Damage...

>Assumptions are dangerous things <g> Except for online races, I always
>drive with "realistic" damage and whilst GPL can give some nice
>intermediate damage effects (bent suspension, damaged tyre, etc) I do
>generally agree with Jon that the damage is toned down from what it should
>be. Personally I wouldn't describe it as a "total lack of realism" but it's
>pretty much arbitrary whether some quite substantial impacts actually cause
>damage or not.

"Total lack" was too harsh, Mike & Richard. My point wasn't to say GPL
is any less, just that I feel that Papy are proceeding greatly with
the car physics, but are leaving the damage trailing behind <end pun>
To summarise my post, I *hope* that Papy can improve it, and that
given what they have, they could have made the "Realistic" mode of
damage, as "realistic" as the car physics are. This would make the
realistic mode a very elite afair, and IMO give total satisfaction if
completing a race in it. I expect I could string together a lot of
laps at many of the tracks with no sign of a spin, so the fact a tyre
will be ripped off after a small brush with a barrier, is only at the
back of my mind. This only goes for about 4 tracks, and not Kylami for
sure! As Richard points out.. ;)

It's more than a good point..... Yet all the same, if we could have a
realitic damage mode, just with the current values lowered a bit, I
would prolly wreck more Lotus's in training and have to escape to get
a new car, that I would be a lot more conservative.

Thanks!


-:http://www.simproject.com/ccg/

JG_

Damage Realism in GPL/N3

by JG_ » Sat, 12 Dec 1998 04:00:00




>>The one thing that bugs me about GPL, is the total lack of realism
>>when it comes to the damage model...
>>Does anyone agree here? GPL's physics engine is awesome, every aspect
>>is modelled, tiny bit too much brake, the wheel locks, but only if XYZ
>>conditions are also true.

>Given unlimited processor power and unlimited development time I would
>like to see realistic damage. Two things however ;

>1) Coding such a thing will detract from development time (not a big
>issue), but also from processor time (this is a big issue). I think we
>will see realistic damage, but GPL was just a first step along the
>road within the limits of what time and power allow. Where this will
>go in future will be incredible.

I agree totally. It just hit my after a few laps last night how superb
the car felt, and then suddenly I was hitting some barriers (of only a
few microns width), and the car was none the worse.

That is true, but I am sure even GPL's damage model should have been
upped so that it was just as it should be, "realistic" You really
would have to have immense skill. If Papy want to make the most
realistic simulator ever, then surely a realistic damage level that
literally means a brush can end your day than so be it... In Icr2,
practically any tap of an AI gave you front wing damage, which kills
handling on ovals.. Heck, I wouldn't use that mode for racing in GPL
(of any distance), I wouldn't expect anyone to race online with that
mode. But at least it would be there so you could go out for a "spin",
with what seemed like a true sense of what happens when things just go
a *tiny* bit wrong, and a lot wrong. Realistic currently doesn't do
that for me...

Yes, I again fully agree that GPL is justified. I just think it is an
area that is glossed over and thought I may see what people think by
this thread.. I would be very pleased to see a much more radical
damage model for N3, I guess it would be called something like
"dynamic impact reaction" (forgive me for not being that techincally
minded) whereas the body shell could recieve damage of nearly every
percievable type based on the contact with other cars or the scenery.

I shall state again that this is more a dream and I am aware of the
reasons why this is not possible. As you say, the aim is to simulate
driving, and this should always be the main objective. I just feel
that when placed in the most realistic driving simulator to date, the
car feels great, but the rest of the calculations in the environment
could have been increased to provide that feel of danger, in the
relevant modes. It feels "dumbed down" (?) to me, which is precisly
what they haven't done for the car physics. This could be corrected
IMO without needing a new damage calculator...

Thanks,



-:http://www.simproject.com/ccg/

JG_

Damage Realism in GPL/N3

by JG_ » Sat, 12 Dec 1998 04:00:00


Heh. :) I do have this tendancy to lose control and find a nice hard
object to combine with every now and then.

I guess your view is the view that many have, and I generally agree. I
mustn't foget, we have the best simulator to date, yet improvments
will happen in every Papy release, I just feel that this is important,
and may be overlooked by many...

Thanks,



-:http://www.simproject.com/ccg/

John Walla

Damage Realism in GPL/N3

by John Walla » Sat, 12 Dec 1998 04:00:00



>Also, if I get moderate damage to a tyre or suspension, often the first you
>know about it is when you get to the next corner. It would be nice to
>include a visual representation of the damage, as per ICR2. (I'm talking
>about the sort of damage that doesn't throw the steering completely out of
>whack here!).

That would be nice, although often in real life the first indication
of damage is when you get to the next corner and one side doesn't
brake or a wheel locks - that's when you next stress your car. I think
GPL could do with a few more aural or visual clues as to when
something has gone wrong or is about to go wrong, but generally it's
pretty good. I actually quite like the "not knowing" exactly what is
wrong and having to very carefully feel out what the car is no longer
capable of without unexpectedly throwing it off the road. Then you
need to work out the fastest way around the track with a compromised
car - it's a nice challenge.

Cheers!
John

John Walla

Damage Realism in GPL/N3

by John Walla » Sat, 12 Dec 1998 04:00:00



That isn't an argument for making it more realistic though, since
damaging the car isn't the target of the sim. If I blow my engine
every other lap that doesn't mean the sim should more accurately model
the way I do it, in fact it means I should try and stop doing it.

In GPL I guess it would be pretty difficult to get minor toe-in damage
at the speeds people generally go off the road. From what I see of
online races the typical accident is a triple lutz end over end barrel
roll with a mid air pike, and no matter how you dice it that is race
over in any formula. In defence of GPL's current damage model I should
say that I had an online race last night at Zandvoort, chasing Matt
Rowe for first place. Into Panorama I realised I was going in too hot,
and _just_ failed to keep the car on the road. I gently bumped the
fence, kept the car straight and rolled back onto the road without
losing too much. After that the car was a pig, locking up the loaded
LF very easily and picking up wheelspin out of every corner, so I had
to feel my way back around the track and find out what I could and
couldn't do. In that case the damage seemed to work pretty well.

It is something that would be nice to do better, and I think it's
something we will see in future products - it's just not the number
one priority to do "properly" as Mike pointed out.

Cheers!
John

Marc Collin

Damage Realism in GPL/N3

by Marc Collin » Sat, 12 Dec 1998 04:00:00

I agree that the damage model leaves something to be desired, although in
balance it is not that bad.  There are two correlated problems that I
believe to be much worse, however:

1: the AI cars do not seem subject to the same damage model as we are

2: minor mechanical and damage problems were fixed in the pits in those
days, but we can't do that.  Racing for 7/8ths of a full length race and
then being "eliminated" because of a minor incident than could be fixed in a
short pit stop is unacceptable to me.  I don't care about pit stall or crew
graphics, but if I have the ability to limp or drive back to my stall, I
want to suffer a time penalty (even a canned one) and then proceed in the
race if the problem is determined to be repairable.  For me, this is by far
the number 1 problem with GPL.

Marc.


>Hi All,

>The one thing that bugs me about GPL, is the total lack of realism
>when it comes to the damage model...

>What I mean is, nothing else is canned, but the damage, since day one
>of sims I guess, has been. In GPL I expect, with realistic damage, to
>be able to bend the suspension up, if I were to hit the wall at 45degs
>at 20mph. These care weren't that strong were they??

>Here we are with the pinnacle of computer racing sims. and the damage
>modelling seems as advanced as ICR1. Was it Coulthard at Adelaide who
>crashed a Williams in the pit entry? Do this in GPL with realistic
>damage and you may not even get a flat, heck people cheat at the
>Lesmos at 180Mph.

>Does anyone agree here? GPL's physics engine is awesome, every aspect
>is modelled, tiny bit too much brake, the wheel locks, but only if XYZ
>conditions are also true. I am no expert on this, but from all I have
>read it is VERY realistic. However, when the car comes into contact
>with anything, it's back to some very simple maths equations, and only
>3 or 4 canned "types".

>To me, and this is where I'll state me opinion, this does matter...
>Maybe it doesn't for many, but I would like to know. I get the idea
>having no visual clue that you have a flat is a pain, but I see it
>more deeply as: why can't we knock the toe in, squash the front tyres
>in toward the car body, like ICR2, but with greater calculation of how
>far/bad that suspension gets crushed in. Have the wheel go all
>wibbly-wobbly, then maybe drop off at a certain time later.

>I appreciate with body damage, that in GPL it's hard to model crumple
>zones. These things were pretty solid wern't they? I don't know really
>so I won't go on, but this is very important I feel for N3.

>Clips with the wall in N2 can have really dodgy impact sounds but F9
>reveals nothing is damaged. A slightly lesser impact and you may be
>surprised to hear the spotter start "looks like..." In N3, with NROS
>and general online racing, I think it's really going to be necessary
>for N3 to show us a damage model that allows as much simulation as the
>car itself. Every little scrape will need to have an impact on the
>body of a stocker, and this should play as much a part of a car's
>speed in a race, as the calculations with the tyres and the road..

>Thanks for listening,

>Jon.Guest


>-:http://www.simproject.com/ccg/


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.