rec.autos.simulators

Quick TOCA 2 Ramblings

John Walla

Quick TOCA 2 Ramblings

by John Walla » Fri, 29 Jan 1999 04:00:00

On Wed, 27 Jan 1999 23:51:21 +0000, Paul Jones


>I wasn't looking hard enough - I'll recheck. "Canned" is an interesting
>concept in programming though - you almost never want to do it.

It depends what you are modelling I suppose. If you are modelling
lateral g, and it's effect and interaction with camber, contact patch
etc then it is better not to have a canned effect. If, OTOH, you are
not modelling all of these things then you can write a "lean" effect
for when the car corners at beyond a certain speed/turn radius.

Certainly within TOCA2 the "lean" effect reacts more slowly even than
my own road-going Honda Accord, and since the touring car will be
sprung far more stiffly and with less total movement it is clear that
the effect bears no relation to the on-screen events.

The important question is whether that actually matters, and frankly I
don't give a damn. If it was a sim I'd be a mite suspicious of it, but
as an arcade racer it really doesn't matter - the overall effect is
way more important and TOCA2 does very well in that regard.

Obviously anything which is not "real world" is canned, but this to a
greater or lesser degree rather than a black or white yea/nay. GPL
models in far greater depth than anything up until now, but then as a
sim that's something it should aspire to. Even then it doesn't model
changing track conditions, debris pick up, marbles or a host of other
things. TOCA2 is a different beast in that I would expect things to be
canned - if someone told me the TOCA2 programmers had spent weeks
coding the camber change due to body roll I'd be askiny why they
didn't spend that time and money giving me another couple of cars and
tracks.

A bit like the catapult launchers the other FFs seem to use at the
start. Can anyone even kep them in sight off the line? I usually catch
them up into the first right hander after the chicane complex, but
they're seriously quick off the line. Unfortunately they're too slow
thereafter.

Cheers!
John

Jack

Quick TOCA 2 Ramblings

by Jack » Fri, 29 Jan 1999 04:00:00

John, are you sure about the marbles thing? It certainly seems that there
are marbles on the outside of Speed Trap at The Glen, for example. Or is
that just the supposed negative camber there?

Paul Jone

Quick TOCA 2 Ramblings

by Paul Jone » Fri, 29 Jan 1999 04:00:00


> The important question is whether that actually matters, and frankly I
> don't give a damn. If it was a sim I'd be a mite suspicious of it, but
> as an arcade racer it really doesn't matter - the overall effect is
> way more important and TOCA2 does very well in that regard.

But the realism is very important and it's hard to see how the complexity of
the algorithms won't be proportional to the realism. I don't see why an
arcade game shouldn't be realistic. This is what lets NFS down so badly.
There is nothing to driving those cars. It's almost <amount throttle down>
proportional to <velocity> and <steering wheel turn> proportional to <change
in direction>. Admittedly you can turn the car with the brake but I need
more, and the cops don't give me that.

Again, I don't accept "canned" but I will accept that the algorithms are much
simpler than GPL's and thus the realism is less than in GPL. I would much
prefer for TOCA2 to max out on realism. More tracks and cars are fine but I'd
rather a hyper-realistic simulation of the BTCC. TOCA2 is probably the best
we're going to get until TOCA3, and it is fun.

What I found is that if you hold the revs between 8 and 9 on the tach until
the green light and then floor the throttle while clicking up to 4th as
quickly as you can you can overtake 4 of them before the first right hander
(Hawthorne). For some "canned" reason the car only really starts to pull in
4th even if you set the ratio for 3rd to be the same as the default 4th.
There after it seems to me that you never want to drop below 4th even on that
tight hairpin, whatever ratios you set. Seems to me that you get more grip
from the lower gears but speed is compromised. Acceration doesn't seem to be
affected by the gear selected and the car picks up pretty well in 6th.
Realistic, huh :-) I think back to my school boy physics and think, no.
Shame. They done a great job on the graphics and the look of the car pack is
gorgeous. Just a little consultation with a school physics teacher, yourself
or Doug Arno might have got their physics up to an acceptable level. This
said the game is amazing fun.

Cheers,
Paul

ymenar

Quick TOCA 2 Ramblings

by ymenar » Fri, 29 Jan 1999 04:00:00

Paul Jones wrote

I take this as a good explanation, and what we expected (well I did) for
Toca2.  I had fun, but for me I'll rather get a Cooper at the Ring.
Personnal flavors, but I do respect that Codemaster did a great job with
their title to appeal the market they targeted.

I will argue with you.  I launched Toca1 after... and Im sorry but the
physic engine (or game engine) is not different at all.  It's mostly the
same. Same effects on medium-speed corners (the canned oversteering spin),
same effect braking into a corner, or turning into a slow corner (the
chicanes at the end of the lap).  There's no major upgrade to the physics in
Toca2.  Probably some little adjustments, but nothing more.

<snipped good stuff about canned physics>

The problem with canned physics is that even if the company tried their best
to immitate the feeling, it will always be wrong in heavy-weight transfering
situations.  You can't immitate real life physics with canned algorithm.
You can with mathematical calculations. Take per example the Ubi game
engine.  Their tracks are flat.  it only goes up and down.  No camber, no
banking, no heavy elevation changes (like ex: Zandvoort out-of-track
mountain climbing hehe).  When the track loads, inside the track files there
is data that contains XYZ position of bumps on the tracks. Each time the car
pass that bump, the algorithm calculates the Z-position of the***pit
travels, without affecting the physical look of the suspension.  Now take
the same with GPL, and look at what mathematical calculation will do.  You
will see first of all visually the bump in the track (Monaco near the
***).  You will see the wheel go up/down in the Z axis, the suspension
act realistically under the physics, the shock absorbing the rotation of the
wheel up/down, the***pit who will visually act on the bump, the other
wheels who will have more weight on them, right rear wheel will have the
biggest weight on it, etc.., etc..

That's what makes GPL so incredible compared to the other game engines using
canned algorithms.  Not because GPL is difficult, but because it's the
closest at the moment to real-life.  I remember saying when the GPL demo
went out (almost a year ago !), that I don't see in the next two full years
a company having a better game engine than the Papy one.  That's mostly what
will happen.  I see only ahead the Trans-am Racing and the Rally 99 game
engine that could be as good as the Papy game engine.

Hmm, that means the physics are canned ;-)

Hmm, "upgrading" to the MGP game engine would be more a downgrade ;-)

-= Fran?ois Mnard <ymenard/Nas-Frank>
-= NROS Nascar sanctioned Guide http://www.racesimcentral.net/
-= SimRacing Online http://www.racesimcentral.net/
-= Official mentally retarded guy of r.a.s.
-= May the Downforce be with you...

"People think it must be fun to be a super genius, but they don't realise
how hard it is to put up with all the idiots in the world."

ymenar

Quick TOCA 2 Ramblings

by ymenar » Fri, 29 Jan 1999 04:00:00

Paul Jones wrote

well, it can be realistic, but is it nessecary for an arcade game ?

Well, let's remember they tried to hype Toca2 saying it would be
hyper-realistic.

But you have you 2 good points. I don't really care if it's canned physics,
since I feel it's not the major focus on it.  Arcade game needs a good
feeling of speed, good eye-candy, fun tracks, average AI that fights with
the driver, and probably less complicated physics to please the mass market.
Anybody know what are the secret tracks ? I wish for a long Brands Hatch ;)

Thanks for the hint ;-)   the FF cars also seems to be slingshooted when the
cars go by the frontstrech another time. Watch them exiting the last slow
corners.  Im wheelspining, trying to gain momentum with the less wheel spin,
but those AI just power off like a Formula 1 car ;-)

-= Fran?ois Mnard <ymenard/Nas-Frank>
-= NROS Nascar sanctioned Guide http://www.nros.com/
-= SimRacing Online http://www.simracing.com/
-= Official mentally retarded guy of r.a.s.
-= May the Downforce be with you...

"People think it must be fun to be a super genius, but they don't realise
how hard it is to put up with all the idiots in the world."

John Walla

Quick TOCA 2 Ramblings

by John Walla » Wed, 03 Feb 1999 04:00:00



I think that will be a camber effect. I'm as sure as I can be about
the marbles without knowing for sure (if you know what I mean!) :-)

Cheers!
John

John Walla

Quick TOCA 2 Ramblings

by John Walla » Wed, 03 Feb 1999 04:00:00

On Thu, 28 Jan 1999 20:40:43 +0000, Paul Jones


>Again, I don't accept "canned" but I will accept that the algorithms are much
>simpler than GPL's and thus the realism is less than in GPL.

Now there's a statement I disagree with - algorithms are a means to
achieving realism, but they alone do not ensure a game will feel
realistic. I'd far rather have something "canned" to hell but feeling
right than programmed to the "n"th degree of complexity and feeling
awful.

Even with TOCA3 I doubt we'll see such realism, since it is clearly a
console title now and being ported to the PC because it can be and
sales will be good. Developing an utterly realistic title for the
Playstation would be neither possible nor good sales (IMO), and the
only alternatives would be developing solely for the PC or to develop
two separate releases - neither very likely (again, IMO).

AFAI am concerned it achieves what it sets out to do and then some,
and while I would like a totally realistic BTCC sim I also really like
a fun one and this fits the bill perfectly. No disrespect to
Codemasters but if a BTCC sim were to be made I'd prefer it to be done
by GC or Papyrus since we'll likely only get one shot at it and I'd
like it to be perfect.

Where I think Codemasters have really got it is the "rough and
tumble", door-rubbing, love-tapping closeness of it all, something
that even Papyrus have failed to capture in past titles. Even in N2
the cars would spin with even minimal contact, whereas I'd expect a
hugely heavy NASCAR to be able to take nad dish-out quite a bit of
punishment, especially at low speeds around Bristol etc. TOCA2 is
great for this (perhaps a bit too exaggerated, but let's not split
hairs!) and I love leaning on cars round corners and squeezing between
a duelling pair as the sparks fly :-)

All in all it's certainly one on my "buy" list, and the addition of
Formula Fords and other cars to drive is just great. Even the
inclusion of mini track guides corner by corner - nice attention to
detail. It's almost like some of the people have done some background
research here or on the web to hear some suggestions and ideas people
made to improve past products, but it couldn't be that. Could it?

Cheers!
John

ymenar

Quick TOCA 2 Ramblings

by ymenar » Wed, 03 Feb 1999 04:00:00

John Wallace wrote

I agree to disagree with that John ;-)

The problem here, is that we're close to the "third generation" of racing
titles now. GPL is the first step into it IMHO.  You would rather like the
canned to hell but feeling right. I can see that.  But how could a software
programmed to the "n"th degree of complexity feel wrong ?

I can't see how a hard-core simracing title, using real-life physics without
any canned feature can feel wrong.  Physics don't change wether it's virtual
or real.  I just can't imagine myself how a realistic game engine can feel
wrong. If it doesn't feel good it's because it has some canned features no ?

And if whatever the case it does feel wrong and the physics are true, thus
that would suggest that the real-life version of it is feel wrong, no ? :)

-= Fran?ois Mnard <ymenard/Nas-Frank>
-= NROS Nascar sanctioned Guide http://www.nros.com/
-= SimRacing Online http://www.simracing.com/
-= Official mentally retarded guy of r.a.s.
-= May the Downforce be with you...

"People think it must be fun to be a super genius, but they don't realise
how hard it is to put up with all the idiots in the world."

Paul Jone

Quick TOCA 2 Ramblings

by Paul Jone » Fri, 05 Feb 1999 04:00:00


> John Wallace wrote
> >Now there's a statement I disagree with - algorithms are a means to
> >achieving realism, but they alone do not ensure a game will feel
> >realistic. I'd far rather have something "canned" to hell but feeling
> >right than programmed to the "n"th degree of complexity and feeling
> >awful.

> The problem here, is that we're close to the "third generation" of racing
> titles now. GPL is the first step into it IMHO.  You would rather like the
> canned to hell but feeling right. I can see that.  But how could a software
> programmed to the "n"th degree of complexity feel wrong ?
> I can't see how a hard-core simracing title, using real-life physics without
> any canned feature can feel wrong.  Physics don't change wether it's virtual
> or real.  I just can't imagine myself how a realistic game engine can feel
> wrong. If it doesn't feel good it's because it has some canned features no ?

> And if whatever the case it does feel wrong and the physics are true, thus
> that would suggest that the real-life version of it is feel wrong, no ? :)

Hi John, Fran?ois,
The issue I take here is that no programmer in their right mind would "can" a
driving model - arcade or simulation. "Canned" as in "canned laughter" in a tv
comedy is analagous to certain programming tricks like sticking bitmap textures
on objects. That is canned. The simulated movement of a vehicle whether
realistic or not will almost always be governed by algorithms. We should get out
of the habit of talking about canned driving models when what we really mean are
poor or over simplified algorithms.
Of course, complexity does not guarantee realism. At the level of considering
car physics, the world is a very complex place and even if you get the physics
of each individual component correct (ie it obeys all Newtons, Hookes etc laws),
you have to consider how each interreacts with the other, and the total input
into the final system each component might make. Gross simplicity is almost
guaranteed to fail when applied to the total car physics, as the resulting
n-dimensional performance curve is guaranteed to deviate from reality and pretty
soon sim drivers will spot this (probably before turn 1). But simplification
will be necessary because the physical world out there outscales the computing
power of any computer that will ever be built - ever, ever, ever. The trick to
good simulation (apart from running it on the very fastest machine you can get
your hands on), is to identify the level of complexity each sub-component should
be assigned. You could spend out almost your entire computing budget on the
behaviour of say, fuel ignition and grossly oversimplify tyre behaviour. You
will then end up with something that is worse than Need For Speed.
Thinking about this makes me want to develop a driving sim - sounds fun doesn't
it. John (or anyone else in the UK) fancy having a go at it? I'm a very
experienced technical programmer (I've done a variety of non-racing sims and my
Bridge program won the gold medal at the 1st Computer Olympiad). My physics and
maths are adequate. My racing and automotive skills are rudimentary and I'd like
to team up with someone with the converse ie. rudimentary programming skills,
good racing automotive knowledge and at least passable physics and maths. I've
got one or two ideas that I think would be revolutionary. I guess the end
product would be the "engine" for the game which we could sell to a games
developer. The graphics, though important, don't really interest me, unless
anyone out there would like to take that aspect on. It would have to be a spare
time project because I need my job to finance my family. Just a thought, really,

Cheers,
Paul

John Walla

Quick TOCA 2 Ramblings

by John Walla » Fri, 05 Feb 1999 04:00:00

On Thu, 04 Feb 1999 01:40:50 +0000, Paul Jones


>The issue I take here is that no programmer in their right mind would "can" a
>driving model - arcade or simulation. "Canned" as in "canned laughter" in a tv
>comedy is analagous to certain programming tricks like sticking bitmap textures
>on objects. That is canned. The simulated movement of a vehicle whether
>realistic or not will almost always be governed by algorithms. We should get out
>of the habit of talking about canned driving models when what we really mean are
>poor or over simplified algorithms.

Those are two different things, leastaways in "ras-speak".

A simplified driving model is one which models only that which is
necessary to achieve the effect the programmer is looking for. At the
basic level you could program an acceleration curve linked to the
joystick input position without worrying about engine revs, flywheel
mass, temps or whatever.

A canned model is not a driving model, it is an effect within that
which takes over when a certain set of conditions are met. Examples
are GP2's spin and spin-turn or TOCA2's "lean" effect. If cornering
force >= "x" then "lean" the view window.

It depends what you mean by "fail". The final target is something that
feels real, be that by hook or by crook, and how it is achieved can be
argued afterwards. Something like Mario Kart 64 has given a lot more
pleasure to me than MGPRS2, and while it may not be remotely "real"
nor based upon real-world physics it isn't difficult to see which has
been the greater commercial success (quite apart from the fun).

Cheers!
John


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.