rec.autos.simulators

   I have a tough time going back and forth between my TPTCC DTMHeat league
and running the F1 2k2 mods.  The steering is different in Heat,  More
flowing, So I have a tendancy to oversteer when I first go back to Heat to
get ready for the next round(which I haven't done yet, and it's Troi
Rivers<Becanour> which I don't have many laps at yet.)  Gotta focus more.
dave henrie

I'm fairly sceptical in this respect. While it didn't bother me much (I got
rid of NH because of controller assignment and AI/gameplay issues before any
mods surfaced), to me the difference in physics rate seemed really tangible
(as it is in F1 2k2 when toggling the half_rate setting). It might go
unnoticed if one were to use Heat as one's sole training aid, just like one
can't tell if A on a piano is really 440Hz when you hear it alone, but
compared to other titles there is a difference, IMO. Just like a trained
musician can hear differences of even a couple Hertz when instruments play
in unison.

Regardless of individual handling there is, IMO, a certain crispness to
handling in GPL/NR2k2 and F1 2k2/GTR2002 which just isn't there in
"standard" NH. Would be interesting to have people comment on how they feel
this situation is with any of the mods.

I think Dave means the bit where Malcolm says it's more important to
simulate the feel than get the actual numbers correct. As in the past I go
along with that to a certain extend.

IMO GPL, NR2k2 and GTR2002 are very good and most convincing at creating the
illusion one's piloting a high performance car on ones computer. All of
these push real numbers around, so as hardware gets more powerfull, capable
of calculating more complex models in real time and provinding more feedback
to the user (motion) this is, IMO, an avenue to pursue.

While I do enjoy other efforts which, supposedly, lean more towards
"simulating the experience", it's my experience that after some use there's
always something that pops up to destroy the illusion for me. I'm ready to
be proved wrong on this count, but not able to test this particular setup at
the moment.

Jan.
=---

As for NH - no idea as it has never been available here and no obvious
suggestion it was a must have to get by importing.

Cheers
Tony

David G Fisher

Hehehe... ;-)

Jan.
=---

Right about what, again?  Five years ago....  Um, was it that a rally
game was better than GPL?  I'm getting old and my memory is fuzzy.

Many people will post simply because the given article may be of
interest.  You, however, often post so you can "prove" you were right
in old arguments that you won't let lie, for example:

Modern F1 cars are not easy to drive (quoting J.Stewart IIRC)
GPL is not a 100% accurate (this very thread)
Papyrus get things wrong (The Jasper article snippet)

Well blow me down if all of the above are not completely obvious.  Hey
- we agree!

Where we disagree is perspective.  You seem to think that Papyrus, or
people who like their *games*, should somehow be corrected, be taught
the gospel according to DGF, to SEE THE LIGHT!  Yay, you've spotted
that these games aren't like total realism.  Whoopee.  What's wrong
with them just being as close as an approximation of a given racing
series as the developers can muster?  And if there are one or two
delusional souls out there who are kidding themselves into thinking
that simracing is totally authentic, why not let them have their
jollies?  Is it that important that you are seen to be right?

Tell you what, Dave, if Ralf gets to be F1 Champion before Juan then I
will paypal you a tenner if that will get you to shut up about that
particular argument.  Please, remember this post.  I am a man of my
word.

Nope :-)

Mark

UPDATE FOR THOSE WHO HAVE ASKED FOR SPECIFICS
-------------------------------------------------------------------------------------------------

The ONLY reason I am continuing to post against this topic is that I
have been asked by a large number of people in all the various forums
and newsgroups to be specific about what the UNREAL elements of GTR
are. I have tried to explain how difficult this is to do as I may be
talking about slight variations in the control inputs required and the
way the car reacts to both those and to the actual processes of
car/track/phyiscs interaction.

I have stated many, many times this is not in any way whatsoever a
criticism of the SBDT who have done a sensational job with GTR2002. I
am total admiration of them and from posts within this topic they do
not seem to be offended in any way and I hope that is the same for any
GTR2002 fan reading this. I have also had email communication with
Doug Arnao who also is not offended by the content of my postings.

I could equally as well do a similar post on GPL/GP4/Netkar/LFS or the
Nascar Heat mods that I referred to in my earlier posts. On the
question of handling realism and taking into account computer sim
limitations none are fully realistic. All have realistic elements, all
have unrealistic elements. I didn't for one moment expect anyone to
doubt that, but it seems many did and asked for specifics - so what
follows is the start of that process.

What good will it do? Well, maybe it will spur the developers of
products such as Netkar (and I only mention this as an example) to
strive to lessen the unreal elements that might creep into their
simulation products. Also, it might just save someone some
embarassment and possibly even injury should they get it into their
heads that real life cars will handle identically to those they have
driven on the PC. Thankfully, fear may well prevent them from going
fast enough to need to call on the instinctive reactions they have
picked up in driving sims

At the moment I am following 2 paths :

1) To try and do something to demonstrate that the driving style that
gets results in GTR is not that which gets results in real-life, also
that you are able to drive GTR in an unrealistic fashion.

2) In 1) you might be able to "spot" the differences in driving
technique from listening to the engines revs etc. and see the way the
car appears to behave. But this does not really cover what it feels
like vs. the real-world or how the technique required differs. Please
don't forget that again, my purpose here is to comment on handling
physics and not any other aspect of the game.

This post is related to section 1

As I said, it is not just the premise of the super-quick to exploit
handling oddities of a sim. We all run the same program and are all
open to these oddities. Initialy they may hamper us but as we continue
to drive and go quicker we develop a technique which gives the results
we want - faster lap times.

To show these up fully though I did need a GTR lap replay from soneone
who had learnt to exploit these, i.e. a "fast" guy. I also needed some
good real-life footage that would demonstrate how that same car/track
would be run in real-life.

I got very close to this by finding, on the internet, a VCR of a GTR
lap from a "very" fast guy and also an in-car lap, from the same
track, with virtually the same car. Now, I could have used footage
frmo another track and another car but I know many would say "well
it's a different track", or, "it's a different car". So, I got as
close as I could.

The first thing I did was to write to the guy who had produce the GTR
lap, explain my intention, point him to the article and ask if he
would consent to me using HIS lap in this way. Had he not replied and
given permission I would not have used it. He understood that it was
neither a criticism of him or GTR.

The reply I got was far better than I'd expected. It seems he was also
someone who had real-life racing experience and agreed with much of
what I was saying. Even if he had not agreed with me, but still agreed
to have his lap used, I would have posted his response here, out of
courtesy.

Without further delay I'd like to introduce this guy, his name is
Dominik Binz (a.k.a. Slasher_GT). The reason I picked his .vcr is that
his was one of the few I found that I could also get a good quality
real-life video for comparison of.

I did a quick check at the time of writing and Dominik is No. 6 at
Find The Limit (GTRank) - his Mugello time was in the top 10 when I
first looked but has slipped a little since - Dominik agrees that his
line was not as good as it could have been.

Here, in full and un-edited, is his reply to me following my initial
request to use his VCR file :
======================================================

I think you've found the right one here :) I'm not only a simracer, I
also had the pleasure to test a real touring car. A VW Lupo GTI
(slick tyres, 140hp, 750kg, race suspension) last year in October
at Oschersleben for two days in a talent triage. I did quite well
and I can say that I had qualified myself to be allowed to do full
season. It was very professional organized by Volkswagen and
we learned how to drive a race car fast.I also own the international
C License to compete in every racing series except F1 :)

Well and to answer your question. What I do in GTR2002 to drive
fast is not comparable to the real life situation. I was quite
impressed when I was playing GTR for the first time. I took a NGT
car and the first laps I did were at Oschersleben. I found it pretty
similar to the things I was also doing in the Lupo (though a Lupo
is not a 996 lol). But I was more or less carrying the car through
the track at that point because I wasn't sure about the
physics. What I discovered is the faster I tried to go the more
unrealistic my driving became...driving this sim on the limit is
far away from reality. Ok you also have to do the right line in
the sim, find your braking points etc but the driving on the edge,
I mean the things you do to be fast are very different from the
things you would do in real life to be fast. For example braking!
I lock the tyres in nearly every bend as you have seen! In real life
my tyres would have been ruined after 3 corners ;) But I have to
brake so late in addition with a high use (unrealistic high use)
of engine break to be fast. In real life my engine would have
blown and or I would have spun cause of too high revs on the
back. Etc etc. Endless list.

Though I have to admit that Mugello is a bad example, also for
simracing. My lines were quite bad and also I exeggerated some
things. I was overdriving the car in that lap. I also didn't have
much practise. I overall have done around 1000 laps in GTR. On
10 tracks, means 100 laps on every track. I have done laps for
all tracks with GT and NGT cars. Means 50 laps with a 2 different
cars on every track. That's not much. With more practise and a
better setup my lines would have been smoother and more fluid
in Mugello :)

Sure you can use my Vcr to show that it's not right what I'm
doing cause I know it's not right :) No problem here, hehe. What
cars have you been driving in your life? Nice posts btw, I was
following this discussion since it was opened but was too lazy
to reply ;)

========================================================

I am aware that there are different fidelity levels when recording a
.VCR file. I believe Dominiks lap was not set on the highest value
hence is not a 100% represenation of exactly how the car behaved when
he as driving it. As I understand it, it is like sampling at a lower
frame rate.

It is pretty close however and the best we have at the moment. You can
easilly spot how Dominik is having to drive the car to produce the
fast lap. Things to note are the whole braking and corner entry
process. Now, of course Dominik is going for the fastest lap possible,
with no damage/wear modelled - so he is able to abuse (lock) the tires
under braking. Of course, no one would dream of driving like this in
an online race (with damage/wear) switched on and many wouldn't push
the car like this even in testing BUT, the fact remains that when you
start to push the car it will behave unrealistically adn will "lull"
the driver into driving unrealistacally.

The "normal" (race school taught) approach to a corner is to brake
smoothly and progressively, ease off the brakes prior to turn-in, turn
in smoothly but directly whilst appying a "balanced" throttle (neither
accell not decell). To hold the car on the balanced throttle, ideally
at the limit of grip until the apex, then to start to accellerate
whilst unwinding the lock (if you are on the limit you can accellerate
whilst holding the same lock).

Of course there are slight variations between corners, some you get on
the gas before the apex, some after, as you improve your technique, on
some corners you may hold some braking on as you begin to turn
(trail-braking) but that is it. Using the available grip where it is
needed most (grip is finite, can be used for braking, cornering,
accell but increasing the requirement for one detracts what is left
for the others - traction budget/circle).

I have sought approval from Dominik to post the above text but he has
also asked me to include the following. As you see, Dominik, like me,
wishes to make clear that there is still great benefit in sims beyond
simply providing ultra-realistic handling phyics and that GTR is a
fabulous product. I recognise this desire so as not to be criticised
for things that you are not actually commenting on. The following is
complete and un-edited.

===================================================

I want this to be added too:

But for a game it's not necessary to be 100% realistic. Ok it
shouldn't handle like a chewing gum. The fun factor is
very high in GTR and that's cool for me. Also Doug Arnao
admitted that it's not very close to reality but ...

read more »

I'm "not" pushing heat but you can order it here :

http://www.strategic-plus.co.uk/

Also, I think you can get it from Oval Racing News, see
info here : http://briscaheat.m4driving.sm/faq.htm

Some also say you can find it on ebay.

Maxx

Jan,

On Mon, 03 Feb 2003 18:45:49 GMT, "Jan Verschueren"

Essentially, you create a ghostlap (heat does this by default
for your fastest lap) you copy itn and rename it to reside in
your mod directory. Then, next time you race the AI it will
use YOUR ghost as the basis for it's AI.

I'm not sure what the relative sampling rates are but it's
a small issue in relation to he overall physics mdoelling
IMO.

I think there is still a little lack of crispness as you refer to
it. Maybe it's (partly) this that makes the Heat tire model
feel more convincing to me. Just speaking of GTR, the tires
feel like they are solid *** with no flex. As soon as you
turn and get any degree of slide you seem to get a fairly
constant squeal - in Heat the tires feel like they are flexing,
maybe it's just the physics engine lagging a little before
"biting" - I think in fact it is more tham this as there are
variables in the physics files for stiffness of the tire walls.

I agree, I'm all for higher sampling rates and the future progress
of sims but Heat as well as those you mention push real numbers
around.Certainly GTR/F12002 has more "numbers" to deal with
than Heat but that plus a higher sampling rate does not mean it
will automatically produce a more realistic feel. In many ways I feel
it is BECAUSE of the extra effects being catered for in GTR that
it thows up more UNREAL elements - clearly, in terms of progress
we should work on improving that level of physics model rather
than sit on the the laurels of Heat's less sophisticted but, IMO
more realistic feeling (at the moment) handling.

Yes, the illlusion is much more than the handling phsyics. The
Mini-Cooper in Netkar feels spookilly realistic at times (in terms
of handling) but the current level of sophistication in terms of
graphics and such is not as convincing as some other products.

Maxx

Well, no one can acuse you of not following through:)
*Much* appreciated you taking all this time to respond and start this
rather interesting issue..

Cheers,
Willem

Yes, thank you Maxx. Very much appreciated; the time you spend.

MRSisson

--  

LOAD "GPL",8,1
RUN