rec.autos.simulators

Well there is no doubt it is an evolution of previous titles in the
series but things do often evolve into something great which is what I
believe we have with NSR.

The new physics as mentioned in an earlier post are great, you now have
controllable oversteer and can put the car into a drift/slide that you
can collect up.

I was originally comparing NSR to the current king of driving simulations
GTR not NR2003 (which it competes admirably though hardly perfect).  

Mitch
--
Remove "nospam." to reply.
SuSE 9.2 Pro KDE 3.3.2a

You are really comparing apples with oranges in that case.

Fair enough Achim; just couldn't resist baiting our omnipresent new friend.
Was thinking it might be ymenard in disguise... ;-)

But physics and driving feel is the number one thing for me.  I haven't dug
around in the files a lot yet, but I see the suspension files remain from
Nascar Thunder 2004, which Doug Arnao did for them a couple of years ago (at
least according to the file headers.)  But for a bit of NSR physics
treasure, how's this bit from the HDV file.

DiffPumpTorque=250.0                // at 100% pump diff setting, the torque
redirected per wheelspeed difference in radians/sec (roughly 1.2kph)
DiffPumpRange=(0.0,0.05,1)          // differential acting on all driven
wheels
DiffPumpSetting=0
DiffPowerRange=(0.35,0.05,18)       // fraction of power-side input torque
transferred through diff
DiffPowerSetting=0                  // (not implemented for four-wheel
drive)
DiffCoastRange=(0.0,0.05,1)         // fraction of coast-side input torque
transferred through diff
DiffCoastSetting=0                  // (not implemented for four-wheel
drive)
DiffPreloadRange=(10.0, 5.0, 1)     // preload torque that must be overcome
to have wheelspeed difference
DiffPreloadSetting=0                // (not implemented for four-wheel
drive)

All the related diff setting values are at zero in all the SVM setup files,
which means you have a limited slip diff.  That's one reason you can easily
do big lurid slides in NSR cars (at 170 on pavement).  Cup cars of course
have Detroit Locker rear ends.  Back when I was messing with the demo, I
changed the HDV settings to mimic a locked power-side diff and the rear
actually behaved a lot more like the cars in NR2003.  Admittedly, slideways
is not the fastest way around, it is very easy to do, lap after lap.  Drive
it in, turn it down hard and toss it sideways, then just floor it and steer
your way out.  One-handed if you like.  Bit like Ratbag's Sprint Car game,
but maybe even more forgiving here.  (At least in that one you had to really
work the pedal to find some bite off the corners.)  And as people noted
about the demo, you can do one-tire burnouts against the wall - which is
tough when the rears are locked together.  The tire temps show it's not a
graphical glitch.  Oops!

The ISI engine can produce a solid and challenging driving experience, but I
think the above offers a little evidence that EA/Tiburon have gone the route
of making it "seem" real to the masses rather than making it real in the
data and letting users learn how to deal with the result.  And as the masses
are generally sure they could win the Daytona 500 given the chance, they
love the ability to recover from losing the rear end with ease.  If it was
as easy to play outside of the friction circle as it is in this game, we'd
all be stars.

All of which means there's much less of a learning curve for people to get
up to say, 98% speed, and users will find themselves crashing a lot less.
The flip side is that driving NSR is much less involving than in NR2003.
The limit is much more narrowly defined in NR2003 but the game gives much
better feedback on where it is exactly, so that you know when you're
approaching it and have the drama of trying to tiptoe JUST on the safe side
of losing it.  Flirting with disaster, if you get my drift.

In it's favor, NSR should be more accessible to a wider audience than
NR2003, as many are not prepared to or interested in getting on top of a sim
the way post-GPL Papy sims require.  And who knows; EA/Tiburon may have hit
the difficulty level in the sweet spot for the wide acceptance they want.  I
mean, we in RAS tend to think of Papy's Nascar efforts as big sellers, but
they're market pipsqueaks in EA terms.  Am I incorrect in thinking Papy
Nascar games didn't sell as well after adopting the GPL physics engine?  My
guess is NSR's exactly what they wanted it to be; real enough to call a sim
(which is certainly is), but not so much so that it turns people off.  EA
wants to sell a boatload of these games to recoup the licensing fees and
they know that folks like us are in the distinct minority.  So they've
tilted it a little in the easy fun direction, which means there's hard fun
to be had in it.  Guess I'm more of a hard fun guy.

SB

...

Indeed, NSR errs a tad on the side of 'user friendliness' for my personal liking
as well. However, as you've outlined, the settings are there, it only takes
someone to use them.

What I was getting at was a lot more simple, though. I think you'll have noticed
that as well, but I am under the impression that NSR combines Papy's excellent
tire deformation model with ISI's highly sophisticated suspension (and
transmission, as you outlined) model, plus the configurability ISI sims have
always had.

I.e., NSR is a hybrid of a new class, combining the previously strictly
separated highlights from two worlds.

In its default state, NSR may not meet the expectations of *** simmers, but
if the above is true, it is a respectable simulation with interesting potential.

I don't like excessive copy protection either (although I don't have a prob with
it on my machine), but NSR may be a little gem in the physics and modding
department and should be given a very close look before discarding it
lightheartedly.

Achim

...

NSR errs a tad on the side of 'user friendliness' for my personal liking as
well, but as you've outlined, the settings are there, it just takes someone to
use them.

That said, what I was getting at at was that I am under the impression that NSR,
on top of having the highly detailed suspension and chassis model plus the
configurability of ISI sims, has a tire deformation model that seems to be as
sophisticated as that of N2k3.

If I'm right, NSR combines two assets which previously had been strictly
separated, which makes it a pretty interesting simulation in the physics department.

Everybody is entitled to their opinions, but I think in the best interest of the
simming community, NSR should not be discarded lightheartedly. It might have
hidden potential which deserves to be discovered, and used.

Achim

Gotcha; just meant it didn't feel like Papy's tire model to me.  I actually
graphed and tweaked the slipcurves in the demo and messed with the dropoff
function to create a less forgiving tire, which in concert with power-side
locking of the rear diff did make the cars more edgy and rewarding to drive.
But with all the talk about the demo physics being an old build and not
representative of the final build I just thought I'd wait and see what the
final looked like.  Not all that much different.  Wish I'd saved some of the
demo physics files to compare values!  As delivered, the thing that came to
mind was that it felt about like Nascar Heat in terms of feedback.  I'll
have to reinstall that one for comparison though, as Heat hasn't been
installed in a while... ;-)

But yeah, it is the same basic tire model as other recent ISI products (and
the previous Nascar Thunder games), which is to say it's a complex and
well-thought-out one.  The game does have a solid physics engine under it,
so there's no reason it couldn't deliver a GTR-level driving experience with
the right variables in there.  But since everyone's files have to match for
online playing, the fix would pretty much have to occur at the company level
and I have trouble seeing that happen.  I think they're aiming elsewhere.

SB

Jeez, I really should proofread before I send this junk!  Meant to say
"there's LESS hard fun to be had in it ".

Better now...

Dunno.  I've never had a look at the guts of Papy's tire model, and the TBC
files in NSR look just about like they did in prior iterations with the ISI
engine.  Don't have any old NT files handy for comparing values, but they
have the same inputs as rFactor save for multiple slipcurves, which the old
model could do anyway.  So all I can say about the tire model v. Papy's is
that it's too forgiving and the car feels really dead as a result of it.
That was actually the first thing I noticed about it after driving it right
after NR2003; the Papy car just feels so much more lively while the NSR car
just feels very damped and dull.

Unlike some, I have no issue with copy protection or with EA or Tiburon for
making the game the way they see fit, and generally ignore all the marketing
hype anyway.  But it escapes me how any end user can say this thing advances
the genre over NR2003 in any way whatsoever.  Did some online racing with it
last night, and it was an all around poorer experience in my view.  A
matchmaking service that can't even compete with the old version of VROC for
functionality, and even more warping than I see in NR2003 - this with 100%
broadband users.  The shallower learning curve and the fact that you can
bang into each other without much effect will no doubt please some, but not
moi.  And that turbo draft...

I bought NSR yesterday just to muck around with it and see what could be
made of it, but unlike some others, the more time I've spent with it to date
the less I think of it.  That may change, but I have to say I'm thoroughly
unimpressed so far.  Hope I run across something to change that.

SB

...

Well, of course it isn't Papy's tire deformation model, but it feels like it.

The forgiving stuff you're dissatisfied with is a very forgiving slip curve, and
that's something that's easily cured in the config files by altering a few
numbers in the slip curves (well, a few - it's 10 minutes of typing).

But the basic thing is, it feels like a pretty good tire _deformation_ model,
and that is something really complex to program for all I know. So, having that
in is quite something.

Achim

Im still in absolute disbelief in the poor reception to GTR.  The demo2
really sets the standard high yet still almost nothing from this group, a
SIMULATION newsgroup.  Half the people that come here need to go back to
their arcade roots and leave the sims to the men ;)

Mitch
--
Remove "nospam." to reply.
SuSE 9.2 Pro KDE 3.3.2a

That'll probably change when it hits worldwide release; as it is the buzz is
sort of trickling out rather than hitting all at once.  The original leaked
press release did make quite a splash here.

But to be honest, some of it has to do with the buggy and problematic nature
of that generation of ISI engine.  Took me tons of tweaking to get the
***y thing to run halfway decent on my box, which runs all non-ISI sims
(and even rFactor) just fine.  And now NSR is looking like another
tweak-fest.  The first thing I noticed before even running it was that I had
to tweak the 3D config util to get all my video resolutions, just as with
the demo.

I have sympathy for the developers though, it's a ton of work to put one of
these games together, even one that uses middleware to get there, ala GTR
and NSR.  I think the difference between those two says all you need to know
about the developer's focus and intent.

SB

Just had to have another look. ;-)

FWIW, I had some old data in Excel to compare to the current stuff, and the
as-shipped NSR lateral slipcurve is considerably less forgiving  than the
curves in rFactor, or any previous ISI sim or mod I had numbers for,
including the original GTR2002.  Can't read the current GTR data, but as for
the overly forgiving slidey nature of NSR, it's coming from elsewhere in the
tire model.  I mucked with the latpeak and longpeak numbers in the demo to
tweak it as that appears to affect the sliprange where peak force occurs.
And the range in NSR is considerably wider than rFactor for those variables.
Need to poke around in there some more.  Interesting stuff...

SB

...

Yep, but a slipcurve cannot be seen isolated from the other tire parameters. If
I wanted to mod the _tires_ for less forgiving slides in NSR, I'd mod the slip
curve.

Achim

...

I forgot to add that the slipcurves probably interact, and Lat curve probably
cannot be looked at alone.

Achim