Zapper's car pack (vannila cars)

[Zapper]

Thank you for all your words. They are all likely to get motivation and my best effort in making these improvements.

Jaguar XJR-9 IMSA is one of the vanilla cars that took the longest path to get right!

I expect that the remaing one, the Porsche 962, should be easier.

[Zapper]

- Porsche 962 IMSA:

You cannot view this attachment.
You cannot view this attachment.
You cannot view this attachment.
(*)
You cannot view this attachment.

Presenting last but not least, the Porsche 962 IMSA from the 1988 season.
Being one of the most successful race cars due to its decade-long career (since 1984) and its extensive use (+100 chassis) in both IMSA and Group C (962C in WSPC and JSPC), the research process to find the version to be portrayed in the game was naturally more challenging.
In this way the resulting 3D model is based on various aspects of the most successful chassis, the 962-121 in 1988, and others such as the 962-120, 962-122 that also competed in the IMSA championship in 1988 and 1987 respectively.
Three paint variants (main sponsors "Blaupunkt", "Havoline" and "Dyson/Royal Crown") were incorporated to portray different aerodynamic package options (rear diffuser and/or front side wing flares) used by the "Dyson Racing" and "Bayside Disposal Racing" teams. Unfortunately a low drag rear wing aerodynamic package associated with a longer body (used for example by 962C in Le Mans races) was not possible to replicate due to limitation of vertices and/or polygons available.

In IMSA specification the 962 in 1988 used a 3.0L air-cooled engine (Type 962/72) with single turbo, instead of the water-cooled twin turbo used in its equivalent 962C in Group C championships. The performance of the 962 IMSA is thus somewhat restricted which translates into 695HP which is more comparable to Jaguar XJR-9 IMSA in game competitor. If it were chosen to use a 962C engine with 720-740 HP performance, it would be more like an Indy's competitor.

As for the instrument panel, it was designed with the options that I considered most suitable for the game, having been composed of several aspects of the previously mentioned 962 chassis (and others such as 962-C04 or even 962C-102). In short, I had to consider that over the years the chassis were always adapted/developed and therefore there is no real correspondence with their specific time period based on the photographs found on the web. Similar to what happened with the Jaguar XJR-9 IMSA, being a two seater car the steering wheel is at an offset position and the right front wing is visible, 5 speed gear is "dog leg" patterned and lastly a speedometer was added where a digital compressor discharge temperature gauge resides, but this time a fake feature that was made on purpose.

NOTE: (*) - The yellow customized color needle appears only with Stunts 1.1 Mod 1.0

[HerrNove]

Congratulations, that was a mighty work! The 962 can finally open its "eyes"

Interesting that the new version has slightly worse values for both 0-60 acceleration and top speed, but the power curve seems to suggest that its performance at typical racing speeds has actually improved.

[Zapper]

Congratulations, that was a mighty work! The 962 can finally open its "eyes"

This one got more research work than making the real hands on stuff 

Interesting that the new version has slightly worse values for both 0-60 acceleration and top speed, but the power curve seems to suggest that its performance at typical racing speeds has actually improved.

If you check the stunts wiki list of DSI's cars, both IMSA cars are actually slower at 0-60MPH than what is shown on their respective descriptions and wich is also true against these new tuned versions.

[Zapper]

Hi you all, this car pack is taking it's final tuning touches, and one of them is the brake parameter in order to adjust each car braking power in match (or approximately at best) to their real stopping distances.

The braking test method takes into account the distances taken from 60-0 mph and 100-0mph and it's measured using the information under "Dragstrip testing" section of Car testing page Stunts Wiki, i.e., "a tile amounts to 204.8 feet".
Here is an example image of how it's done  with the LM-002 on 60-0mph:

You cannot view this attachment.

After doing that to all cars here are the resulting "new brake values" and their real and in game comparisons table:
You cannot view this attachment.

As we can check some braking distances are pretty close and I was pleased to get the 60-0mph stopping distances under 1m and even an acceptable 2m error margin, but for the 100-0mph it was not possible to match namely for the LM-002 case with almost 20m shorter distance.
The 100-0mph stopping distances were the most difficult information to gather on the web, they were some times very disparate so I considered the 60-0mph stopping more relevant in this regard. In this respect some compromise adjustments were taken on GTO and ZR-1, getting a little less stopping distance on 60-0mph and minimize the 100-0mph excessive stopping distances.

I hope with this the car pack gets a better approximation to each car braking capacities judging by their braking technology, curb weight and balance.

Next will be the grip tuning

[KyLiE]

You've done some fine work with this!  Remember that you're dealing with a 35 year old game and even older cars, so there's always bound to be a compromise with the performance parameters.

[Zapper]

@KyLiE , I completely agree given that the physical model has its shortcomings, as we well know, it is still quite capable of simulating through these adjustments for each car and not as something "stuck" within the game's engine.

These adjustments were reminiscent of when, in a game just 4 years younger, "The Need For Speed" had the collaboration of "Road & Track" magazine to help calibrate the cars to make them more realistic. So much that in the same game, it was possible to do acceleration and braking tests on the "Rusty Springs" track (using the F8 key) in order to show that accuracy.

You cannot view this attachment.You cannot view this attachment.

Such a clever touch that does justice to the simulation game it also intended to be, and much to the delight of racing sim fans like myself.

[KyLiE]

I'm a big fan of The Need for Speed!  That feature isn't very well known, despite being included in the user manual.

[alanrotoi]

Oh I didn't know that! I bought a pirate copy of TNFS back in 1995 but I couldn't play it until the next year because my pc was a 486 DX2, not enough power. Anyway, it didn't come with the manual of course

[Cas]

Very neat!  I didn't know this either!

[Zapper]

Hi,

Just to give some update, I begin grip tuning tests for all 11 original cars on 2 types of flat tracks, one high-speed with wider curves (named "DRIFT") and the other more shorter with twisty and tight corners (named "DRIFT2").

You cannot view this attachment. You cannot view this attachment.

As you can see from the screenshots taken with Bliss, for both track types there is a base asphalt path and two symmetrical paths on dirt and ice surfaces for better comparison although these can surely be biased by the game's steering bug.

For now I'm reaching the following grip values for all cars shown in the table below.
You cannot view this attachment.

These values are the result of "base grip" multiplied by their respective asphalt, dirt, ice and grass coeficients. Next to each car grip values are the respective lap times for track and type of surface path taken:
(The table is ordered by fastest DRIFT asphalt lap times.)

These adjustments take into account each car's engine layout (balancing), wheel traction, weight and stability control equipment (in the case of the NSX).

[Duplode]

Excellent work! These lap time comparisons with varying grip/surfaces are very interesting.

As you can see from the screenshots taken with Bliss, for both track types there is a base asphalt path and two symmetrical paths on dirt and ice surfaces for better comparison although these can surely be biased by the game's steering bug.

One way to eliminate bias that could be more convenient than e.g. reversing the tracks is testing on the Mindscape 1990 game version -- with it, left turns aren't faster than right ones.

[Zapper]

One way to eliminate bias that could be more convenient than e.g. reversing the tracks is testing on the Mindscape 1990 game version -- with it, left turns aren't faster than right ones.

I'll try if these new cars are loaded correctly in that version, or alternately, use same path in diferent surface type on distinct tracks.

[Zapper]

Hi,
More grip tests were carried out on the same tracks, but mirrored in order compare with previous results by considering an average lap times for each type of track surface.
With this approach its possible to negate any biased steering behavior bug interference, as far as I get.
You cannot view this attachment. You cannot view this attachment.

Here are all the lap times in "DRIFTM" and "DRIFT2M" tracks (ordered by asphalt path lap times on "DRIFTM"):
You cannot view this attachment.

As expected, on this mirrored tracks the lap times slightly decreases on ice surface path being mostly formed by left turning curves, and increases lap times by opposite cause in dirt surface.

PS: All cars were also tested with similar results between mirrored tracks on Mindscape 1990 game version.