So, one of the reasons I've not posted much about work in progress for a while is that a lot of it has been hard to illustrate. The prime example is, well, this post below from quite a while ago:
I think the issue with the MR2 and Vantage is that their redline and max RPM are set to the same value, so the game stretches the curve another 500rpm and thus the peak power is too high. I'll probably just dump a list of all cars with the same dodgy setting and go through them all at some point.
The best way to show this issue is with some power / torque curve graphs, because it gets fun. Warning: lots of graphs ahead!
A normal torque curve looks like this, from the Megane:
The purple line is the power, the blue line is the torque (multiplied by 10), the horizontal green line shows the peak power (147ps), the vertical red line is the engine redline (6500rpm), and the end of the graph is the rev limiter (7000rpm).
However, some cars in GT2 have the redline and max RPM set to the same value, for example the New Beetle, which should have a curve that looks like this:
Unfortunately, what the game actually does is increase the max RPM by 500 and duplicate the last torque curve point to fill in the gap. As a result, the Beetle's curve really looks like this:
Instead of the correct 115ps @ 5000rpm, the peak is now 117ps @ 6500rpm, so the power ends up increasing again after the redline.
As planned, I wrote a bit of code to list every car affected by this, and... there are over a hundred of them. There also wasn't a one-size-fits-all fix, each car had its own specific solution depending on how it was affected.
For the Beetle, it was fortunately simple - just add a real point to the torque curve at 6500rpm, and set it to something sensible to continue a smooth curve down:
The MGF on the other hand was even simpler - indicated redline should actually be 7000, so I could just reduce the redline and leave the torque curve intact. Before and after:
The Audi TT was one of several cars that weren't noticeably affected as the power spike was lower than the real peak power. However, the power dropped off so unrealistically and dramatically before redline that it harmed performance with auto gears, so instead of changing the spike I smoothed out the drop instead:
This issue actually hadn't gone unnoticed by Polyphony either. The Ford Focus Ghia makes less torque than advertised, which it turns out seems to have been done to scale the max power down to the advertised 130ps. Whoever set the car up must have noticed the incorrect power but not been aware of the game engine messing with the curve, so reduced torque until they got the right power number to show up in the game. The fix here was to add a real point at 7000rpm, then scale the entire curve up until the peak torque was correct, at which point the car made the correct 130ps again:
Some cars were especially badly affected, the Viper GTS shows up in the garage as a 470ps car, but in fact is only really making 440ps before redline and so isn't as fast as would be expected from the displayed figures. For this one I reworked the top end of the curve entirely to hit the real car's quoted 450bhp (454ps). The expected curve, the actual curve, and the revised curve:
Just to show how it appears in the game, here's the Viper's original curve:
Notice that peak power is listed as being at 6000rpm, when the chart itself ends at 5500rpm. The game only graphs the points listed in the data, while the power calculation includes the imaginary extra point added to handle the max RPM increase. It's also quite noticeable that the graph is below 450ps despite the peak power showing as 470ps.
After adding a real point at 6000rpm, it shows up on the graph and the peak power RPM shows the intended 5000rpm:
Unfortunately, things hadn't finished snowballing at this point. I went to fix the '69 Camaro Z/28 and found that it actually has a torquey and low-revving '69 SS 350 engine in it instead of the peaky and revvy Z/28 race homologation engine. I then noticed it had a Mustang transmission in it instead of a Chevy, and realised that I really needed to dump all of the fact-checkable data into a spreadsheet and audit it, like I did for GT1+.
That spreadsheet has been an ongoing project for quite some time now, poring through scans of old brochures and other sources trying to verify values and finding pages and pages of new things to fix. Here's just a tiny snippet of a few rows for fun:
As things stand, I've fixed the torque curve and/or redline for all of the 100+ cars affected by the max RPM issue, but I probably won't make a real start on all of the other stuff I'm finding until I've researched at least most of it.
Just as a sample of the bizarre stuff coming up from one manufacturer alone... the Ford RS200 has an Evo V RS transmission in it, the Escort WRC has the listed dimensions of the entirely unrelated US-market Escort LX, the Taurus has a Toyota Chaser auto, the '99 Mustang GT has the basic 16" tyre sizes but is modelled with 17" wheels... and so on. I'm actually planning to reformat the changelogs to make them more readable due to the amount of changes to come.
