Question on car stats and their power.lut's

[JBolho]

Hi guys.
I have noticed something strange with some cars, i use the TorqueCreator site to calculate and adjust the power.lut for my "Street Tuned" versions and i noticed that the output of the unmodded LUT corresponds to a much lower BHP and Torque figure than what appears on the stats in CM, and official manufacturer numbers.
It didn't bother me until now because i was only messing around with mod-cars (either creating the street version of a car, like the Citroen DS3R from the R3 model for example, [personal use only of course], or creating a Tuned version from an already street/sports car model)
But today i was creating a "Street Tuned" version of the GT86 and when i input the stock LUT into TorqueCreator... it corresponds to barely more than 160bhp?? I know about that whole thing of "transmission loss" but IRL if i buy a 200bhp-rated vehicle and take it to a rolling dyno, it will have around that number, not 40hp less, that much i know from experience.
So what's happening here? Can someone explain? And i'm sorry in advance if this is a dumb question. (I guess now i know why the GT86 is such a snail in-game...)

 

[Stereo]

bhp is rated with no losses (not just the transmission, but accessories disabled), engine directly attached to a dyno. If a rolling dyno is telling you bhp, it's by using a 15-20% correction factor to account for this difference, which is just a technical way of saying "they measure wheel hp, then multiply by 1.15".

 

[JBolho]

bhp is rated with no losses (not just the transmission, but accessories disabled), engine directly attached to a dyno. If a rolling dyno is telling you bhp, it's by using a 15-20% correction factor to account for this difference, which is just a technical way of saying "they measure wheel hp, then multiply by 1.15".

Ok, thanks for explaining that part, but still, TorqueCreator from what i understand is accounting for those 20% already (at least it has a setting for that), unless it's doing nothing...
Coupled with the fact that the GT86 in particular feels so underpowered, i had to check out the files. The 500's for example don't show nearly as much of a difference.

 

[Goblin616]

I just used the tool, and put in the info of my own car, and there's no HP or Torque loss in the graph or calculator. Are you sure you used it correctly?

 

[axeracing]

I'm not familiar with the site you mentioned, but perhaps you are inputting lb-ft instead of nm? AC uses newton-meters.

 

[JBolho]

I just used the tool, and put in the info of my own car, and there's no HP or Torque loss in the graph or calculator. Are you sure you used it correctly?

Yes. What surprised me was, giving the GT86 example, the in-game stats say that the car has 200hp, but if you put the LUT figures in the tool, the result is barely 160hp. That's what I found to be strange.
Furthermore, I don't know about you guys but for me, the GT86 simply doesn't feel like a 200hp car. After some digging I reached a possible conclusion...
I think the devs tried to recreate the transmission loss of the cars using lower LUT values but ended up unrealistically hurting the performance of some cars compared to their real life counterparts. The 500 Abarth for example, feels much like the real thing, and examining the LUT values, the difference in hp is much much lower than the GT86 case.

 

[JBolho]

T

I'm not familiar with the site you mentioned, but perhaps you are inputting lb-ft instead of nm? AC uses newton-meters.

The tool uses nm too.
I'll continue experimenting using different methods of balancing LUT values, real world figures, and perceived performance. I'm recreating my Golf GT and that will give me a good idea of what I'm doing right or wrong.

 

[Goblin616]

Ah... see I put in the values from my own car just looking up the numbers. I'm not sure how or what you're doing. IDK if it makes much difference, but around here the car is advertised as having 205hp. I've been looking for info, and found that at the time the car had 200ps (197hp,) but since 2017 has 208ps (205hp.) Not a big difference. It also makes 205Nm of torque.

 

[Goblin616]

I also know what you mean that it doesn't feel very powerful, but the GT86 is not a very powerful car, just a well balanced one. It weighs more than a comparable MX-5 (ND) - but I've never tried drag racing against them. Maybe I'll try that now and see.

 

[Neilski]

i noticed that the output of the unmodded LUT corresponds to a much lower BHP and Torque figure than what appears on the stats in CM

I don't have CM but I know that it can recalculate the torque and power curves - have you tried that?

I just checked the power.lut and the ui_car.json, looking mainly at the 7000 rpm figures.
The LUT torque figure translates to 168 bhp. The JSON torque figure is higher, and translates to about 189 bhp, but the JSON power figure is 200 bhp... Bit nuts really.

 

[JBolho]

I don't have CM but I know that it can recalculate the torque and power curves - have you tried that?

I just checked the power.lut and the ui_car.json, looking mainly at the 7000 rpm figures.
The LUT torque figure translates to 168 bhp. The JSON torque figure is higher, and translates to about 189 bhp, but the JSON power figure is 200 bhp... Bit nuts really.

The recalculation option affects only the power graph on the car stats, not the performance itself, it's mainly to keep the data coherent on the car stats section.

 

[Neilski]

The recalculation option affects only the power graph on the car stats, not the performance itself, it's mainly to keep the data coherent on the car stats section.

Yes, but I was referring to your comment that the calculated power didn't match the CM stats:

the output of the unmodded LUT corresponds to a much lower BHP and Torque figure than what appears on the stats in CM, and official manufacturer numbers.

If you hit the button to recalculate, they should match, unless something weird is going on.
Edit: but the other part of my message was that the vanilla AC figures are clearly wrong and not even vaguely self-consistent.

 

[aphidgod]

I can't speak for rando internet tool I've never used, but both AC and IRL are working properly.

200bhp ought to result in around 170whp in a RWD car. You say it's showing 168 in the LUT... that seems pretty perfect to me.

The UI graph should show BHP, not WHP, but the physics engine still needs WHP. (Actually WTQ, but you get the idea.) They're not supposed to match up, so again, sounds pretty perfect to me. Those UI graphs aren't always super precise either, unless you use CM to recalculate them from the actual physics data, so don't waste too much time worrying if they're off a bit. They don't matter. Just part of the UI.

 

[Goblin616]

Game engines have to simiulate reality. I wonder if they sometimes modify values from the real for the sake of the game engine to properly simulate relative performace between different objects. In that case they may have to skew the numbers a bit to accomplish this. I always wondered this about (combat) flight simulators as well.

 

[aphidgod]

AC is really quite good about giving correct results when you give it correct numbers. It's not perfect, nothing is, but this part of it needs no "translation" IME.

 

[Neilski]

I can't speak for rando internet tool I've never used, but both AC and IRL are working properly.

200bhp ought to result in around 170whp in a RWD car. You say it's showing 168 in the LUT... that seems pretty perfect to me.

The UI graph should show BHP, not WHP, but the physics engine still needs WHP. (Actually WTQ, but you get the idea.) They're not supposed to match up, so again, sounds pretty perfect to me. Those UI graphs aren't always super precise either, unless you use CM to recalculate them from the actual physics data, so don't waste too much time worrying if they're off a bit. They don't matter. Just part of the UI.

Did you actually check the data for this car? The UI torque figure is not consistent with the UI power figure. How can that be considered correct?

 

[aphidgod]

As far as the UI goes... if the power and torque graphs really don't jive mathematically, that's a UI problem because obviously there's only one objectively correct answer there. One is a function of the other. I guess use CM to recalculate them.

Physics-wise, AC only accepts wheel torque in nm via power.lut for the power curve so there's no way it could possibly have mismatched power and torque curves in game. Be correct in your power.lut and you'll be fine. Whether or not Kunos got their numbers correct I cannot say, but 200bhp ought to present about 170whp. Perfectly normal for a RWD car to lose ~15% to its driveline. (My WRX loses ~19% to its AWD.)

Again, though, the UI graphs have no bearing on the physics. They could read eleventy-billion horsepower and the actual car wouldn't be one millisecond quicker. If that's the only issue, just recalculate them.