So, is the realism of the current torque curves shapes actually important if we consider that almost nobody had extracted them for 5 years? :nol2:

Quote from Bogey Jammer :

Many things have to be modeled in engine physics to get less smooth curves (thermodynamics, fluid mechanics, pure mechanic, variable admission systems, ignition timing, cam profile and many other parameters). Anyway the common curves published in some general cars documentation are quite simplistic too. And the curves measured by dynometers are highly disputable (based on my observations )

But I admit such engine modeling would be exciting and would introduce us to engine setup !

Full engine/drivetrain modelling wouldnt be "required" in order to gain improved realism. Just simulation of realistic behaviour.

Quote from kaynd :

This isn't going to happen in any sim anytime soon... such a physics model, specialized only for the engine is extremely complicated.
"Drown by hand" power cuerves, imitating or copying similar power curves from similar real engines, would be enough for any sim released in the following decade.
One of the difficulties, would be part throttle power oughtput. But anyway any model based on data gathered from real dynos would be much easier to implement.

Agreed. IMO that is all that would ever be required. When you drive a real car you don't need to understand about why the engine responds in a certain way, you only need to understand about how it responds under certain conditions.

Quote from Bogey Jammer :

So, is the realism of the current torque curves shapes actually important if we consider that almost nobody had extracted them for 5 years? :nol2:

To the levels I stated above, yes I believe they are.

Remember producing a realistic sim does not require actual modelling of what is going on in the real world, only that the input/outputs of the model are correct in respect of real world environments.

This is a mistake I think a lot of people on here make. For example it is NOT required that every valve, spark, combustion wavefront is mathematically calculated by the sim in order for an engine response to be realistic. It is only required that input/output parameters of such things as engines and drive trains are correct. Each can be treated as a "Black box". This can be achieved far more simplistically from real world empirical measurements of (for example) engine blocks on dynos under various load conditions and at various atmospheric conditions. Sure still not simple, but far simpler than true modelling and just as accurate. At least in terms of a driving sim.

Remember, LFS is not an engine/car development package. Hence it is not required to fully model such things in order to give an accurate driving experience.

Quote from gezmoor :

It is only required that input/output parameters of such things as engines and drive trains are correct. Each can be treated as a "Black box". This can be achieved far more simplistically from real world empirical measurements of (for example) engine blocks on dynos under various load conditions and at various atmospheric conditions. Sure still not simple, but far simpler than true modelling and just as accurate. At least in terms of a driving sim.

thats right. And what differs LFS from Simbin sims is (as I understand) that Scavier assumed that you should create as many separate "blackboxes" in critical places instead of one big blackbox that mimic behaviour of the car (correct me if I am wrong). That's why they will not spend time over tweaking the whole drivetrain and leave it as simple as possible waiting for the moment when those blackboxes - driveshaft (?), gearbox, engine will be simulated as different parts.

Thanks! I really need it.

I only have one question... how do you output this data? I'm interested in know how.

Quote from KassadGLA :

I only have one question... how do you output this data? I'm interested in know how.

I too

I'm pretty sure he used a (self written?) tool to read it directly from memory and output that to a file. There is no way to make LFS output this by itself.

Quote from KassadGLA :

Thanks! I really need it.

I only have one question... how do you output this data? I'm interested in know how.

http://www.lfsforum.net/showthread.php?p=1251368#post1251368
http://www.lfsforum.net/showthread.php?p=1251556#post1251556

I think those curves would be useful to resurrect the shift light in Z25

Quote from Bogey Jammer :

I think those curves would be useful to resurrect the shift light in Z25

Yes, that is what I want to do, but I don't have skills on programming. That's why I asked you how you did, can you upload a simply program receiving torque data from 1 car in a C++ or VB6 format?

So, LFS continue sending the ShiftLigth packet by OutGauge, but I want to do a LEDs bar, so that when they light all at the optimal time to change. Rather than simply indicating revolutions. And for that I need to calculate in real time

I've made dynocharts before, it's quite easy. You just have to use a replay analyzer to output a csv, and then import it into your favorite spreadsheet or graphing application and calculate the torque from the longitudinal tire force. You could also derive the power directly from certain values in the data, but that's more difficult and I don't know how.

Quote from Napalm Candy :

Yes, that is what I want to do, but I don't have skills on programming. That's why I asked you how you did, can you upload a simply program receiving torque data from 1 car in a C++ or VB6 format?

You still need to have some programming skills to do that, and I don't use C++ nor VB.
I begun to use a self-made program in Python but was too much limited. Then I downloaded a memory hacking software, had to learn how to use it from scratch and fortunately managed to read the torque value after 5 minutes. That was easy. But for the acquisition, I used the script feature of this program, and wrote something in a hybrid custom language (again, I had to learn it from zero). All of this is quite awkward to use, and I will not provide help or tutorial to achieve torque acquisition.

Quote from Napalm Candy :

So, LFS continue sending the ShiftLigth packet by OutGauge, but I want to do a LEDs bar, so that when they light all at the optimal time to change. Rather than simply indicating revolutions. And for that I need to calculate in real time

Again you still need to program it, and find a way to retrieve current gears setup.