Re: [Torcs-devel] Force feedback integration : need help about physics

[Bernhard W. <be...@bl...>]

Hi Jean-Philippe

>>> Currently working on force feedback integration in Torcs,
>>> I am very close to the moment when I'm able to inject some effects
>>> into my wheel.
>> Since we already use the Pacejka formula, adding the aligning force is
>> trivial.  I hope that it works correctly.
>>
>> I guess I can export a couple of variables in some structure..
>> How about
>> struct SteeringFeedback {
>> 	float align; // magnitude of the horizontal force
>> 	float bump; // magnitude of the vertical force
>> 	float bump_frequency; // in case the force updates are too slow
>> };
>>
>> Now, there bumps may be a bit tricky.  The simulation's dt is currently
>> 0.002 if I remember correctly.  I think that the control feedback is set
>> to 0.01, though I could be wrong.  In that case, you wouldn't need a
>> bump_frequency, just a magnitude for the vertical forces and then how
>> that is handled would depend on the actual feedback wheel used.
> 
> If I understand correctly (raceengine::ReUpdate/ReOneStep), the physics engine
> call period can't be lower than 0.002s (RCM_MAX_DT_SIMU), and the human input
> device read period can't be lower than 0.02 (RCM_MAX_DT_ROBOTS).
> As I want to update feedback force in the device each time Torcs reads it
> (it's easy to implement), this means that we can't render a vibration
> due to a run on a -say - 0.5 m period kerb, if the speed of the car
> is higher than 45 km/h (if Shannon is right ;-)
> (2 * car speed (m/s) / bump period (m) < feedback force update frequency)
> 
> Or 1m period kerb at 90 km/h ...
> 
> It'a a bit poor I think (if I am correct).
> 
> On the other hand, it seems that my wheel can't render periodic forces
> under 50 or 40 ms period ... so, it may be poor, but the device can't do better.
> 
> And of course the final question is : what a real car mechanical steering chain
> can render, as far as high frequency vibrations are concerned ?
> I don't know that ...

It does actually not matter, this is the wrong question, IMHO. The question is 
how far can human perception distinguish between right and wrong feedback, and 
how realistic is force feedback at all (I would guess you have no such or just 
little feedback on the steering wheel, what really gets the vibration is your 
whole back/body.

> After some researches about force feedback over other racing cars simulation
> games, I tried rFactor (and found it a dangerous competitor against Torcs (;-).
> In rFactor, the aligning force injected to the steering wheel decreases
> as the grip on the front wheels loose the grip (with some mysterious damping
> formula) ... I don't know if this "effect" will already be in the "Magnitude
> of the horizontal force" you would export from Torcs physics engine ...
> but if not, it would also be very interesting to export some kind of grip value
> on the front (=steering) wheels.
> 
>> I think I'll add the structure in interfaces/car.h and
>> add it to the tPrivCar structure.
> 
> I can imagine you are very very very busy with family, work and many passions
> (this is interesting life ;-) ... but could you give us an idea
> of when you could be able to release such a patch ?
> Or if not, could you find some time to give us enough tips about simuV2/V3
> to enable us to try and achieve it ourselves ?

Ok, to keep it short, I am not interested at all in force feedback for TORCS 
currently, I think somebody would first better invest time to integrate a 
portable force feedback API into SDL (look at the README for "my" priorities). 
Here quickly the reasons:
- If you want "this" kind of simulation go to the shop and buy RFactor, Race07, 
GTR XX or whatever.
- The "unique selling proposition" of TORCS is its simplicity, portability, 
performance and stability. Because it is small a developer can quickly dig into 
the sources and achieve its customization (e.g. for a certain type of research, 
just for fun, whatever).
- I have seen lots of people which hooked in an left, so the TORCS core has to 
stay small to be maintained over the long run, almost nobody is able or willing 
to do the "dirty work" (e.g. look at Trigger, it is nice but nobody takes 
currently care of it, IIRC).
- Regarding force feedback: I bet the ordinary TORCS user has no steering wheel 
at all, so the feature is mainly a mainteinance burden with almost no benefit.

If you do not agree with this, you are of course free to create/distribute your 
own fork/patch of TORCS.

Bernhard