Currently head movements come directly from acceleration. For a given acceleration, head is moving X centimeters.

IRL this is a different. There are many biomechanical models of human upper body. One can find very simple models where head-torso link is simulated with springs and dampers, and head is a 5kg mass.

This is not too difficult to implement, and gives very realistic head movement. One can easily find spring and dampers value for average human body in literacy. Such spring/damper model is used in ISI engine, parameters are in the HeadPhysics.ini.

It would be great to have something like that, and a head which tries to fight lateral acceleration and to stay level when the car tilts (there was an unofficial addon script to do that and it looks very good).

Quote from Mikkomattic :

I was just racing the Mini Cooper in GT Legends and I really super enjoy the G-force effects on the camera. Seems very genuine. But it never stutters around like it does in LFS. The things I note about it is that:

Most movement seems to have a springing/bouncy effect to them, taking a second or so to 'settle', reducing their sharpness. I believe rFactor has the same thing (and I do find rFactors camera movement more convincing that LFS).

This is exactly what the spring/damper/mass model brings...smooth and more natural movement, no stuttering. If you take into account that the acceleration is transmitted to torso through a rather soft layer of foam and a flexing seat back, it adds another spring/damper. Usually, it takes 0.2 seconds for a car movement to go through seat->torso->head, and tiny movements are filtered out.

Last edited by Juls, Sat 9 Aug 2008, 10:48.

Quote from w126 :

Is it the same mechanical trail that is defined below?(http://www.tamusae.org/index.p ... w&id=29&Itemid=48)

Imagine a car sliding laterally with no polar body rotation (to make it simple) and the wheel centered. Wouldn't you have some force on the steering wheel?

Sorry my mistake. Mechanical trail is constant. I should have written: torque caused by mechanical trail is zero when wheels are centered.
Remember I talk all the time about a car rolling in usual conditions on a flat and horizontal road, not sliding. On flat road, wheels rolling not sliding, you feel no torque on the steering wheel when wheels are aligned, and when you turn the steering wheel, you feel a torque directly proportional to steering wheel angle.

Anyway, I checked with Yoda from x-sim and this is exactly how it works in LFS. FFB is proportional to steering wheel angle.

Quote from Shotglass :

logitech wheel? if so try changing the force setting to something in the 101-105% range... logitech muddles the forces round the center a bit with the standard setting

Yes, tried that, but I think this does not come from the wheel. When I turn in LFS FFB does not increase linearly with steering wheel angle. It increases more like a parabol or a sine function...do you see what I mean?

When I have time I will check it...there is a tool somewhere to see precisely which force is sent to directX. I will use a constant speed, then turn the wheel a given angle, wait for steady state and check the force strength. Then I will see if it's a line or a sin curve.

OK I checked right now...FFB in LFS increase linearly with steering wheel angle. Don't know why I was feeling it different, but it is exactly as it should be...a line first and then a bit slower increase, because it is in fact something like sin(atan(wheel angle)). Once again I was wrong

No way...the only thing missing are bumps...tiny bumps acting differently on left and right wheel...local areas of the track with slope changes like the end of T1 in SO6 if you put the left wheel on this area where the track is a bit larger. There the slope is different and you HAVE to take it into account when close from the limit. This is great!

Last edited by Juls, Mon 4 Aug 2008, 20:31.

Quote from w126 :

I don't understand this part. That would mean you can't feel any force when the wheels angle is 0, which is not true. Using cosine in that formula would make more sense to me, that is if I have even the most remote understanding of what it's all about.

I talk here about cornering forces....forces applied by the road on the wheels through the trail (mostly mechanical trail caused by caster*sin(wheelsangle)). Obviously these forces are zero when wheel is centered (in LFS and any other sims), and increase with wheel angle. Mechanical trail is almost zero when wheels are in line.

Last edited by Juls, Mon 4 Aug 2008, 20:03.

After watching carefuly what happens with my car (FWD, so geometry is supposed to have a scrub radius close from zero), I noticed:

- I do not feel most bumps through the steering wheel. I feel them through acceleration, and if I feel like the wheel is trying to shake my hands, it is in fact the contrary happening: my arms and hands try to shake the wheel because they are pushed by lateral and vertical acceleration on bumpy surfaces. Maybe in iRacing they put a part of lateral acceleration into FFB to simulate this effect.

- I can feel bumps smaller than tyre width through the steering wheel. They impact the tyre on one side only, and it is like scrub radius is not zero anymore. Smaller bumps give a lot more feeling even on cars with zero scrub radius. Maybe that is what is missing here.

- In my car, FFB strength seems to be directly proportional to steering angle. I mean if I turn the wheel 40 degrees it seems to resist two times stronger than when I turn it 20 degrees. I know this is not very logical when you think about caster effect...which is proportional to sin(angle).
There is this document, about steering wheel torque in a nascar car:
http://www.mscsoftware.com/sup ... manHaas_steering_feel.pdf
Look at figure 1. Steering torque is directly proportional to lateral acceleration (which is proportional to steering wheel angle before grip limit).
There is another document from a Renault driving simulator test. They try different laws for steering torque/steering angle dependance, and the default law they use is linear.
http://www.pervasive.jku.at/Te ... vingSimulator_Toffing.pdf
Look at figure 3.

I have the feeling FFB in LFS is not proportional to steering angle, and that is why it is very soft close from center and a lot harder further. Should it be proportional? Is it proportional most of time IRL?

Edit: Finally it makes sense for me...FFB should be directly proportional to steering wheel angle (tyres not sliding).
It comes from rack an pinion mechanism. To explain quick and dirty (I know it is very very simplified):

- Rack displacement is proportional to steering wheel angle.
Disp=c1.SteeringwheelAngle
- wheels angle are in asin(Disp/steeringArmLength)
WheelsAngle=asin(Disp/steeringArmLength)
=asin(c1.SteeringWheelAngle/steeringArmLength)
- trail, hence FFB is in sin(WheelsAngle)
FFB=c2.sin(WheelsAngle)
=c2.sin(asin(c1.SteeringWheelAngle/steeringArmLength))
=c2.c1.SteeringWheelAngle/steeringArmLength
this is directly proportional to SteeringWheelAngle.

Is it the case in LFS? For me it does not feel proportional, but it may come from the FFB wheel mechanism. IMO it feels like in LFS wheels angle is steering wheel angle divided by steering ratio. In that case...FFB is no more proportional to steering wheel angle.

Last edited by Juls, Mon 4 Aug 2008, 11:28.

- Drifting is an art, please stop hating us!

- Are you kidding? Only hotlapping is an art and drifting is pathetic!

- What? Did you say hotlapping is an art? Let me laugh! only racing is an art, and hotlapping is ugly and stupid.

- What? Did you say racing is an art? Only racing with RWD cars is an art. FWD cars suck.

etc...etc...

look at the right side....most popular fights: Chuck Norris vs God (Dieu)

Reasonably new....
GT legends from 2005 is still reasonably new.
Richard Burns is already old but it is the only rally sim we have.

But sorry, rally trophy (2001!) is OLD....if you take it every other sim since 1990 will follow

Edit: OMG there is GPL too...1998... then you need to add Virtual Grand Prix 1, which was released on Amiga in 1998 too...maybe racing sims from the 80's?

Last edited by Juls, Thu 17 Jul 2008, 07:26.

+1

Maybe this is the solution... in the WR list should be written if you use analog clutch/h shifter.
So you can sort records more precisely according to controllers used.

Because using mouse or wheel/clutch/shifter is like playing a different game.

Quote from Theafro :

Let's face it, Real-life race teams will spend thousands on chasing tiny gains like that so i guess it's fair game, If chasing tiny little margins turns you on, so be it.

Unfortunately, you can say this for anything...like timer cheats. There is always a limit between what is legal or not, the devs have to decide and try to make the sim as fair as possible.

For me, as soon as you use another piece of code (exe, script) - which is not included in the game - to go faster...there is a little problem. Scripts like aimbots have ruined most FPS games. I expected something similar would not occur with racing sims.

Rally Trophy is a *NEW* simulation?
Wow....then I am very young

This week I found a new one... Virtual Grand Prix 3

This is an online/offline single seater simulation, with a tyre model similar to LFS one (brush-like), a learning AI, it is moddable and comes with a track editor.

Unfortunately no-demo. Pay per year system.


It seems virtual grand prix sims have been there for a while (from 1998!!!)...very confidential.

Forgot the linK:
http://www.vgp3.com

Last edited by Juls, Tue 15 Jul 2008, 16:56.

Imo vgp3 has a reasonably good collision system but no visual damage.

In this crash video on youtube the guy had no brakes at all and came full speed....he wrongly configured his controls.
http://www.vgp3depot.com/forum/viewtopic.php?t=85

Here is another video with strong crashes and it looks better (be careful heavy music):
http://www.endoftheworldtv.com/vgp3/vgp3_crash.mov


Anyway...it would be better to have more videos from this sim...showing gameplay, physics...they write it has a learning AI like LFS had, but with realistic behavior.

Last edited by Juls, Mon 14 Jul 2008, 07:42.

Quote from Sir moi 407 :

Well I don't get why the track must be real to make the game more immersive

People are fascinated by names. Brands, labels... Real tracks, real cars do not bring physics realism...but "names realism". Ask someone to try the best sim in the world with a fictional track, and the most crappy sim in the world with a crappy reproduction of a real track....and in most cases the person will prefer the second one. That's the way it is.
(did you notice there are texture packs with real advertisment to improve "realism" for most moddable games, and people like it?)

Power of names. Just names. Names like Monza suggesting an entire universe and bringing the player into this universe. It is so powerful it can hide most defects of a product. (and LFS has almost nothing to hide...)

"If you having great product and using power of names together...great success to you!"
Lao Tseu

Last edited by Juls, Mon 14 Jul 2008, 07:26.

South City circuits please !!

Quote from tristancliffe :

I'm not entirely convinced a car launching itself over a small hill whilst going backwards is 'proof' of accurate aerodynamics in reverse effects, especially as the Radical isn't exactly known for it's reverse flight dynamics.

But then again, you can believe what you want to believe, if it makes you feel clever.

Wow... GOOOOOOOOOOOOOOOOOOOOOD MOOOOOORNIIIING JUUUUUUULS. Thank you Tristan, good morning too, have a nice day.

I did not put this video as a proof iRacing does aerodynamics well. It was just to show there seems to be some aerodynamics simulated in this simulation. Watching the video before I was not even sure, it could be a bug in collision system. But from the post before it seems they have worked on something.

I do not know for the tyres, but sure you heat up too fast. This is not realistic.

Last edited by Juls, Sun 13 Jul 2008, 07:24.

Quote from Crashgate3 :

Doesn't iRacing do this, abeit not in real time?

I seem to remember reading they'd put the car models in a 'virtual wind tunnel' at lots and lots of different angles, so while not actually calculating it on the fly, the sim can provide a dynamic aerodynamic affect depending on the cars speed and movement.

Sure aerodynamics can be modeled that way...with precalculation stored in a table. It is far far enough for a beginning.
http://www.youtube.com/watch?v=3N39u_pSJo8

In that case it is resonance...engineer's nightmare...bridge is strong enough to stand the wind force, but starts oscillating more and more because his natural oscillating frequency is similar to wind change frequency.

I like the very calm guy crossing the bridge, smoking, like nothing strange is happening

Quote from w126 :

F1 teams have supercomputers a thousand times more powerful then a typical gaming PC (http://www.tgdaily.com/content/view/36905/135/) and they are still very far from doing aerodynamics simulations in real time. Did I understand it correctly, do you want it in real time in a computer game? It's around 40 years away it seems, if we are lucky.

You need supercomputer to obtain very refined aerodynamics calcuations. F1 teams need to reach a very high level of precision because they need to compare aerodynamics of different shapes and chose the best one. And difference can be very small.

But racing sims do not need such level of precision. Somewhere in my engineering books there is a table with aerodynamics. It contains the aerodynamics coefficients for many different rectangular shapes and their angle to wind direction. It has been used during more than one century to estimate wind effect on complicated steel structures like bridges, towers....etc. It requires almost no calculation at all and gives a very reasonable result. This table was filled by a guy called Gustave Eiffel.

Car Sound Remixer, with a good sound pack (there is one in the thread) gives very good results.

The difference between LFS and other sims concerning sound, is that LFS builds car sound from a lower level. LFS is a sample based system too, but samples used represent basic components of car sound. Then lots of filtering and combining is needed to build the overall sound.

Other sims use higher level samples, containing already most of the engine sound, but still have to alter these samples, change their tone and frequency, mixing them.

For me the best solution would be something between LFS and other sims. More synthetized than other sims, but based on richer samples than LFS:

- a high level system based on rich samples.

- many more samples than used today in sims, for many more conditions (various levels of throttle, revs decreasing or increasing...etc)

- more filtering, not only frequency scaling of samples as now, but more advanced adjustments in frequency domain.
Starting from a limited subset of samples, the sound engine could generate many more synthetized samples at initialisation time with scrambling/frequency adjustments....etc.


In fact the main problem with LFS approach is that engine sound is far more than a serie of explosion sounds. Explosions in cylinders are the source of broad band mechanical vibrations, and these vibrations will use all parts of the car to output sound.

Many parts of the car engine are resonnant, and will output a sound at their resonnance frequency when they receive these vibrations, at very different frequency than the revs.

Trying to recreate engine sound from one explosion sound is a bit like trying to recreate a piano sound from recording the noise of the hammer hitting the rope. It ignores resonnant parts which account for most of the sound.

Quote from Taavi(EST) :

The free market between Russia and the Baltic States/Eastern Europe during the soviet union really was good, the only bad thing about it was that it had no aim, nothing would have ever came from it, no progress, the collapse would have came anyway.

For example, here we had the largest steel factory in Europe during USSR time (Nowa Huta). All steel produced here was going straight to Russia by train, and we received entire trains of mayonnaise or other useless things as payment. People here were dying of heavy pollution caused by steel factory, and nobody could find steel to make roofs. Only ***** mayonnaise, entire shops full of mayonnaise and not much to eat with it (most valuable goods were going to Russia too).

This is what you call good free market? looks very much like creative history.
Compare eastern european countries during USSR and now. Everything has improved dramatically as soon as USSR ceased to exist. Life level, health, environment...every single thing. And improvement came far before European Union integration.
There is absolutely no way someone can logically write USSR was helping eastern european countries. It is plain nonsense.

Last edited by Juls, Tue 8 Jul 2008, 12:44.

Quote from squidhead :

Everyone cries about Big Bad USSR that was "Terrorizing poor Estonians and Latvians and Ukrainians and everybody else", and nobody tends to think that their asses actually depended on Russians to save them from being wiped out during the war, and depended on USSR to feed them till '91...

Pathetic to read such things from citizens of former USSR. You either did not have to suffer in your flesh from this regime, or you have to demonstrate something for (unclear) political reasons.
Or you are French.

Talking about Poland, over a population of 30 millions people, 5 to 6 millions died during the war, including 3 millions jewish polish citizens...while the country was half occupied by USSR. Where is the protection from USSR you are talking about?

Moreover, USSR took part in the slaughtering and destruction:
Here 1.5 million people were deported in 1940-1941 by USSR....when the "big bad" USSR was an ally of 3rd Reich and they were sharing the country. Many of those who came back came 10 years later, 5 years after end of war.
USSR slaughtered 24000 military officers to ensure the country would not be able to fight for it's independance after 3rd reich fall. This is known as Katyn.
When soviet troops came to conquer Poland (what you call free), there was an uprising in the capital because people wanted to take part to this fight and remain an independant country. Soviet troops waited 3 months on the opposite side of the river to be sure every fighter was killed by SS.
This is during this uprising that Hitler decided to destroy and burn every building of the city to give a "horrible exemple" for other cities. To have an idea of what happened, watch the movie "the pianist".
Civilians families were burnt in cellars with flame throwers.
From a 1 million population, Warsaw had about 10 000 inhabitants alive when the USSR troops arrived.

When USSR collapsed, the country was left with a destroyed economy, a destroyed environment, an average 65 years life expectancy.


It's totally unfair to say 7 millions people killed by Nazis were a child's play. There is no point comparing things like that. The truth is that Nazis commited horrible crimes all over Europe, and this is good it has been studied and revealed as clearly as possible. The truth too is that large scale slaughtering and crimes against mankind were commited too in USSR and China, and should be studied and revealed as clearly as the former ones. Your comment is the proof there is a lack of information and education concerning these crimes.

Last edited by Juls, Sun 6 Jul 2008, 20:12.

Quote from JJ72 :

try a UFR on southcity it does pull around quite a lot, I think the effect is definitely there.

Yes, UFR gives a very good feeling for bumps, because it if FWD and it loses grip one side or the other because of different normal of the track under each wheel. It is called something like torque steering if I remember well.
I would like to feel bumps as well with RWD.

Anyway, looking with forces activated, curbs and grass do not have any braking effect on the wheels. It gives only vertical forces. Sand has an effect and pulls the steering wheel. Other surfaces not.
I can see that most bumps (surface normal changes) in LFS do not give longitudinal forces at all. Only very big ones like speed bumps on Kyoto.

And this time I am sure it is not my imagination

Last edited by Juls, Sat 5 Jul 2008, 14:16.

Quote from bbman :

Rubbish... Apart from the racecars, you have quite an amount of lift at speed, so you should feel LESS, as the wheels are already less loaded...

Forces caused by the track and track irregularities increase with the square of speed. It compensates easily for this lift. If you were right we could take speed bumps at 100kph as we could feel less than a 10kph.


Something is missing with bumps in LFS. More narrow bumps first. But something else too.
I am not sure what else is missing, but I can't help thinking about it when I can drive one wheel on curb or on grass without the steering wheel pulling.

And I suppose this is what they focused on with iRacing...it looks like bumps in iRacing have a different effect on the tyres/car and pull the steering wheel noticeably, and this requires correcting all the time and makes the experience more immersive.

Anyway, this is only an attempt to explain differences between sims FFB. Only one thing is sure...we need to feel tracks are less smooth in LFS, whatever it requires to change/improve.

Last edited by Juls, Sat 5 Jul 2008, 13:06.