It would explain a lot if pneumatic trail is missing. I suppose tyre patch does not twist (or only at very low speed), but takes a trapezoid shape every time the track normal changes from vertical (on bumps or slopes), and is no more symetrical->steering torque.
I read somewhere nkPro was modeling pneumatic trail, and maybe that is why you can feel a lot more the bumps in nkPro.
Very well explained here: http://www.lfsforum.net/showthread.php?p=539489#post539489
Last edited by Juls, .
I do not think it is so easy. When steering wheel is centered and left tyre meets a bump while right tyre stays on flat surface, there is a stronger force applied to the left wheel. OK.
But remember steering wheel is centered...so tyre contact patch is almost centered too...and the force coming from the bump should not give a steering force because pneumatic trail and mechanical trail are aligned with the force and there is no torque resulting. It was stated here several times that you almost do not feel bumps in LFS when steering wheel is centered because forces applied on wheels give almost no resulting streering torque in such conditions. But from RL it seems something is missing...I can clearly feel bumps with my road car, soft tyres and soft springs, and sure it can not be explained by camber.
The steering force you feel on a bump when steering wheel is centered can only be explained if there is a rather strong change of the tyre contact patch shape on the bump.
Last edited by Juls, .
I noticed that IRL, I can feel bumps even if my wheel is centered.
Do not know where it comes from physically, but bumps do pull the wheel in one direction even on straight line where wheel is centered. It seems that when road is not symetric (for example a bump under left wheel and flat under right wheel) it has a strong effect on steering force.
In LFS you can almost not feel bumps when the wheel is close from center. Or bumps impact on steering force is too small to be noticed at 50% FFB. Or maybe tracks do not have enough bumps. There is something to improve there.
I suppose iRacing does not try to keep all the bumps in the track mesh. It seems to me track mesh is rather simple, and surface information is kept somewhere else....so that they can keep very detailed track normal map without needing a very detailed track mesh. It seems to me the best way to keep very detailed info from the laser scan without killing the framerate.
Do not know where it comes from physically, but bumps do pull the wheel in one direction even on straight line where wheel is centered. It seems that when road is not symetric (for example a bump under left wheel and flat under right wheel) it has a strong effect on steering force.
In LFS you can almost not feel bumps when the wheel is close from center. Or bumps impact on steering force is too small to be noticed at 50% FFB. Or maybe tracks do not have enough bumps. There is something to improve there.
I suppose iRacing does not try to keep all the bumps in the track mesh. It seems to me track mesh is rather simple, and surface information is kept somewhere else....so that they can keep very detailed track normal map without needing a very detailed track mesh. It seems to me the best way to keep very detailed info from the laser scan without killing the framerate.
Last edited by Juls, .
For the uf1 I know who was the dad, and I have a candidate for the mother:
http://www.caradisiac.com/medi ... /collection/renault_5.jpg
Until now, people "seriously tired at the lack of content and development speed" had no alternative. They had to wait, use other sims, and come back to buy LFS when a new release is ready. They were never lost for LFS community, because no other sim was able to give a similar feeling/online racing.
Now, it seems to me there is an alternative...iRacing. I did not try it, but from reviews it seems it is good enough (and updated fast enough) to prevent tired people from coming back to LFS after trying it. And as soon as you start to pay for one year, it costs so much that you have to use almost only that sim and no other anymore. It concerns only a minority of LFS users because of the cost, but it is a threat. Things start moving!
There is a tool called glovepie which already allows to emulate trackir using a wiimote.
http://video.aol.com/video-det ... wiimote-trackir/466746856
http://video.aol.com/video-det ... wiimote-trackir/466746856
freetrack is supposed to support tracking using a wiimote very soon. They are beta-testing this feature.
With a wiimote located below your display, and two tiny IR reflectors or leds on a cap, you should be able to have a head tracking with 100 or more updates per second (currently used webcams do 30 updates per second) without using the CPU as tracking is performed by wii hardware (currently tracking is performed with heavy CPU calculations).
With a wiimote located below your display, and two tiny IR reflectors or leds on a cap, you should be able to have a head tracking with 100 or more updates per second (currently used webcams do 30 updates per second) without using the CPU as tracking is performed by wii hardware (currently tracking is performed with heavy CPU calculations).
I would really like to have the same kind of job. Because in my job, however fast I work, people paying me always agree I work too slow and none of them give reasons to defend me
It should be a tool, not an authority. Servers can choose to use it or not.
Servers using the safety rating accept only drivers with their safety rating in the specified range. With this system you can create servers accepting everyone but recording statistics for safety rating, or servers with safety rating completely disabled, or more restricted servers for good drivers...etc. That way nobody has to drive offline only and it is always possible to recover from a bad safety rating.
Servers with safety rating enabled should record some crash information every time there is a crash/spin, and send this information to LFS master server. This information then is stored in database. And if safety rating rules are changed, it is possible to compute again safety ratings for everyone using new rules on stored data. This is how it works for iRacing, because they wrote somewhere they changed the way safety rating is calculated and updated it for everyone...looks better than a safety rating reset everytime there is a change in rules.
It seems iRacing does not try very hard to find which driver is guilty. You can easily be victim of a wrecker and get a penalty. It seems unfair at first glance but the most important thing is that wrecker gets a penalty every time he is involved in a crash. On the long term, a wrecker will be involved in more crashes than a good driver, and his average rating will be lower...even if the good driver received several undiserved penalties. It works a bit like car insurances Bonus Malus System (No Claim Discount in UK). Most of time you get a penalty if you are guilty or not in a crash...seems an unfair system...but finally it makes a difference between good and bad drivers.
Last edited by Juls, .
Good question. Safety rating is an average. If you are a wrecker, you should be involved in many more crashes than a careful driver...on average safe drivers should have a better safety rating than others, even if sometimes they can have a penalty because of another guy.
I suppose with time they will be able to refine the safety rating, and decide automatically which crash is really a fault or not. From the trajectory of vehicles before the crash it is possible to identify typical crashes for which one driver is clearly guilty.
Sorry, but ban currently works like that: you almost lose control and slow down in front of a guy trying to do a qualifying time...you did not wreck, you did not touch him, he is furious, he calls for a ban, 10 other guys vote yes without knowing why or what, and that's it, you get banned.
The kick/ban vote is not satisfying.
There should be a safery rating...server can use it or not. If they use it, they specify which range of safety rating drivers is accepted, and race time statistics are saved to update safety rating. It would ensure high quality racing if you want it, and more relax racing when you want too, on servers where it is disabled. After all, safety rating is something like an automated kick/ban system....less prone to emotions.
Take LFS:
- a constantly evolving online racing simulator
- a pay per release system
- a 3 persons dev team
Make it big, full time hard job. Increase the dev team to several dozens. Then you are obliged to ask users to pay more often than every release...pay per month for example. And you have to be sure to attract many users....a good idea is to have real cars and tracks, people like it.
What you get is iRacing. iRacing is almost LFS concept with a large company and investment. Both sims can be very successful, they offer similar concept for different simracers.
For me LFS should increase dev pace (I suppose it requires increasing dev team)...iRacing is like a big factory, maybe too big, but sure LFS looks sometimes too small. LFS product is mature and is asking for more.
- a constantly evolving online racing simulator
- a pay per release system
- a 3 persons dev team
Make it big, full time hard job. Increase the dev team to several dozens. Then you are obliged to ask users to pay more often than every release...pay per month for example. And you have to be sure to attract many users....a good idea is to have real cars and tracks, people like it.
What you get is iRacing. iRacing is almost LFS concept with a large company and investment. Both sims can be very successful, they offer similar concept for different simracers.
For me LFS should increase dev pace (I suppose it requires increasing dev team)...iRacing is like a big factory, maybe too big, but sure LFS looks sometimes too small. LFS product is mature and is asking for more.
OMG, this is the Grail...I have been searching so long for this information:
" to work out a car's orientation matrix, the calculation order is : roll,
pitch, heading
to work out the acceleration or velocity in the car's coordinate system, a
programmer should create horizontal (x) forward (y) and up (z) vectors
(using the heading, pitch and roll) and take the dot product (scalar
product) of these with the acceleration or velocity vectors supplied in the
outsim packet."
Maybe it could be somewhere in documentation (maybe it is?), because rotation order has to be known to properly convert world coordinates into car coordinates. I suppose most 3D gurus know about this order...but at least for me it was not so obvious (and angles come in the opposite order in OutSim packets!)
Here is the rotation matrix with this rotation order:
http://planning.cs.uiuc.edu/node102.html
edit: in the formula above you have to switch pitch and roll angles. Apply this matrix with alpha=LFS heading, beta=LFS roll, and gamma=LFS pitch and it seems to work fine.
" to work out a car's orientation matrix, the calculation order is : roll,
pitch, heading
to work out the acceleration or velocity in the car's coordinate system, a
programmer should create horizontal (x) forward (y) and up (z) vectors
(using the heading, pitch and roll) and take the dot product (scalar
product) of these with the acceleration or velocity vectors supplied in the
outsim packet."
Maybe it could be somewhere in documentation (maybe it is?), because rotation order has to be known to properly convert world coordinates into car coordinates. I suppose most 3D gurus know about this order...but at least for me it was not so obvious (and angles come in the opposite order in OutSim packets!)
Here is the rotation matrix with this rotation order:
http://planning.cs.uiuc.edu/node102.html
edit: in the formula above you have to switch pitch and roll angles. Apply this matrix with alpha=LFS heading, beta=LFS roll, and gamma=LFS pitch and it seems to work fine.
Last edited by Juls, .
No problem, I always steer clear of these servers after the ban
For me the best thing would be pseudo-randomly generated bumps.
When you select the track, you give a few parameters, like average bump height, average bump length, average bump width, and a large number which is used to seed the pseudo-random algorithm.
And it takes the track and generate bumps at loading time, slightly moving track points. And the large seed number ensures that bumps are the same for everyone with the same parameters. Using the AI path it can generate very realistic bumps you get on a real track (more often in turns, oriented according to path...etc).
And using different seeds give different bumps-> infinity of variation for the tracks. Every set of parameters give a unique track,everyone has the same on a given server.
When you select the track, you give a few parameters, like average bump height, average bump length, average bump width, and a large number which is used to seed the pseudo-random algorithm.
And it takes the track and generate bumps at loading time, slightly moving track points. And the large seed number ensures that bumps are the same for everyone with the same parameters. Using the AI path it can generate very realistic bumps you get on a real track (more often in turns, oriented according to path...etc).
And using different seeds give different bumps-> infinity of variation for the tracks. Every set of parameters give a unique track,everyone has the same on a given server.
It happens very often too on non-oval racing servers. From my experience, one guy gets furious because of you crashing, and he starts a ban vote, and 10 other guys vote yes without asking for any reason.
This side of LFS is not very well balanced...but it is a difficult problem to solve. Same issue for all multiplayers games: how to prevent morons from ruling.
This side of LFS is not very well balanced...but it is a difficult problem to solve. Same issue for all multiplayers games: how to prevent morons from ruling.
I suppose X axis is the position on the track, not the time. So for every track point, you know what was the time difference between the two drivers. Looks nice and very useful!
It seems to me that this time difference graph is the integral of 1/v1-1/v2, that is to say (v2-v1)/(v1*v2).
Intuitively it looks right, it increases or decreases when speed difference is positive or negative, and slow areas of the track (like turns) makes the biggest time difference (v1*v2 small).
Reading this graph before and after a turn you know how much time you lose or gain during that turn.
Does it improve framerate for autocross layout objects too ?
It seems I did not understand at all what a time diff graph is
Could you explain more what is a time diff graph? (what do you take for X and Y axis)
Could you explain more what is a time diff graph? (what do you take for X and Y axis)
I see what you mean...a graph that shows the black spots where you lose most of time compared to a WR.
You get exactly that when you display speeds in synchronized mode. And it is better than a time difference, because it shows directly what should be your minimum speed in corner to go as fast as WR.
You get exactly that when you display speeds in synchronized mode. And it is better than a time difference, because it shows directly what should be your minimum speed in corner to go as fast as WR.
Last edited by Juls, .
Brilliant tool!
Very intuitive, very useful, fast and nice.
Other sims do not have a user friendly, integrated analyser.
This is another strong point for LFS and should be proudly advertised !
Very intuitive, very useful, fast and nice.
Other sims do not have a user friendly, integrated analyser.
This is another strong point for LFS and should be proudly advertised !
Hello, I am working on a motion simulator and would like to have it running with LFS.
I can see here accelerations LFS sends are expressed in world coordinates, not in car coordinates, and have to be processed.
I saw somewhere the formula to do it in a post from Scawen, maybe one year ago, but I do not manage to find it anymore. (when I saw it I was thinking "nice, it is there...I will find it easily when I need it...")
Maybe someone who have it could post it again? Thank you.
I can see here accelerations LFS sends are expressed in world coordinates, not in car coordinates, and have to be processed.
I saw somewhere the formula to do it in a post from Scawen, maybe one year ago, but I do not manage to find it anymore. (when I saw it I was thinking "nice, it is there...I will find it easily when I need it...")
Maybe someone who have it could post it again? Thank you.
DMR is a very good server. Best place for LX4, no doubt! For me the best FFB/sense of speed/immersion in LFS comes with LX4/LX6 on South City circuits.
I think some people racing online are in fact AI drivers pretending to be human beings. Online gaming is a very good way for AIs to heal their inferiority complex that way. Personnaly, I always ask AI drivers to do that when they feel down after reading the forums. We find together some very simlpe algorithm to imitate bad drivers behavior, and it helps them a lot to know some humans can be simulated with such simple algorithm:
1- accelerate
2- if bump then ban
3- goto 1
I think some people racing online are in fact AI drivers pretending to be human beings. Online gaming is a very good way for AIs to heal their inferiority complex that way. Personnaly, I always ask AI drivers to do that when they feel down after reading the forums. We find together some very simlpe algorithm to imitate bad drivers behavior, and it helps them a lot to know some humans can be simulated with such simple algorithm:
1- accelerate
2- if bump then ban
3- goto 1
Last edited by Juls, .
Good idea, and it should be impossible to ban a user before he was kicked first.
And the kick/ban vote should have a reason written on it.
It was so quick I did not even see it before I was banned so I did not have opportunity to see what was wrong and change car.
A popup should appear: A kick/ban vote has been issued against you for the following reason:....
The "funny" thing is that virtual racers are far less tolerant toward new drivers doing mistakes than real drivers risking their lives. I suppose having to brake during qualifying because the previous car is slower than you is common?
Last edited by Juls, .
proper AI needed quickly
Today I wanted to use the LX4/LX6 online. Unfortunately there is no server with a proper combo. So I chose a server allowing all cars...it seemed to me such a server is a place where people can use any car.
Haha, what a mistake...as soon as I took LX6 I was banned. All this is boring. When I have network lag, other people think I am drifting and ban me. When I lose control (on a training server, not even in a race)...they think I am a wrecker and ban me because one guy had to brake a bit to avoid me. When I chose a different car and stay behind just to learn better to keep my distance and drive cleanly they ban me without any warning message. Why do they allow it then?
Please improve again AI so that I do not have to play online anymore.
Haha, what a mistake...as soon as I took LX6 I was banned. All this is boring. When I have network lag, other people think I am drifting and ban me. When I lose control (on a training server, not even in a race)...they think I am a wrecker and ban me because one guy had to brake a bit to avoid me. When I chose a different car and stay behind just to learn better to keep my distance and drive cleanly they ban me without any warning message. Why do they allow it then?
Please improve again AI so that I do not have to play online anymore.
FGED GREDG RDFGDR GSFDG