Well, the spike in iaStore.sys that was freezing up LFS on my system was only about 30%. In fact, until I widened the CPU History column in Process Manager, the blip was too small to see (only a few pixels high), but it was enough to jack-up LFS.
I would look for any activity that lasts about as long as your freezup in LFS.
If it turns out that csrss is the only culprit, you can go into the thread view to see if there is anything unusual going in in csrss. Set the update interval on PM to 5-10 seconds, so that the offending thread will still be on top when you are able to tab back to PM.
That's all I can think of ATM. Good luck, man.
I don't get it. We are all playing S2 already, aren't we?
Who cares when they label it "done"? What does that even mean? Live for Speed is an ongoing project. Download the latest version, play it, enjoy it, and if you don't like... try to find something better (good luck).
Who cares when they label it "done"? What does that even mean? Live for Speed is an ongoing project. Download the latest version, play it, enjoy it, and if you don't like... try to find something better (good luck).
Last edited by Eric Tetz, .
I had a similar problem that was bugging me for months. In my case it was Intel's Application Accelerator causing periodic .5 second freezes. That's probably not your issue, but the steps I took to troubleshoot my problem may work for you, too.
1. Download Process Explorer. It's tiny and free and doesn't install any garbage on your system (doesn't even need to be installed). This is a more advanced version of Task Manager. It will let you get detailed CPU/disk usage information about processes, modules (DLLs) loaded by a process, even individual threads within a process.
2. Right-click on the column view on the right and click "Select Columns". Go the the Process Performance tab and check CPU Usage and CPU History. This will give you a seperate CPU and disk-usage history bar for each process in your system.
3. Go back to LFS and play. When it freezes up, quickly TAB back to Process Explorer (as soon as Windows will let you) and look at the history bars. Look for a CPU or disk-usage spike in one of the processes, corresponding to the freezup in LFS.
From there, how you troubleshoot depends on the process that's causing the problem. In my case, I noticed that the 'System' process had regular CPU spikes, at 90 second intervals or so. I had to find out what part of the System process was causing the problem.
If you right-click on a process and select Properties, it will give you a larger view of that processes CPU/disk usage. It will also give you a Threads list, which gives you CPU usage for individual threads within that process. At this level of granularity, it's actually showing you how many context switches (number of times that Windows gave CPU time to that thread) per update interval.
I noticed that a driver named iaStor.sys was getting an inordinate amount of context switches when my LFS freezes were happening. I googled iaStor.sys (Process Explorer has Google support built-in) and found that iaStore.sys is Intel Application Accelerator. I update the driver for that, and my problem went away.
No way I could have solved this without Process Explorer.
Oh, I should mention that this is a sort of last resort troubleshooting method which is done after you do the more obvious stuff, namely:
1. Open Task Manager or Process Explorer and shut down every process that is not needed to run windows. You can use MSCONFIG to prevent a lot of these from starting in the first place.
2. Open Services (Control Panel, Administrative Tools, Services) and shut down every service that is not needed to run Windows. You may have to look up some of these to determine what they do and if you need them, but it's time well spent.
I do audio recording on my machine, so making sure I have no background tasks interrupting my recordings is an absolute must. My task list is pretty much empty, and I only have about 10-15 services running, most of which are core, required Windows services.
1. Download Process Explorer. It's tiny and free and doesn't install any garbage on your system (doesn't even need to be installed). This is a more advanced version of Task Manager. It will let you get detailed CPU/disk usage information about processes, modules (DLLs) loaded by a process, even individual threads within a process.
2. Right-click on the column view on the right and click "Select Columns". Go the the Process Performance tab and check CPU Usage and CPU History. This will give you a seperate CPU and disk-usage history bar for each process in your system.
3. Go back to LFS and play. When it freezes up, quickly TAB back to Process Explorer (as soon as Windows will let you) and look at the history bars. Look for a CPU or disk-usage spike in one of the processes, corresponding to the freezup in LFS.
From there, how you troubleshoot depends on the process that's causing the problem. In my case, I noticed that the 'System' process had regular CPU spikes, at 90 second intervals or so. I had to find out what part of the System process was causing the problem.
If you right-click on a process and select Properties, it will give you a larger view of that processes CPU/disk usage. It will also give you a Threads list, which gives you CPU usage for individual threads within that process. At this level of granularity, it's actually showing you how many context switches (number of times that Windows gave CPU time to that thread) per update interval.
I noticed that a driver named iaStor.sys was getting an inordinate amount of context switches when my LFS freezes were happening. I googled iaStor.sys (Process Explorer has Google support built-in) and found that iaStore.sys is Intel Application Accelerator. I update the driver for that, and my problem went away.
No way I could have solved this without Process Explorer.
Oh, I should mention that this is a sort of last resort troubleshooting method which is done after you do the more obvious stuff, namely:
1. Open Task Manager or Process Explorer and shut down every process that is not needed to run windows. You can use MSCONFIG to prevent a lot of these from starting in the first place.
2. Open Services (Control Panel, Administrative Tools, Services) and shut down every service that is not needed to run Windows. You may have to look up some of these to determine what they do and if you need them, but it's time well spent.
I do audio recording on my machine, so making sure I have no background tasks interrupting my recordings is an absolute must. My task list is pretty much empty, and I only have about 10-15 services running, most of which are core, required Windows services.
Last edited by Eric Tetz, .
Wow, I've never done that. I can hardly wait to try.
Any other secret LFS enhancing tips I should know about?
Seems like a fun course, but it's extremely hard to see where the course is. You have to go super slow at first, staring at the white lines on the ground, so you can memorize where the turns are.
I like the cone setup in training track #36 "Autocross Course - MRT". They use orange cones throughout, with a striped cone + blue cone lying on it's side to indicate the inside of each turn. In a few cases they use yellow cones to indicate turn direction or a dead end. The whole thing is super visible, even on your first time through.
Yes, as long as you don't create a profile for LFS, it should use the global settings.
BTW: I just found that I had the FF settings to 50% in the Profiler, and 100% in LFS. I just reversed that (100% in the Profiler, 50% in LFS) and found that I much prefer the way it feels. Just an FYI, in case anybody else wants to try it. I seem to get stronger FF with the wheel turned, without excessive notchiness in the center.
You ought to try the G25 before making sweeping proclamations about "current wheels". It's about three times faster than the DFP, and has far less internal resistance to rapid steering inputs.
That's exactly how you drive the G25. It turns faster than I would want to have to spin a wheel on my own. Clearly it's not completely realistic, but it's a huge step up from the DFP.
You have to force it there anyway, with your non-FF wheel. You've just chosen one form of unrealism over another.
Letting the wheel slide through your hand is unrealistic (with most modern FF wheels) because it's too slow, but at least it's an approximation of the real thing. Tossing a non-FF wheel around like you're playing Super Sprint seems to me just as great a compromise -- making the "sim" feel like an arcade game, no resistance, no weight, no sense that the wheel is attached to an actual vehicle, with your steering inputs totally not reflecting what you would be doing in a real car.
The compromise I chose when using the DFP was to greatly reduce the steering lock compared to the actual vehicle. With the G25 I use up to 720 degrees and it feels great. I accept that it's not 100% realistic, but IMO the alternative is even less realistic.
In short, I don't see how you're being any more of a purist by choosing to drive a wheel that you can toss around like it's attached to nothing. We're both choosing compromises; which is preferable is a matter of taste.
Last edited by Eric Tetz, .
There will never be an even playing field. Even if you only allow racers with wheels, some of them will have $500+ wheels while others have $30 wheels; some will have $5000 super machines, others will be running on their $500 Dells.
That said, I agree that the server operator should be able to enforce restrictions in cases where players are choosing options that increase the difficulty of the simulation.
For example, let's say you get a clutch pedal and turns auto-blip off. You have increased the difficulty of your simulation and are likely at a disadvantage to players using the auto-clutch. However, if you run a server today, people using auto-clutches are free to join. In order to be on an even playing field, you would have to dumb down your version of the simulation, choosing auto-clutch.
In that case, a server option to only allow fully manual clutches makes sense, so people playing a more difficult version of the simulation can have a fair race.
It makes no sense to have a server option that only allows auto-clutchers. If you are using auto-clutch, nobody joining has any competitive advantage on you. You are no forced to dumb anything down to complete. Option not necessarily.
If you are using a mouse/kb, you are at a disadvantage (in most cases) to players using wheels. In that case, an option to allow only mouse/kb users makes sense, so people playing a more difficult version of the simulation can have a fair race.
It makes no sense to have a server option that allows only wheel users. Mouse/kb users joining your server don't have an unfair advantage, they aren't forcing you to simplify your experience in any way to complete, so the option to exclude them doesn't make any sense.
Last edited by Eric Tetz, .
Don't forget videogame consoles.. they are notorious (*looks at his 360's blinking red lights of doom*).
You're assuming it's a design error. How do you know it's not a manufacturing problem? My wheel doesn't do it, and I've been using and abusing it for months now (as have thousands of others).
At least you're not assuming, as sshhaabb did, that Logitech is 'taking us for a ride', willingly deceiving us to make cheap buck with a shoddy product. That's really the only reason I replied; that kind of assumption is not only ludicrous, it's really uncharitable to some stand-up guys who have been more active in the community than any retail manufacturer I've ever seen. I'm sure they'll do right by you, one way or another (though I understand it sucks to be you in the meantime).
What does that have to do with anything? All I'm saying is that you are drawing a cartoonishly extreme conclusion (i.e. the G25 design is fundamentally flawed! all of ours will fail! the only solution is recall and redesign! Logitech is trying to swindle us!) from a sample size of one.
Oh my, how fast people turn. :uglyhamme
Hundreds of glowing reviews, then one guy has a problem with his unit and suddenly it's a LOGITECH CONSPIRACY TO STEAL OUR MONEY! THE G25 IS COMPLETE GARBAGE! RECALL! CLASS ACTION LAWSUIT!! ZOMG!!! :scared:
Well, I started by saying 720+, but explaining how some cars get physical locks and others don't confuses the point. The point is simply that doing it this way is set-and-forget. Set to any value >= 720 and you're done.
For all practical purposes, it doesn't matter whether you set it to 720 or 900, because you're never going to come anywhere near hitting the lock either way. The locks matter more in the cars that have only 270 degrees of steering.
The only downside is lack of physical locks, however:
Exactly.
So you have two choices:
- Set the profiler and LFS to 900. Your wheel matches the in-game wheel for every car in the game and you never have to touch it again.
- Set the profiler and LFS to match a specific car. Every time you change to a car that has a different steering lock, you must:
- TAB away from LFS and open the profiler.
- Select Edit -> Specific Game Settings
- Edit the degrees of rotation to match the specific car you want to drive.
- Go back to LFS and edit the 'wheel rotation' setting to match the car.
- TAB away from LFS and open the profiler.
The downside of #2 is that it's a pain in the butt, which I do notice.
So my preference is #1.
Try it.
(1) Set the DFP/G25 to 720+ degrees in the control panel (I just use 900).
(2) Set 'wheel rotation' in LFS to the same value.
(3) Set 'wheel compensation' in LFS to anything above 0 (.01 works; why this needs to be set to a non-zero value, I don't know, perhaps it's a bug)
Result: the rotation of the wheel on your desk and the wheel in the game will match exactly, for every car in the game. You won't need to go back to the control panel every time you change cars.
The only reason to change the lock in the control panel to anything less than 720 is so that you can get the FF locks for a specific car (like lyd said).
You don't want to enable 'centering spring' unless you are playing a game that does not support force feedback. LFS will generate centering forces when they are appropriate via the physics model.
'Combined pedals' is for games that don't support separate axis for gas and brakes. LFS does, so you can turn this off.
I have 'degrees of rotation' set to max (900), though I don't think there are any cars in LFS that support more than 720 degrees.
I leave all the other the global parameters (overall effects, spring effects, damper effects) at their defaults.
In game, 'wheel turn' should be set the same as 'degrees of rotation'. For me that's 900.
'Wheel turn compensation' should be set to .01. If you set this to 0, LFS will not match the turning of the G25 to the in-game wheel. If it's set to any non-zero value, the G25 and in-game wheel will match degree for degree.
'Throttle brake axes: separate' should be on, assuming you un-checked 'combined pedals' in the profiler.
The turning of your wheel and the in-game wheel will match exactly, but you won't get any real steering lock. For instance, when driving the MRT or an F1 car you will be able to turn the G25 past 270 degrees, but the in-game wheel will just stop at maximum lock for that car.
'Combined pedals' is for games that don't support separate axis for gas and brakes. LFS does, so you can turn this off.
I have 'degrees of rotation' set to max (900), though I don't think there are any cars in LFS that support more than 720 degrees.
I leave all the other the global parameters (overall effects, spring effects, damper effects) at their defaults.
In game, 'wheel turn' should be set the same as 'degrees of rotation'. For me that's 900.
'Wheel turn compensation' should be set to .01. If you set this to 0, LFS will not match the turning of the G25 to the in-game wheel. If it's set to any non-zero value, the G25 and in-game wheel will match degree for degree.
'Throttle brake axes: separate' should be on, assuming you un-checked 'combined pedals' in the profiler.
The turning of your wheel and the in-game wheel will match exactly, but you won't get any real steering lock. For instance, when driving the MRT or an F1 car you will be able to turn the G25 past 270 degrees, but the in-game wheel will just stop at maximum lock for that car.
We've got nothing on the flight sim guys.
http://www.737sim.com/737_Pics/Intro_page_pit_small.JPG
http://members.aol.com/wietlpachm/howto/web28.jpg
http://www.737sim.com/737_Pics/Intro_page_pit_small.JPG
http://members.aol.com/wietlpachm/howto/web28.jpg
I never said anything remotely like that. I think LFS is better online for a lot of reasons.
I know what you are talking about. I quoted you, didn't it? Barriers = reproducible at will. In a race = uncommon.
How is it not?! If a car is inside another car, a collision has occurred. That is all collision detection code does: "is any part of volume X inside volume Y?"
The discussion at hand is about what LFS does with those sorts of collisions. For instance, client X gets an update from the network telling him that car Y is inside our car. How far is he into our car? What was his velocity vector when he struck us? We can use that to create an approximation of a realistic result, right? Right now, LFS does some wacky things under some circumstances, but I'm sure that will be worked out by S3.
I think you mean not as much data per update (rather than fewer updates, which would make teleporting issues worse).
You said yourself it happens, but it's "uncommon". Any differences between rFactor and LFS are relevant in a thread like this.
Happens all the time on autocross tracks. Bump into a barrier wrong and suddenly your car is doing somersaults. That's not an indictment of LFS -- we're all fans here, so there's no need to get defensive about it -- it's just a relevant observation when comparing LFS and rFactor. It's one area where LFS's physics, which are in most other respects superior to rFactor's, will need to be improved in order to support more motorsports and a mod community (eventually).
For instance, rFactor has a buggy mod. It's not very polished, but the point is that some guys threw together some buggie models, described the vehicles to rFactor, and it works. They drive like buggies. They never catch a tire wrong and go flying into the sky spinning like tops. I would much prefer to have such a mod in LFS (would gladly pay $50 for a track editor), but there are a few kinks to work out first before a mod community can just go nuts and not end up with wierd results. As it is now, if you try to build a ramp in the autocross editor and don't put the ramp planks close enough together, it becomes a space shuttle launching pad. ^_^
Absolutely. It's the reason I can walk into Fry's and get one. Not because Logitech is nationalistic (as some Canadians have theorized ^_^), but because Fry's is a HUGE chain that sells a high volume Logitech products.
Just to be clear, you are setting the game specific settings in the Profiler, not the global device settings, right?
In the Profiler, you create a profile for LFS and have it selected (so you see a "Play LFS" button on the top-left of the dialog, after the Select Device and Select Game buttons), then you click Edit menu and select the Specific Game Settings... option on the very bottom. This will bring up a dialog where you can set a specific amount of wheel rotation for LFS.
Well, you said "I don't see what all the hype is about". I'm pointing out that you haven't seen what the hype is about because your computer can't show you.
LFS won't look much different on a high end machine compared to yours (other than resolution, filtering, and anti-aliasing improvements). GTR looks a lot better on a high end machine. The track doesn't change that much, but the car models become extremely detailed.
Last edited by Eric Tetz, .
I had the same issue with GTR's physics: massive understeer, and when the rear end does break loose the car usually ends up facing backwards (almost like it's canned animation). If you play long enough, it still ends up being fun and intense, and you start to be able to catch the rear end, but it doesn't feel quite as legit as LFS for some reason -- but I've never driving real GT cars. Do they understeer that badly? Hell if I know.
Sounds wise, GTR has an extremely good sample-based sound engine. If I have a friend over, and I want to impress them with how cool driving sims are, I don't start them with LFS. I set them up in GTR driving a Lister Storm and a crank the speakers. Roaring down a straight with that thing at full throttle, my brother said, "I don't know if I'm supposed to be excited or afraid." The car sounds like it'll come alive and rip your head off.
That said, after an hour or so with GTR or rFactor I start to get fatigued. The samples begin to sound very repetitive. That never happens to me in LFS. I feel like the sound engine in LFS conveys a lot more useful information.
It's the exact same difference between sprites and real-time 3D. Sprites can easily be photorealistic, but the illusion is blown when you get them in motion. Samples are the audio equivalent of sprites. LFS's approach is the audio equivalent of real time 3D rendering. LFS's audio is more realistic 'in motion' than any sample-based approach I've heard.
That's not the force feedback, per se. In other words, it's not necessarily a problem with the software providing inaccurate feedback information to the wheel, it's a weakness in the mechanics of most force feedback wheels. They are very slow, and have a significant amount of latency in the force feedback, especially when changing directions (which causes that oscillation problem).
If you want to see how good it can be, check out the G25. It uses two FF motors, each with reduced gearing. It has much less latency and it's about 3 times faster than the DFP. Still not as fast as a real wheel, but fast enough that letting the wheel automatically countersteer feels much more realistic, even when using 800+ degrees of steering lock.
Perhaps you need a better GPU? The cars in GTR with all details cranked to max are astonishing, IMO.
Sounds wise, GTR has an extremely good sample-based sound engine. If I have a friend over, and I want to impress them with how cool driving sims are, I don't start them with LFS. I set them up in GTR driving a Lister Storm and a crank the speakers. Roaring down a straight with that thing at full throttle, my brother said, "I don't know if I'm supposed to be excited or afraid." The car sounds like it'll come alive and rip your head off.
That said, after an hour or so with GTR or rFactor I start to get fatigued. The samples begin to sound very repetitive. That never happens to me in LFS. I feel like the sound engine in LFS conveys a lot more useful information.
It's the exact same difference between sprites and real-time 3D. Sprites can easily be photorealistic, but the illusion is blown when you get them in motion. Samples are the audio equivalent of sprites. LFS's approach is the audio equivalent of real time 3D rendering. LFS's audio is more realistic 'in motion' than any sample-based approach I've heard.
That's not the force feedback, per se. In other words, it's not necessarily a problem with the software providing inaccurate feedback information to the wheel, it's a weakness in the mechanics of most force feedback wheels. They are very slow, and have a significant amount of latency in the force feedback, especially when changing directions (which causes that oscillation problem).
If you want to see how good it can be, check out the G25. It uses two FF motors, each with reduced gearing. It has much less latency and it's about 3 times faster than the DFP. Still not as fast as a real wheel, but fast enough that letting the wheel automatically countersteer feels much more realistic, even when using 800+ degrees of steering lock.
Perhaps you need a better GPU? The cars in GTR with all details cranked to max are astonishing, IMO.
Yes! When I got my G25, I started going through the LFS tutorials (they are little controlled experiments for braking, acceleration, etc.) Then I got to the the MTR autocross test, and I never left. ^_^ So much fun doing that with the G25.
I did that track over and over until I got a pro score, then I went looking for more autocross tracks. To my surprise, there aren't many. This is one of the most fun things in LFS, IMO.
It would be nice to have some very small and curvy tracks (like real kart tracks) in LFS, just for racing these things. They are so well balanced, and it's so easy to control understeer/oversteer, they are just a blast to push around little courses like that.
FGED GREDG RDFGDR GSFDG