Is Outsim data filtered?
I have been working on a motion simulator for a while. During the whole process I used LFS Outsim acceleration data to test the simulator.
Only recently I started using other sims...and what a surprise, acceleration from other sims is a lot sharper than what LFS sends. For example I can perfectly feel every bump on the kerbs in other sims, but in LFS these bumps are smoothed....like if some filter was applied (and I need to tweak a lot the signal to feel them).
It is easy in LFS to guess what acceleration the driver receives, because the head moves accordingly...and when I compare this head movement with the acceleration from Outsim data....it looks like there is a smoothing filter applied somewhere.
Is there something to do to remove this smoothing? Or at least I would like to know what is the filter applied, so I can compensate for it. It may be a moving average.
Only recently I started using other sims...and what a surprise, acceleration from other sims is a lot sharper than what LFS sends. For example I can perfectly feel every bump on the kerbs in other sims, but in LFS these bumps are smoothed....like if some filter was applied (and I need to tweak a lot the signal to feel them).
It is easy in LFS to guess what acceleration the driver receives, because the head moves accordingly...and when I compare this head movement with the acceleration from Outsim data....it looks like there is a smoothing filter applied somewhere.
Is there something to do to remove this smoothing? Or at least I would like to know what is the filter applied, so I can compensate for it. It may be a moving average.
Last edited by Juls, .
If rFactor2 has good graphics and sound, good content out of the box, a tyre model close from LFS one, improved FFB... we get a clear winner.
Maybe the LFS community is not their main source of income, and they develop versions of LFS for professional customers, like this Czech simulator renting business, or this fair where they introduced the scirocco.
It would make sense. I doubt licenses bring enough money for them to live.
It would make sense. I doubt licenses bring enough money for them to live.
LFS uses 90 degrees default Field of View. By default in iRacing the FOV is 78 degrees, so you see less on the sides and the car takes more of your view. I suppose you can change it somewhere in options.
Did you try iRacing with 90 degrees?
+1
no communication > hype and anger > less communication...etc
This is a vicious circle which can finally cause the devs to stop enjoying their project. LFS was built around a friendly relationship between devs and players from what I understood.
Community can brake the circle if everyone stop complaining at the same time and wait religiously.
Devs can brake it with a dev blog with comments disabled.
What is the easiest thing to do?
More and more threads like this...it shows something. Not clear what it is, but noone can continue ignoring it.
It happened several times for me. In such situation I quit. Life can give you unexpected gifts...good or bad.
Everything looks steady and peaceful, but every 10 seconds, there is an incredible opportunity somewhere. And these opportunities never occur in boring, frozen places.
You may find something worse, and you may find something better. That's life. Life is movement. If you move more, you have very good chance to find a really good life. In french, still-life paintings are called "nature morte"= dead nature.
Everything looks steady and peaceful, but every 10 seconds, there is an incredible opportunity somewhere. And these opportunities never occur in boring, frozen places.
You may find something worse, and you may find something better. That's life. Life is movement. If you move more, you have very good chance to find a really good life. In french, still-life paintings are called "nature morte"= dead nature.
No, it can't be. NFS is LFS improved. There was LFS, then MFS and finally NFS. Better graphics, better sound, better physics... all pretended sims like iRacing LFS and others fail to simulate grip.
Only NFS, thanks to it's evolved state-of-the-art physics engine, has the ultimate grip.
Come on... LFS does not even have blur!
I agree... and all this crazyness about Scirocco and other updates is a consequence of the "wall of silence".
Too much hype and too many announcements is never a good thing, but on the other hand, giving so little information is the best way to make people crazy too.
Imagine something very very simple....a web page with a simple list for the 10 or 20 top priority development tasks in LFS. For every task, a completion status, maybe percentage. No need to announce, no need to explain. List is there and gets updated for example every week. No need for posts from devs which finally get lost.
Scirocco enters the list, and later completion status becomes postponed because other higher priority tasks come on top...shadows bugs, ABS, relocation, whatever. Even the patch test forum is now silent...
Last edited by Juls, .
In France there has been something similar since...1998 or even before (I was using it in 98). Mappy did it or bought the company doing it.
Example here...click on a street on the map, and move the view with the arrows.
http://photos.pagesjaunes.fr/h ... Wx0smw==;x=254.2;y=193.35
Nobody complained about privacy stuff...it's public place pictures...but now it is Google, and like Microsoft they have MONEY. Sure European Union will manage to find a reason to ask Google to pay a fine for it.
Example here...click on a street on the map, and move the view with the arrows.
http://photos.pagesjaunes.fr/h ... Wx0smw==;x=254.2;y=193.35
Nobody complained about privacy stuff...it's public place pictures...but now it is Google, and like Microsoft they have MONEY. Sure European Union will manage to find a reason to ask Google to pay a fine for it.
I hope they do not lack so much common sense.
If I have to chose between two employees...one which sold one car for a 1 euro campaign....and other who sold 100 cars for a 10000 euros campaign, I take the second one...unless the first one accepts to be paid with a percentage instead of currency of course.
We live from volume*ratio...not only ratio.
Last edited by Juls, .
That's why car producers ask all the time for sims to include their products, but for an unknown reason sims developers insist to refuse and pay license fees
Sorry, it does not cost 0 because it requires time from technical staff. And it costs even more because it requires to work on new things that can not be reused later (if as you think this is for marketing). Time consuming, not reusable=>expensive.
And how does it compare with alternatives??
Imagine you are VW marketing and compare it with arcade game like NFS.
NFS guys pay the licence, come and scan the car shape, do not take time from your technical staff because they do not ask for any data or technical question, and sell a game to millions users in which the car looks better than real, handles easy and reaches 400 kph. Not mentioning the massive advertisment campaign NFS has.
Now, the week after NFS team leaved, you receive a call saying LFS want to use your car. They need a lot of data and time from technical staff, it will not look fancy like your other advertisment, it has a very limited audience, no advertisment campaign, and they will not pay a license for it. You accept that only if you have another idea than doing marketing with it. R&D, maybe later organizing events with the sim...etc.
Je vois bien que tu tiens a cette idee, mais c'est vraiment tire par les cheveux. LFS? Combien de divisions?
To sum it up, who benefits more from having a car from a famous brand in a sim? The car producer or the sim? So who pays for it?
Last edited by Juls, .
You really believe that it is interesting for a worldwide automotive brand marketing to have it's products inside a computer simulation? IMO, it is not interesting at all for them for marketing....it requires a lot of work and bring them less than any advertisment page they buy in any local newspaper.
That is why sims developers usually have to pay for it. If they do not pay for it...it's because the brand decided to sponsor them that way, or because it is interesting for other reasons. But certainly not marketing.
OK, so it looks like you use 48volts and get 4 times more heat than default G25. With two small fans why not! It still looks reasonable.
The specified motor voltage applies only when the motor is not blocked. When blocked, current through the motor is far more important.
If you take a 24 volts motor and block it, it will burn.
In a wheel motors are blocked most of time. That is why wheels HAVE to use motors with a far lower voltage than specified.
What is important is not voltage but current. In a wheel, motors reach their max current and heat with a lower voltage than specified, because they are blocked.
Second point....heat in DC motors is produced in the rotor...which is enclosed inside the motor. For the heat to reach the outside of the motor, it has to cross the air gap between rotor and stator. And air is a very poor heat conductor. As a result, even if you cool the outside of the motor with a fan, the rotor inside will reach a very high temperature.
I suppose the G25 motors should be able to survive this, but you can not say you stay within motors specifications. What is the current max going through motors with your mod?
Do not forget that stewart platforms do translation. This does not appear on the wikipedia page, but they can translate. And if you recline the actuators of the stewart platform, then you increase it's translation range and decrease the rotation range...
With a large enough stewart platform you can do everything...translation for the ramp, then tilt while the platform translates back to the center. Expensive simulators using stewart platforms do exactly that. They translate to kick real acceleration, then they tilt to use gravity while they translate back. This concept mixing tilt and translation has been used since the 70's. And the algorithms to mix the translation with tilt are called "washout algorithms", you can find documents if you google it.
Force Dynamics and others use other concepts (almost rotation only) because it is less expensive to build.
Yes but 2 people more in a team, and it takes longer for the pot to turn around.
Sorry, kidding, but I just pictured the scene...this is pure fiction of course..." Damn, I told you not to hire a new dev, man, now when the thing comes back it is already half smoked, this is bad for vibes man".
I have the weird feeling you do not read at all the answers. I am not dreaming.... you are trying to explain FD how they should do their sim?
That's why you need to understand this: what looks right is often not what feels best. Motion simulators HAVE to trick body and brain. They HAVE to do very strange things to make you believe you turn hard in a car at 100 km/h while you are almost not moving at all.
Frex is not simulating the acceleration forces on the body. Frex works the same as your idea. It simulates the action/reaction between your body and wheel/pedals.
It makes you feel like the wheel is pulling your arms when you accelerate, and pushing against your hands when you brake. To do so it moves your body backward/forward instead of moving the wheel like in your idea, but it gives the same result. The range movement can of course be decreased, but it seems it feels more realistic when it moves more than IRL, because you have to compensate for missing cues in the inner ear.
Once again, what looks right (10-20mm movement) is probably not what feels best.
edit: your last idea is a motion simulator with rotations+translations. More expensive simulators than FD work like that. Often they use Stewart platforms because it is a very compact system allowing all rotations and translations.
http://en.wikipedia.org/wiki/Stewart_platform
Look at this one sirocco
40 degrees rotations, 60cm translationshttp://www.symetrie.fr/Ressources/Files/SYMETRIE-sirocco.pdf
Imagine this one:
http://www.symetrie.fr/Ressour ... ge/l_hexapode_aquilon.jpg
Last edited by Juls, .
This is not so easy to spot on a video what's wrong or right with a motion simulator. What looks first like strange, unrealistic movements, often appear to be very well done.
When you accelerate/brake, the FD tilts backward/forward until the gravity pushes the body backward/forward.
And the location of the center of rotation is very important, but it is difficult to explain (with my english).
If the platform is tilting forward/backward around a low center of rotation (like blue tiger simulator), then it causes conflicting acceleration cues during the movement. Imagine you accelerate, the platform tilts backwards around a low center of rotation. Picture it. The wheel and seat are moving toward the back in user perspective. This feels not like accelerating but like braking...wheel is pushing backward against hands.
But once the backward tilted position is reached and the platform stops moving, you feel the gravity pulling you backward like an acceleration. And this is right.
->we have conflicting acceleration cues.
Now imagine you accelerate and the platform tilts around a high center of rotation, above head, this time forward. Wheel and seat are moving forward, seat pushes against your back and wheel pulls arms, like in real accelerating vehicle. And once the tilted position is reached, the gravitu is pulling you backward like in accelerating vehicle.
-> acceleration cues are coherent.
FD is one of the very rare motion platforms that rolls/tilts around a high center of rotation to avoid conflicting cues. When they tilt/roll the platform they ad the exact up/down movement needed to have a high center of rotation.
So this up/down movement which looks strange for you is in fact very carefuly engineered to avoid conflict between proprioceptive (muscle) cues and inner ear cues.
Here guy from FD explains this very important point a lot better than me:
http://www.youtube.com/watch?v=ghncbQVCQVM
If you tilt/roll someone and translate a little bit at the same time in a very precisely calculated way, you can fool completely the inner ear, the person will really think she is accelerating in a straight line. This trick is the base for stewart platforms, most military motion simulators, big entertainment motion simulators...etc.
Last edited by Juls, .
It does not have to move in a "natural" way. It has to reproduce acceleration you feel in the vehicle. Rolling and tilting in a carefully chosen way around a high pivot point is a very good way to achieve this when you have little or no translation degrees of freedom.
Do not think they are idiots. Read about washout algorithms and motion platforms. This is a very complex matter.
Before 10 years ago, it was easier to catch the sim virus, because sims were at the same level graphically than other games and advertised like other games.
So we were simply trying all nice games, attracted by novelty and fancy graphics/sound, and we discovered some of them were deeper than others.
Now, sims with their dated graphics and lack of advertisment (compared to big games) are more difficult to find. You have to know what you are searching for, there is very little chance to discover a sim by luck. So almost only people who feel the Need For Sim find them.
And when I wish LFS were developped faster, this is because I always regret there is no more sim with up to date graphics, sounds and physics. The sim world, generally speaking, needs a great, fancy, well advertised sim to open the eyes of people and show them what racing sims can be, and that what they usually consider like a realistic game is often crap.
I like what Simbin try to do. They produce sims or sim-like games, and with lots of graphic and sound work, try to make virtual motorsports popular. They managed to have virtual races on TV in Sweden.
The world desperately needs popular sims. Somewhere, working in the dark, heroes can save the world
So we were simply trying all nice games, attracted by novelty and fancy graphics/sound, and we discovered some of them were deeper than others.
Now, sims with their dated graphics and lack of advertisment (compared to big games) are more difficult to find. You have to know what you are searching for, there is very little chance to discover a sim by luck. So almost only people who feel the Need For Sim find them.
And when I wish LFS were developped faster, this is because I always regret there is no more sim with up to date graphics, sounds and physics. The sim world, generally speaking, needs a great, fancy, well advertised sim to open the eyes of people and show them what racing sims can be, and that what they usually consider like a realistic game is often crap.
I like what Simbin try to do. They produce sims or sim-like games, and with lots of graphic and sound work, try to make virtual motorsports popular. They managed to have virtual races on TV in Sweden.
The world desperately needs popular sims. Somewhere, working in the dark, heroes can save the world
Last edited by Juls, .
I did not want to say Scavier makes a luxury product, but craft in the traditional, pre-marketing age way.
It has little to do with luxury, brand association, imagery, emotion manipulation...
To take another comparison, Italy...I heard about guys there who still make tailored shoes, in small workshop for same price than mass-produced shoes.
It works because their products are known for their quality, and because they do not have any of the costs modern companies have.
Guys working in these workshops do almost no marketing and do not need any, and would rather die than work in a different way. Very different from Ferrari and others.
You completely missed my point. Their target, like for the shoemaker, would be to keep doing what they like, and live from it, as long as possible. This way of thinking may be very rare among software developers, but is still very common in Europe, and often prevalent among traditional craft workers. Quality of life first.
It has little to do with luxury, brand association, imagery, emotion manipulation...
To take another comparison, Italy...I heard about guys there who still make tailored shoes, in small workshop for same price than mass-produced shoes.
It works because their products are known for their quality, and because they do not have any of the costs modern companies have.
Guys working in these workshops do almost no marketing and do not need any, and would rather die than work in a different way. Very different from Ferrari and others.
You completely missed my point. Their target, like for the shoemaker, would be to keep doing what they like, and live from it, as long as possible. This way of thinking may be very rare among software developers, but is still very common in Europe, and often prevalent among traditional craft workers. Quality of life first.
Last edited by Juls, .
Very interesting thread here. IMO all this is a question of philosophy. With the recent posts from Eric and Scawen, I understand better how they see LFS (maybe).
They do not seem to consider it like a software company, but rather like a traditional craft company. For example a company producing limited series hand crafted cars.
In such craft company, the main objective is to create the best atmosphere for working. Too much production, too big team, and you introduce additional constraints which prevent you from fully focusing on the product. You have to cut costs, remove marketing, use a reduced team, work in small workshop to give the best result without getting prisonner of commercial constraints.
This is a great, ethical idea. I would like to create similar business (but hardware product)... product first, marketing to hell, never do anything that could result in a quality loss. But sometimes I wonder how this model can apply to software business. Writing a large piece of software like LFS is not like producing hand crafted cars, one after the other.
It is rather like building a castle. LFS grows all the time, each part added requires maintenance, repairs, improvements, and will consume another part of the workforce. For example, Eric has been carefuly adding textures for years, progressively...imagine the work required to make all these textures high-res now!
IMO, building larger and larger software without scaling the team leads to a progressive decrease of new features development rate and motivation. This is unavoidable, and in fact it is not a problem, it's their choice, it's their product. A bit frustrating sometimes for those who think LFS has the potential to eclipse all other racing sims like I do.
I admit this is very easy to dream about what Scavier should or should not do...I go back to work
They do not seem to consider it like a software company, but rather like a traditional craft company. For example a company producing limited series hand crafted cars.
In such craft company, the main objective is to create the best atmosphere for working. Too much production, too big team, and you introduce additional constraints which prevent you from fully focusing on the product. You have to cut costs, remove marketing, use a reduced team, work in small workshop to give the best result without getting prisonner of commercial constraints.
This is a great, ethical idea. I would like to create similar business (but hardware product)... product first, marketing to hell, never do anything that could result in a quality loss. But sometimes I wonder how this model can apply to software business. Writing a large piece of software like LFS is not like producing hand crafted cars, one after the other.
It is rather like building a castle. LFS grows all the time, each part added requires maintenance, repairs, improvements, and will consume another part of the workforce. For example, Eric has been carefuly adding textures for years, progressively...imagine the work required to make all these textures high-res now!
IMO, building larger and larger software without scaling the team leads to a progressive decrease of new features development rate and motivation. This is unavoidable, and in fact it is not a problem, it's their choice, it's their product. A bit frustrating sometimes for those who think LFS has the potential to eclipse all other racing sims like I do.
I admit this is very easy to dream about what Scavier should or should not do...I go back to work
Last edited by Juls, .
Concerning sounds, it seems at first glance LFS and rFactor have very different sound systems but in fact they are quite similar. They both use samples modulated with engine frequency to recreate the sound. But rFactor uses long samples (few seconds) while LFS uses very short samples repeated more often.
Instead of using one sample modulated over the entire frequency range like LFS, rFactor uses several samples, and each one modulated on a frequency range...like one for 1000-2000 RPM, another for 2000-4000 RPM...etc.
They both read samples at different frequencies. Not so different.
FGED GREDG RDFGDR GSFDG