i own both games that would be AWESOME if the games joined and then we would be able to add our cars to beamNG and crash test them +1
Well spoken! I agree with the idea that BeamNG feels sort of stuck between two worlds or types of gaming. The real time physics model is processor heavy and unlikely to spawn a multiplayer game or one with a large number of beamed objects interacting, thus a racing based gameplay may not play to the strengths of the game. I mean the point of racing is trying not to crash. Sure it adds a little danger that will make it risky to cut corners but that doesnt require a complex physics model. The popularity of Ror is the sandbox nature and modability. If you could streamline the jbeam creation you could implement the design work like Automation then use the BeamNG to really experience the fruits of your labor and ideas. With a consistent design side of the game, cars made by different people would have the same look and feel, and be constrained to the same limitations, thus being comparable from one design to another. No going into notepad and doubling the hp to win lol.
You mention Orbiter, a buddy of mine plays that religiously. I think you're spot on with the statements regarding "games" like Obiter and BeamNG. They cater to a niche crowd for sure but the future implications of the technology and software are readily apparent. Being part of this is the allure of it, being part of a dream come true to those who have always dreamed of crashing cars realistically (me), or to pilot a uber realistic spaceship in an uber realistic gravity simulation (my buddy). It's all about being part of something awesome that can...well..just BE.
Thank you for your reply Aborath. My point is more that BeamNG is not (and does not seem that it will in a near future) ready to be something like Orbiter (but with cars) in the realistic, hardcore sim sense. So if I were its developers I would look for other "niches" other than racing (where there is already strong and good competition) and try to find the place where the techno and software can make sense. For BeamNG "just BE" means just being a technology demonstrator (Maybe it is but developers never told us so...) so I think it should focus in being a "game", in the best sense of the word: an streamlined car design game (like somebody said before) with superb physics to test, crash test, race, etc. your own creations, just like in Kerbal Space Program.
From what i understood out of old forum statements and whatever i have picked up over the last 2 years or so that the beamng name has been out there on the world wide internetz, it actually is more or less just a tech-demo, of a game engine they hope to sell to other games. For our pleasure and their own development purposes they will make a game out of it, called "Drive". But i personally think it is more to prove gameengine-popularity, (It's already been sold xx amount of times !!!) and potential, as well as finding all bugs native on their own platform before actually selling it. So honestly though i am having a lot of fun with this engine, i do not expect the game to be aimed at a very wide targetgroup, not for it to become a in depth hardcore simulation role playing deep diving whatever kind of game. I expect a highly modifiable simulation engine they runs solid realistic physics and simulations, that can easily be adapted into any direction a potential buyer would like to take it, or what the community builds on top of it. I personally, am hoping that it will soon turn into some sort of RoR type of game but better, with more and better looking vehicles, doing advanced simulations on advanced machinery like cranes etc... But it will honestly be a little surprise where the DEV's take us. In any case i am already having a lot of fun, and it easily has been my money worth, so it can only get better from now on, lets hope for the 'bestest' and coolest thing know to mankind !
Concerning the emphasis on "racing", and speaking for myself. I love the challenge of taking a basic geometric description of an object and some real world parameters and simulating how the real thing would behave. Even in extreme cases like racing. When you open our G-meter, you don't see "curated" values in there designed to "look" correct. Our G-meter shows the actual calculated forces that get applied to the vehicle. If our G-meter values are similar to a real G-meter's, it is because the calculated forces are similar to the real life forces. The same applies to the wheel weight distribution widget and all the other vehicle sensors. I'm immensely proud that we have achieved all these up to now, without "tweaking" the simulated physics reality to "look" correct. With above, i don't want to downgrade the other games. Striving to simulate the "experience" (or present an "enhanced experience") requires a huge effort too.
Exactly, please keep the simulation correct. The vast majority of people are too stupid to appreciate a correct simulation, too bad for them. The devs already stated as well that the game isn't going to be primarily focused on racing, that's just the first bit of gameplay they are adding. There is going to be a variety of things to do when they manage to start getting a "career" mode going, hauling freight, racing, offroading, trying to reach spots that are hard to get to, and I'm sure they will think of a bunch more.
yes, exactly, I think this is the reason most vehicles we currently have can carry cargo,(D series, H series, T75, Pigeon, the suv we're getting.) it's all there to give us a very challenging driving experience, instead of just being a cargo game, or a rock climbing game, or a racing game, I think of it as going to be a sort of driver for hire simulator life thing. or I could be absolutely wrong and we're just gonna have to race big rigs once in a while.
A correct simulation is one that gives the correct results, not the one with the "correct" method. Different approaches to simulation can give more accurate results when replicating/predicting a small slice of reality, or they can be less accurate and model a greater picture. With finite resources of money, time, computing power, etc, you can't have it all. Hardcore racing simulators are at one end of the spectrum, BeamNG is at the other. BeamNG's driving dynamics will never hold up to the scrutiny of a hardcore racing sim unless it dives down the same tire model rabbit hole that a racing sim does. The node/beam approximation contacting the road with coefficients of "kinetic" and "static" friction, is really just too coarse. The racing sim's simulation model isn't designed to reach as far as to model collisions, airplanes, or boats, or even a ball rolling down a hill. But, compare the telemetry between one of these sims and an actual race car and you'll find they are pretty damn close. Like really really close. The number of fudge factors and ugly curve-fitting equations being used is irrelevant if the model gives the right outputs for inputs within the scope desired. The beamNG node/beam system is a very general approach to physics simulation, which is why we can build all sort of vehicles and structures with it, and it will approximate all of them with some level of accuracy. With a little tweaking of coefficients and the art of making a jbeam skeleton, it can do a surprisingly good job of replicating the experience of driving a car on a huge range of surfaces and crashing into all sorts of objects. However, there is no real correlation between a node/beam structure and an actual car's structure, it's simply been tweaked and tuned through some intelligent trial and error to give the right results. It's no more or less "correct" or empirical than any other method of simulation.
I own Automation Supercharged Edition and to be completely honest, it isn't worth it at the current moment. This is all due to the fact there are no "extras" developed for this edition as yet, not saying there wont be any "extras" for it, they are just yet to be implemented (Which is quite sad for the huge amount of people who forked out the $35 USD). I was prepared to wait a little while for updates of any kind, sadly it's been far beyond what I would call "Active Development" unless you are fortunate to be apart of the exclusive Beta Test Team (which receive updates close to monthly) where as My Automation has not updated in a little over 4-5 Months now... Still, stating the above I do enjoy the "game" but there is so little to do within it right now it's only good for a 30-60 min session before you are completely and utterly over it and want to keeel eeet!
If you compare the physical data from the telemetry apps to a real vehicle it is very close, what's not is perceived feeling of reality (for some people). The biggest problem with the game right now is the tire physics, which can be improved. This game isn't designed solely for racing once again though, but with some more work it can do fantastic job. Just the suspension simulation alone makes up for the short comings imo, all the added nuances of driving that you don't find in other racing sims are huge. The chassis flexing, engine torque reaction, the suspension and steering components flexing and bending if you put too much strain on them, the fact that its made of actual separate components with their own weights and physics to them not just a solid block with a bunch of code making it "feel" realistic. There really is no comparison, this is the next generation of driving simulators. It's just in a very early stage of development and needs ALOT of work, but its getting there, and its made ALOT of progress. If they are able to fix up the traction and odd things that happen to the tires at higher speeds, it will give alot of racing simulators a run for their money, and that's just in late Alpha, early Beta territory. Wait until we can load up a flat deck trailer with 50,000lbs of freight and pull it around with the T75 with the engine twisting the chassis from the torque reaction, and the trailer bending from the weight like they do in reality, all the beautiful little nuances you would never get in any other simulation.. then you can switch over to a race track and enjoy all the tiny physical nuances of a racing vehicle on a track.. then switch over to a 4x4 and rat bag the shit out of it on a trail and watch as it slowly (or quickly) falls apart just like it would in reality. This isn't a single purpose simulator just for racing, it's been designed from the start to do a very good job of simulating anything with wheels. So it's not gonna be absolutely perfect at any one thing. Keep in mind this is the first commercial product to use this kind of soft body physics simulation for vehicles, the only other project I know of is Rigs of Rods and that's been free from the start and eventually open sourced.
Hey Kilo, cool to see somebody else from BC, i'm in Victoria! Well, if the solid block with a bunch of code has the same resulting vehicle performance, then really the only difference with soft body is that you can see the stuff flexing, which is making it "feel" realistic, when in reality both may be. You can simulate the effect of chassis twist and all that without seeing it, which is less fun perhaps, but less taxing on the system. The fact that a lot of these effects come naturally out of the platform is great though. I would say that a real-time, multibody spring model has probably always been the pipedream of the makers of sims, (A similar but not real time approach is used by packages like ADAMS to simulate cars for automotive engineering) but probably abandoned due to know-how, computing power, and some of the logistic problems associated with the model. There are other tradeoffs with a multibody spring model. For example, resonance. When you have a system of springs with so many degrees of freedom there are all sorts of driving frequencies and natural frequencies to contend with. Hence some of the problems with the tire at high speed, instabilities in the jbeam structures, jiggling components, etc. Do a burnout in the Bolide 390GTR in slow mo, the rear wheels are shaking toeing in and out about 20 degrees! Alone, a fixable problem, but just an example. Resonance is a huge part of engineering in real life as well. Cars do experience wheelhop occasionally, tractors can start hopping when towing too much load, etc. When I turn my drift cars steering too far the inner tie rod tries to over center and the wheel shakes like crazy. If I hit a pothole at that moment I'd probably snap a tie rod. It's not that resonance doesn't exist in real life, its just really hard to get a simplified model to vibrate at all the same frequencies and amplitudes as the real, "infinitely complex" thing. Devs and modders will probably be chasing resonance problems forever to diminishing extents as they are innate to the type of model. As well, with the emphasis on crash models, due to the approximation of the node/beam structure, what performs well in modelling a crash may not work well at modelling the fine points of say the rigidity of the suspension attachment points. A huge amount of labor in "tuning" and testing has to happen there to achieve realism. I think that is more art than science at this point. Not trying to be negative, I'm still here and supportive of BeamNG, and I think in a broad sense, its the ultimate vehicular playground. And I can't even begin to imagine myself tackling the kind of math and code to make a physics platform like this work. I feel like it would be so unstable by nature, the fact that it works so well even in alpha is amazing to me. So don't mind me, I'm just a pretty fresh engineer who loves to debate the pros/cons and get down to the nitty gritty
Concerning the rigidity question. We have reached (some time ago) in internal tests, diamond stiffness. The problem is that a so rigid simulation would tax a lot the current generation CPUs. Also, as a fellow engineer, i love the debate .
I know it would most likely be unreasonable, but would it be possible for it to be in game as an option for us with nice computers?
It would be good to see BeamNG help out with the car handling on Automation Game and Automation Game helps out with the engines and sounds in this game...thats what I like seeing 2 different development teams helping out each other...thats a really great idea
This thread has become very interesting in many ways. I enjoyed the contributions of Kilo, Goosah and others about how simulators should be. It would be very interesting to know which model ("feeling realistic" or "being realistic") are F1 or aircraft professional simulators using provided they are very useful in the real world and cost millions. I have no idea but I guess "being realistic" must be the goal for F1 sims where they need to simulate the effect of the smallest change in configuration, parts, track, etc. while in aircraft the "feeling" should be paramount as their goal is to train pilots in procedures, systems, emergencies etc. I don't know, they are probably a big, complex mix of both ideas... But Beamng, as, almost, every other sim we can enjoy in our PCs, it is nothing less but nothing more than a game. So it needs to have a game model in addition to the simulation model. The "you'll be able to race, transport things and transport things over difficult terrain" doesn't seem too interesting or groundbreaking for me and also doesn't seem to bring the most of this still imperfect but very promising sim model. Finally a suggestion for the devs, how about a game play options survey among the current users? Nonbinding of course but with some real options to choose from. Sorry again for my English and thank you all for your posts that are generally very high above the "honor your developers cause you don't want to make them angry " theme that is too frequent in this forums.
Of course. What I was saying is that: 1. Beamng deserves a good game play because does not currently have one and won't have it according to the current plans (race/delivery/off-road delivery). My personal suggestion is car design+ car testing. 2. It would be a good idea that its current users could have a say on that (game play) in a survey (with realistic alternatives to choose from). 3. Generally, this forum is not very open to (constructive) criticism. That is not good neither to developers (that are running an alpha/beta/whatever test with us in order to develop a successful game) nor to users (that have paid for a promising product that IMHO is loosing its appeal as time goes by)
