Beamng vs reality: Real world personal experiences

That rollover was achieved mostly with launching a big, metal stick into the ground.

So, yeah.

 

I finally got my Mobius camera and a suction mount so I can take this comparison a little further now with personal real world visual evidence. Just a little peak at what the camera is like and how I can mount it, any suggestions or critique is much appreciated

 

Very cool!

 

Would be better if the camera would be farther from the vehicle, so the interaction of the road with the tire is more visible, but I know it's dangerous. And it looks like the camera has low framerate and bad image stabilization, but it's not that bad, actually it's pretty good.

 

the camera runs a genuine 1080p recorded at 30fps, the original footage on my pc plays without frame drop or disruption and is super impressive but once uploaded to youtube it takes a big hit in both detail and smoothness. The camera has no image stabilisation at all, its an absolutely tiny piece of kit so I wasn't expecting stabilisation on it. But it could be stabilised in post production without much effort.

Strangely enough I had plans to mount the camera further out from the car, but on public roads that would pose a potential injury risk so would only be able to do this on private roads or with some form of fold away system like wing mirrors do in case it hit something. I am looking into this as I want to replicate the side view from Beamng as closely as possible

 

I think most of the handling problems will be solved with real steering wheel support with force feedback. As of now, keyboard and controller is hard to control sometimes. I haven't tried it with my wheel yet because I don't see the point without force feedback.

 

very nice video phipck!! but i please You to record a ride along the same route as in your very first video in this thread. it had more speed and some bumpy parts too i think to remember, would be interested to see that again form the outside view using the new mount.

---

sorry, just in case the following impression was stated before by another poster here, did just fly over the thread fast, as there is much useless stuff within.


i allways had the feeling rollovers on tarmac are working fine...or realistic if you want to say so.
But rollovers on grass or any surface which is mainly made of normal soil/earth would make the car rollover much more in real life, i think.
Take a look at rally crash compilations, the edges of the car get rammed into the terrain by the rotational forces, sometimes even throw up pieces of the turf and by that get way more friction from the contact with the ground, unlike a car rolling on tarmac or concrete. so i have the impression the cars roll way to less if you fly off the track, mostly because it looks like the bodywork recieves to less friction from the surface.
Or is that just happening because most maps soil does not have real depth throu a depthmap? (speaking about normal grass not mud)
Think i have to make some comparative tests and maybe a video laters.


Of course i have no idea if more friction between metal parts and surface layer would actually produce unrealistic node/beam behaviours like ripping away from their original position, if that is a reason for the low friction of soil towards metal. metal friction seems fine while sliding over tarmac on your roof.

 

I haven't been back down that route since making the video, but it was always my intention to re- record that route with an external mounted camera. I have the camera in the car and most likely will be near where that road is on Friday, so definitely going to record that road again

 

Although it may be too bouncy, it is certainly seems amplified from the cockpit view because the camera is static with all car movement. In real life you self stabilize your head some and your brain does the rest of the work. Think of those go pro cameras, have you seen how bouncy they look when someone runs with one? It doesn't look anything like that when you run yourself. This is what is going on in beam ng from what I can tell.

 

Besides what Ethan said above, You also have to remember that the bumps in these roads are huge. You wouldn't be driving that fast over a road in that shape.

 

That doesn't explain the jiggly behavior at mathematically flat terrain like the grid map though.

Also, it's it possible to lowpass filter the rev and speedometer needles? There is a lot of aliasing on those, making everything feel extra jiggly.

Sent from my GT-N7100 using Tapatalk 2

 

i think you should revisit the tires-not-being-round problem. a lot of people seem to agree that grip is a problem. you simply can't have accurate handling characteristics if the tire is not perfectly round: it has to make contact with the road at all times.

real-world performance cars often can handle slightly over 1g of lateral acceleration without breaking traction. i can't even come close to that number in beamNG. try modeling yourself an F1 car and see if you can achieve 5g's in a corner with the current tire/downforce simulation.

 

I spent a lot of time playing recently, and I honestly think real steering wheel support will fix most complaints about the handling. I am able to get cars to handle well and realistically if I have a sensitive enough thumb with the controller, but it is difficult, and it would be difficult driving a real car with a 360 controller as well. The game is too sophisticated for such sensitive controls.

 

There already is wheel support.

 

No force feedback. Can't feel the road without force feedback.