Bikes in Assetto Corsa

@NightEye No I didn't mean it that way. I didn't meant it this way:


But this way, only to have it rotating in dual axis.

 

I thought you were referring to the trustmaster type controller
While the one on that video is a 1000 times better than what's available, it still wouldn't feel right for people that ride bikes IRL I guess, as the axis of rotation is so different from real life. Wouldn't mind testing that out, though!

 

Yeah, I think at a minimum for a bike simulator, you want legs simulation so the thing will stand up at low speeds (and for example you can roll it around to point a different direction without falling over), and probably "picking the bike up after a crash" simulation so you don't reset to pits every time.

I did manage a couple laps of Silverstone but it was not easy. Was on the old tire model though, so maybe new camber sensitivity gives it more grip when leaning.

 

This, which is an evolution of the one in the video above looks interesting.
I am a rider, so yes, it is not like riding a bike, as your body does not do what we do on a real bike, but probably if played in VR as close as one can get at home. I particularly love to gear selector. This might be something a real rider can enjoy. Not on the same level as a proper rig for car, but still a right step toward simulating bike riding.

 

That shifter
Still, counter steering would feel weird - if not wrong.
Gotta give GPBikes a try again

 

That thing should go on kickstarter and then go on production, finished and full could cost something like G27 maybe, maybe a little more. It shouldn't be impossible to produce small simple detachable righ to mount these on and to sit like on a bike. And should have vertical axis for catching powerslides

I think would sell well.

In terms of FFB it is maybe a bit weird for bike, what forces can be felt through a handle bar ? I guess only bumps ? I don't think bike has steering resistance.

I downloaded GPbikes demo today, I think it should be something with G25 on small steering lock, like in baronesbc video.

 

it's amazing how all these devices manage to completely miss the point of how a bike is actually driven

 

Yep, i used 75deg in the logitech profiler.
Would be perfect (maybe) a controller of the size of the device in the video here above (by the way congrats), but with 2 axes: the main one to control the body of the bike, the other for its steering so to allow the countersteering.
Another solution is that the new bike sims start to think a more simple and precise way to control the bike, because the hardest thing right now is to put the two wheels where desired.
NetBike was in the right road, complex physics, but a simple way to control it.

 

@Lord Kunos I have no idea how bike is driven, only bicycle haha, I wonder if proper device to control bike simulator exists at all ? What does this one miss:



Should it be able to read how driver is changing center of mass by his body movements, or something else/more ? Guys earlier in the thread pointed out that it is not so much necessary to actually lean like that for a sim. Or maybe it is ?

@vegaguy 5555 Here is an idea to you, design of bike simulation rig and devices with full functionality and ability of tracking how driver is throwing his weight all around, pulling, pushing the bike in every direction possible.

On the other hand, maybe this time actual bike would be cheaper

@baronesbc Yeah it would be fun enough just to have a bar which would be swinging around offset angle like in that video above, and turning around vertical angle few degrees for countersteer. Should be entertaining enough.

What do you mean "hard to put wheels where designed" ? Hard to predict what they do ? Or it is so because controller is not designed for that task. By the way, those few laps around Laguna Seca does not look like it is super hard to point it where you want, as you go very close to the edges, except that one shake before corkscrew, that one was scary

 

True, but I could see how reversing the movement on that handle bar to mimic counter steering could be acceptable.

@mantasisg 2 "I have no idea how bike is driven, only bicycle hah" they steer the same way essentially, you need to push on the right handle bar to go right. contrary to what people who do not ride think sometime, and sadly some who ride as well, moving your body on the bike does not steer the motorcycle, pushing the handle bar does.

For those not riding ( yet) or who ride but might want to know more, the DVD by Keith Code : Twist of the wrist 2 will tell you all you need to know to understand how to properly ride a motorcycle.

 

I play GP-Bikes with this analog(progressive) motion sensor controller in VR and it feels great.

lean the controller up to accelerate, lean down to brake, lean left and right to turn.
Microsoft Sidewinder Freestyle Pro bought in the 90s.
Moving your body in any game is a no go for me.

 

From my experience with netBike the ideal bike controller would be as follow:

- Torque sensor mounted on an axis to simulate the handlebar. This for bikes is ESSENTIAL, bikes are not driven by wheel position but by torque, the driver pushes on the handlebars, he doesn't really turn them
- Very strong motor on a different axis, this will output the angle of the bike.. the point here is not to fight this rotation but treat it as an output.. I push on the handlebar (explained above) and the bike leans
- Possibly a strong motor on the same axis as the handle bar to simulate the small rotations of the wheel.. IMO this is not essential
- Separate axis for front brake on handlebar, rear brake as pedal, gear pedal, cluch on the handle bar.

Of course, once you have all this you'll need a software that is correctly using handlebar torque as input of the physics system.. I am not 100% sure but I doubt any software other than netBike does that... they all use the input as the "desired lean angle" and work out the necessary physics to get to that angle, depending on the complexity of the game.

And in all this, driver movements will still be totally absent.. which is clearly non ideal.

So couple a very complex and unique hardware device designed to work with a software that can only be really appreciated with a very complex and unique (and probably very expensive) hardware in a genre that is even more niche than car simulation and you have a clear picture of why this is very unlikely to ever happen

 

Driver movements could be simulated by the software. Usually you just want maximum body movement at higher speeds, "negative" body movement at slow speeds. Any sort of stability control script could control body movement if that would prevent a driver from falling.
But yeah, it wouldn't be worth it

Stefano, is there any way to obtain netBike? I'd love to give it a go but all download links I could find are down.

 

Please include sidecars!

 

Already asked Aris to create the possibility for asymetric suspension because I wanted to make a sidecar racer, but he said no

 

LOL, I was about to reply how my Bike Simulator would look like, but Stefano got it already covered.

Pushing on the bar (no actual movement) leans the bike over, stopping to push LEAVES it there at that angle, reverse push to stand the bike up. I.e during the turn you don't need actual force to hold the lean. The software takes the lean angle as input (and the increase rate in lean angle), not what you do on the handlebar.

Problem being, centrifugal forces are missing so any kind of realistic hangoff will see you slip off the seat. I'd still advocate a kinetic-style camera which rewards proper body position with more grip in turns and tuck-in with more straight speed.

But at the end of the day as I said before it's not essential because these things are subconscious. Give me a detailed-output gamepad with enough buttons and triggers to have separate brakes, shifters and throttle, it does just fine simulating the mental aspects of bike racing. Just as a cheap wheel will do fine to simulate a car on an entry level, even with limited 180° rotation and rumble motor instead of FFB.

 

I didn't quite understand why would it need very strong motor then if leaning ends if pilot stops pushing handle bar, as GCCracer wrote. To make simracer to give realistic amount of push to make bike lean. Do you need a strong push IRL ? Or maybe reverse push is something what needs greater force, which for me seems more likely.

I have to watch more slowmotions from Moto GP, it won't make me an expert, but at least that is what I can do now

Edit: It looks like Moto GP drivers are really trying hard to keep their bikes in an angle while cornering fast, I guess centrifugal forces is the reason why it would need very strong motor. I guess if they would stop pulling a bike down in high speed corners then they would fall outside the corner ?

I think it is essential for simulating this:

And again, why it would need strong motor for this, does it take a lot of force to turn it like this. And hat is strong motor, is G27 strong for comparison?

Hey, maybe thats excelent idea, which could eliminate the problem of all body movements missing ! If device could tell to the software where drivers head is, then it wouldn't really be necessary to have lower body leaning too. Torso is quite flexible, at least mine lol I can lean in quite an angle hahaha With trackIR or VR maybe it should be possible to tell software when you are leaning forward for speed, leaning back up for braking, leaning left and right. I just tried it behind the desk, and I think it would be OK.


Gotta try GP bikes, and netBike. Is netBike available somewhere ?

 

It can be quite a strong push needed on the bars in real life, I remember at least one story of a World Superbike rider bending a handle bar. But since you basically have "no movement" (only a few mm at the end) a torque sensor works better than anything else.

I think Stefano meant a strong motor for lean because the entire bike mockup and rider weight would be on that.

If I had to design something cheap and yet useful, it would be a desk-mounted handlebar with torque input (actual rotation is unrealistic), calculating a "rate of axis movement" based on how hard you push, and when you stop pushing the axis value stays constant until you push opposite.

Add to that a twist-throttle, brake lever, floor set with a brake and gear change lever. Maybe a "rumple pad" you could stick on your lower back or under your seat for giving situational feedback. That would work for any of the current PC titles, give precision and enough input without being overly expensive, and simulate adequately.

 

But leaning would still be wrong, as you´d have to hang onto the bike, while in a real corner you´re way closer to neutral due to the centripetal forces.

I don´t see a good way to do that. And you´d still need front / rear weight distribution sensors to be able to utilize your body weight during breaking...

The way higher relative weight of the rider compared to the total weight makes this hard to simulate, I´d guess. A gamepad would be enough for me, everything else sounds like a failed attempts.

 

It needs to be strong enough to overcome the strongest player... the role of that motor would be to turn at whatever angle the bike is leaning in the sim.. the player won't have to be able to move it.. in order to move it the player will have to push on the handlebars (and yes, this might take quite a bit of force for fast direction changes), the software will see that, recalculate the physics/bike angle and drive the super strong motor to that angle.

So to recap, you don't lean the bike directly, you push on the handlebar in order to get to the angle you want.

Handlebar is the input
Lean angle is the output