Vehicle Physics Pro beta, feedback welcome

Does the truck have solid axle suspension and cabin suspension?

 

Not in that example, but could be easily implemented with a couple of VPVehicleJoint components.

 

In regards to the VPVehicleJoint can that include any kind of joint like spring? Do you discuss that in greater detail somewhere?

 

Speaking of suspensions, generally most cars have McPherson suspensions for their rear axle, these suspensions act differently, when they are under pressure the wheel -as is shown below in the video - kinda goes forward. Now I was wondering if it is possible to recreate this behavior in VPP or not.

Watch the video here. It was hard to find a good video, I guess that video is good to see what I mean.

​

 

Yes, you can configure a damped-spring behavior with adjustable rates, and optionally a maximum force to be applied for keeping the parts together. VPVehicleJoint is a convenient wrapper that creates ConfigurableJoint components in runtime.

That's not a McPherson suspension, but a trailing arm suspension. In that video the pivot is at the bottom of the chassis, so the wheel moves forward as its center rotates above the pivot point.

I can speak of it directly, as my own car (Citroen C5) has a dynamic suspension with McPherson-type set-up at the front wheels and trailing arm at the rear wheels. When I extend or compress the suspension to its limits then the front wheels (McPherson) move vertically, while the rear wheels (trailing arm) also experience a slight longitudinal displacement.

Still, in VPP the current suspension set up is the simple vertical setup provided by the WheelCollider. I plan to experiment with advanced suspension types in a later version.

 

Thanks for the info. Glad to hear that the "later versions" would have those kind of suspensions.
My own car is exactly like yours - in case of the suspension setup, well it is not pneumatic (or dynamic) - (I have Peugeot 405 Persia) but I believe I was wrong in calling the suspension names .

 

Hi Edy!

Any rough guesses when the first version of VPP is released or is there still some major features to be implemented? Eagerly looking for that standard version.

 

Estimated date is late September. The last feature planned for this release is the Steering Aid, and it's nearly completed. Next step is preparing the demo scenes, examples, 3D, and wrap up the package.

 

Awesome! All the best for that last stretch and hope you can celebrate this big milestone soon!

 

I want the free version of this asset for now.

 

Hi! Looking through Audio add-on component and testing demos, do you have plans to update the engine sounds to be more versatile? Mainly, it would be great if there was the possibility to use different samples for engine under load, engine no load (like releasing the throttle while driving) and engine idle.

Trying to achieve a decent sound for those situations with just one sample feels quite tricky. Pitch and volume don't seem to give enough room to play and even with good sample the engine sounds to be missing that "something".

Then again, those engine extras (turbo & transmission) bring a really nice touch! One nice addition would be to have that exhaust backfire (burning fuel in exhaust) sound when releasing throttle from load.

EDIT: Out of interest, a question about Unity 5.5: Does the update to Physx 3.3.3 bring any positive/negative changes regarding VPP?

 

@mikko_to The engine audio and effects for Vehicle Physics Pro will be provided by The Audio Guys:

http://theaudioguys.co.uk/
https://twitter.com/VehiclePhysics/status/761132043627528192​

I agree on the current audio component being very simple. Still, it can be used as example for writing a custom audio component. I plan to improve the included engine audio component for combining several samples. It won't be included for the first release though.

No significant changes. The new contact manifold system seems to bring slightly more realistic reactions on impacts, but there's no changes on the WheelCollider at all.

 

Great! Maybe they can also provide some tips on how to improve the audio component when you have time to fiddle on that.

Just noticed when trying the Ferrari Playground demo with manual gears: When revving the engine on neutral / clutch pressed, revs take quite long to go up and down (testing with keyboard), as if the flywheel would be really heavy. Also, when the car is upside down (definitely the most common thing in car games ), it feels that tire inertia doesn't affect it either, revs take about the same time to go up/down with engaged gears or neutral. Design feature or something what we can affect?

Good to hear about Physx 3.3.3! Still lots of work before the first release of VPP?

 

The settings for the Ferrari are complete and shamelessly invented. Try reducing the engine inertia value for a lighter flywheel.

Wheel mass indeed affects the rev time. But the actual mass is quite small compared with the engine power, so the effect result mostly negligible. Try putting a lot of mass in the wheels (x10, x100) and you will definitely notice it.

No more features will be developed for this release. I'm now preparing the 3D models, art, scenarios, examples, tutorials, etc. I'm also writing a blog post on the latest developed features.

 

It's been a long time that we've heard about this package. Could we say that hopefully it is near to the final release date?.

 

Yes! I'm writing a blog post with the latest news and developments. In addition, new demos for mobile and desktop will be released hopefully this month.

I'm working on preparing all included assets, 3D, example scenes, tutorials, etc.

 

Just tried out your demos and I have to say the driving feels quite nice. Been very happy using CarX for my game but can imagine switching to your system - just need to test your physics on my cars to be able to compare. Keep up the great work!

 

Hi, @Edy , Anychance for mac build demo version ? I am using mac os, could not test the current demo.

p/s: Your physics are using Unity's built in Wheel Collider or Custom Collider ?

 

@CoderPro Yes, I'll build the upcoming demo for Mac as well.

VPP currently uses a standard WheelCollider for the suspension and contact detection.

 

Waiting for mac demo version.

Anyway, when you release your kit ? I was waiting it for long time.

 

As soon as the demos are completed. It's the last step for the release. Demos will include the sample vehicles, code, UI, etc so they will be a perfect kickstarter to Vehicle Physics Pro.

 

Yes. And now, waiting for mac demo version !

 

Any way to get on a mailing list, for when this launches? I'd like to be one of the first to purchase.

(I don't use Twitter...)

 

You may create a rule on If This Then That (IFTTT) so you could get an email whenever I post a new tweet.

- The trigger would be Twitter > New tweet by a specific user (@vehiclephysics)
- The action would be Email > Send me an email

 

Cool, thanks!

 

When VPP release ?

 

Hi Edy! Work still in progress? Any unexpected difficulties from new Unity versions? All the best for any development activities!

 

Any updates? Been waiting for this!

 

Any chance for mac os version demos ?

 

Thanks for your interest!!

Intense work in progress! I'm dedicating the first half of the year exclusively to the documentation, tutorials, example vehicles, 3D models, demos, etc. Yes, Mac OS demos will be available as well. I'll develop a complete example scene where you can choose the vehicle, control method and location. Then I'll build the demos out of it.

 

Hi, not sure if this has been asked before, do you ever plan to implement some form of softbody tire system, even if it is faked (e.g. wheel radius calculation or different meshes shown if bump is hit), just hoping

 

Yes, a visual deformation system for the tires is on the works. There's an initial prototype working already, but it still needs some polishing.

 

Whoah! I'd love to know / test more! If you need any testers please count me in

 

Hey Edy, love Vehicle Physics Pro - my team has been using it for over a year, and it’s been working great, exactly what we need. Here’s my question: we originally downloaded VPP back in late 2016, and due to a few factors, we’ve never downloaded an update. In the last year, has the solver been optimized for better speed? We often have 50 cars in our scene, and it really brings our framerate down, even though most of the cars only use 1 integration step (the player car uses 2). I’m wondering if we’ll get a noticeable performance increase with a new version.

 

@Dreamback Thanks! A commit dated 08/12/2016 introduced a significant optimization that removed a RayCast per wheel. That should make a difference. Also, a further optimization run is pending. This one should improve the performance even further.

 

Hi,

Does anyone have any tips for setting up wheel and pedals with Vehicle Physics Pro?

I'm a long time owner of VPP but don't have much experience with it, so apologies if I'm going over old ground.

I have a Fanatec Club Sport Wheel Base and Pedals which are regarded as two different devices on Windows (and I assume its common for the pedals and wheel to be different devices?).

Attaching one VP_Device_Input component to the car gives me the wheel functionality but the pedals dont do anything.
It then occurred to me that I may have to add another VP_Device_Input with Selected Device to 1 instead of 0, to get input from the pedals. But still having problems.

I'm trying to get a video together driving round a CityGen3D scene using VPP, and it was daft using keyboard input when I have the hardware, so thought I'd try and set it up

I'm just using a VP_Device_Input setup from one of the demos, which seems to support Thrustmaster and G27 devices.
But I'm basically just wondering about the workflow for getting my device to work properly.

Any advice gratefully received!


Edit:
Looking into it more, I can get pedals input by only having one VP_Device_Input and Selected Device set to 1, but then I obviously lose the input from the wheel. I get that with Selected Device at 0. But having two VP_Device_Input components (0 and 1) doesn't seem to get the input from both. In that situation they both return input from the wheel despite having different Select Device values.

I assume a vehicle is intended to only have one VP_Device_Input component attached, so not sure if there's a workaround for this as I need input from two devices. Maybe the Fanatec system is less common than I thought and most wheels and pedals are regarded as one USB device (and therefore data for both is passed through one VP_Device_Input component)?

Does VPP support input from several devices and I'm just approaching it wrong, or is this unsupported?

Thanks,

Richard

 

VPDeviceInput is currently hard-coded with the mappings for Logitech G27 and Thrustmaster T500 RS. It's planned to update the component to a more generic system supporting custom mappings. For now, you may just clone the VPDeviceInput component to a new component, then modify it to use the mappings of your device. You may get the device identifiers by enabling Debug Info in the component in Play mode.

 

Thanks for your reply Edy. Just to clarify, are you saying it should be possible to get both the wheel and pedals input via one modified VPDeviceInput. Or I need two modified VPDeviceInputs attached to the vehicle?

My issue is very specifically that the pedals and wheel are two separate USB devices and its seemingly not possible to get both inputs via one component if they are different USB devices.
I've used the debug info and I can see the pedal inputs work, or the wheel inputs work (depending on whether the VPDeviceInput is referencing the wheels or pedals USB device), but I cant get them to work together at the same time.

I assume the other two supported devices are getting all inputs (wheel and pedals) via one USB device. I can imagine any wheel/pedals system that is considered one USB device by Windows is relatively easy to configure. But I'm struggling with the Fanatec setup. Not sure if anyone else has come across this issue, but it maybe very specific to certain devices where the pedals and wheels are different USB devices with their own connection to the PC.
(Or its a common situation and I'm just going about it wrong!)

 

We have three different sets of steering/pedals, and only one of them (Logitech) has wheel and pedals in one device. The other two have them as two separate devices (and they need different settings anyways to make sure all three sets work roughly the same), so we made our own input system, inheriting from VehicleBehavior, replacing VPDeviceInput. That one input system uses a separate “pedalInputDevice” and “steeringInputDevice”, for example:

Code (CSharp):

1. pedalInputDevice_.GetStateUpdate(out m_pedalstate);
2. steeringInputDevice_.GetStateUpdate(out m_state);
3. ...
4. float steerInput = steeringInputDevice_.GetWheelValue(m_state);

 

Thanks for your reply Dreamback, I really appreciate it. It's reassuring to see how you did it and I'll take the same approach.
My only worry now is that I'll get it working and spend too much time street racing around my scenes trying to beat my best laps!

 

Is this system only for flat terrain?

I have tried to make a mesh deformation system, but since the WheelCollider component contains only one Raycast, the deformation of the tire becomes bad on uneven terrain.

 

i want to ask, where are the files for the D-BOX Motion Cueing System ?

 

Yes, it will be assuming flat terrain. A more generic system would involve some kind of softbody physics, which is out of scope at the moment.

Repository Vehicle Physics Specialized Assets (available in the Enterprise license).

 

hi,
i try to make a delayd input for your system but it works not very well.

i have:

Code (CSharp):

1.     private void FixedUpdate()
2.     {
3.     StartCoroutine(ApplayDelaySteer(.5f));
4.     StartCoroutine(ApplayDelayGas(1f));
5.     StartCoroutine(ApplayDelayBrake(1f));
6.     }
7.  
8.  
9.     private IEnumerator ApplayDelaySteer(float HowMuchTime)
10.     {
11.         float vehicleSteer = vehicle.data.Get(Channel.Input, InputData.Steer) / 10000.0f;
12.         yield return new WaitForSeconds(HowMuchTime);
13.         vehicle.data.Set(Channel.Input, InputData.Steer, (int)(vehicleSteer * 10000));
14.     }
15.  
16.     private IEnumerator ApplayDelayGas(float HowMuchTime)
17.     {
18.         float vehicleThrottle = vehicle.data.Get(Channel.Input, InputData.Throttle) / 10000.0f;
19.         yield return new WaitForSeconds(HowMuchTime);
20.         vehicle.data.Set(Channel.Input, InputData.Throttle, (int)(vehicleThrottle * 10000));
21.     }
22.  
23.     private IEnumerator ApplayDelayBrake(float HowMuchTime)
24.     {
25.         float vehicleBrake = vehicle.data.Get(Channel.Input, InputData.Brake) / 10000.0f;
26.         yield return new WaitForSeconds(HowMuchTime);
27.         vehicle.data.Set(Channel.Input, InputData.Brake, (int)(vehicleBrake * 10000));
28.     }
29.  

can you help me and fix this pls?

 

Awesome- I liked the tire/smoke effects

 

This won't work because you start the coroutines every frame. What exactly do you want to achieve?

 

hi, thx for your replay. i want to delay the input. now i found a way but iam not happy with it. maybe someone here have a nice solution

 

If everything should be moved you should check out Queues. Fill a queue with the input every frame and for example in a coroutine wait that time and send all the information to VPP and take it out of the queue.

 

Can someone please help me?

I recently updated from Windows 7 to Windows 10, but now my G29 steering wheel isn't working with the VPP alpha demos. I believe it is something with the SDK not picking up a dll from dx11. Is someone else experiencing this problem?

Another strange thing: before the demos used to crash if the steering wheel wasn't plugged in, but now it starts up fine with no crash.

The Logitech SDK for Unity also doesn't pick up the wheel.

I really enjoyed the VPP demos and really want to get this working.

 

Demos are old and don't work quite good with new Logitech wheels and Oculus SDKs. I'm now working on updated demos.