Hi Devs, BeamNG.Drive is one of the most amazing sims I've used. I know you get harassed about this all the time, but it is SCREAMING for VR support. I have seen your standard response that it's too involved etc, but I ask you to reconsider. If you don't think it's worth it, and you haven't ever had a good VR driving experience, try Assetto Corsa in VR. You will never want to go back to flat screens again. I am not suggesting hand controllers or anything fancy. I refer back to Assetto Corsa and similar that support VR. Thanks, keep up the good work! To my respected forum brothers and sisters: If you do not use VR for sim racing, please try it on a capable system with a modern VR headset with a wheel and pedals before discounting it as not worth the effort, or that it is too demanding on the CPU/GPU. This is not new technology anymore. I'm not ignorant to, or minimising the work involved, but it doesn't require a full rewrite, just some additions to the coding.
I understand they have other priorities but I feel they should start working on it in the next year or two, even before it’s optimized for preformance. Getting it out there is more important than every single person being able to use it at first.
It is very demanding and it is definitely new as far as gaming is concerned. Demonstrably false. You're wording this as if no one on the team ever experienced VR and they just decided to be dismissive for the sake of it. They know what they're doing.
Its been mainstream for nearly a decade now and has been invented for multiple decades, its not new in terms of gaming.
There's a ton of work required to add VR support. No matter how fun VR is, "It's really great" isn't a valid reason to add VR support. How long something's been around isn't the full story regarding its viability or adoption. How do you think most people will afford this? Cheap VR isn't very good VR, generally- and not everyone's rich. I'd rather play on a decent desktop setup then have a decent VR experience.
I've used several different VR devices on several occasions. I'm not going to comment on VR without having tried it
I suspect part of the problem is the physics engine running at an indivisible 2000 Hz, which means you get inevitable render/physics tick timing issues at all common framerates. (ever tried recording BeamNG in 60FPS and counting the frames?) Something like 2160 Hz would divide with all common framerates (30,60,90,120,144,240) and probably make VR more feasible, but the BeamNG team isn't exactly comfortable with a 10% loss of performance, not to mention the necessary tweaking to all physics systems to deal with new resonances and whatever else that might be influenced by the specific tick rate of 2000 Hz.
I thought this would just be adding another virtual camera view and then having it move according to the headset... seems more than that
Define cheap? I'm running on a GTX 1080... Generations old. Easily handles modern physics based titles in VR at higher quality settings. I'm using a Samsung Odyssey+ I picked up second hand for 500AUD. $1000 gets you a new reverb G2... How much do gamers/Sim racers spend on their Ultra wide high refresh monitor? I am happy to sacrifice some details (reflection refresh rate, draw distance, smoke generation, a level or 2 of AA) to have the immersion of VR. If you have a machine capable of running beamNG with all graphics options high, you can easily run a VR setup with a few compromises.
It's also rendering from 2 perspectives. --- Post updated --- No one needs to reinvent the wheel here. All the work has been done in free libraries like OpenVR, and they just need to incorporate this into their engine.
I've looked through the OpenVR library and the sample code and I can say that all the developers would really need to do is just a few things. First, instead of using the current game engine's projection matrix (used for rendering the 3D scene to your 2D screen), you would use the two projection matrices provided by the OpenVR API to render the scene properly to the two eyes. When rendering the scene, you would render it to two framebuffer objects, which will hold the image of the scene for each eye. And finally, you would call the OpenVR API function to send off the two images from the framebuffer objects to OpenVR, which will handle the rest of it to actually display the images in the headset. So it really is not that difficult to implement.
Idk if it's worth it if only 10 people would be using it and 5 of them would be complaining about the graphics. I don't think the demand on VR is that high, most people don't care about it.
I sent a private message to the devs last week, and they instructed me to post it here. So why are you you trying to stifle the creative collaborative process? If no one asks, and meaningful discussions aren't held, how will the devs know that there are or aren't more than 10 users with VR? It might turn out that there is a considerable percentage of VR users here, or there might really only be 10. Regardless, how do you think the feature-request/roadmap decisions are made in software development? It's not disrespectful to give your opinion and make a request, even if it has been requested before.
120 million? So you think BeamNG has sold 60% of the copies that Minecraft did? This forum has ~340,000 users. Let's say 5% of the the forum members, not total license holders, use VR, that is 17,000 users. At ~$27USD per license, that's nearly half a million dollars. ...Gotta have some perspective.
