With the arrival of the new Moonhawk (awesome btw), I feel like there's a bit of a gap. You have the option to add or remove almost all of the parts, Even the driveshaft. None of the other cars (if I remember) have nearly the options the Moonhawk has I was just wondering if the other vehicles will be updated to the same level? And if there will ever be basic standards for all player made vehicles to conform to?
I was thinking about that too, gabester said the Moonhawk is more complex than the other cars, so isn't he keeping to the same standard level as the other vehicles?
I guess it represents the fact that I'm getting better at making vehicles. The D15, Covet, and Grand Marshal are more than a year old, while the Moonhawk was only started a few months ago. I don't think I could really update the older vehicles to the "standard" of the Moonhawk, nor do I want to (I'd rather work on more cars). Besides, it's nice to have cars like the Covet which are much easier on the CPU and GPU that you can run a lot of at once. As for user-made vehicles: there are no "rules" as to how detailed it should be. I can only hope people do things the way I do them
Your cars are great guys but I have a couple things that you could consider. First the Gavril D-15 has too long of a first gear ratio. It's very hard to offroad on rocks and in mud because it never has enough power to get over the rocks or go faster than about 5mph in mud with the 4.5L V8. On the road though, it is fantastic. Second I love how you have coordinated the automatic transmissions to rev accordingly with varied acceleration. But the minimum gear shift is at 3000rpm. Would it be possibe to lower it to say 2000rpm? Finally it would be great to have differend automatic transmission options available. For example the Gavril Vanster models have a 4 and 6 speed option. It would be great to see 4,5 and 6 speed automatic options for the cars. This is by far the most realistic program I have used and I look forward to the future vehicles and game updates!
The gear ratios are taken 1:1 from the real variant but BNG currently has a problem with how the transmission handles take offs.
Yeah, if you switch to manual mode (Q) you can build revs before launching and the tall first gear ratio shouldn't be a problem. We will also improve the "arcade" launch behavior and add manual automatic gears, so you can build revs in N and then shift into D. Also there's no support yet for high/low range gearboxes (but it's planned).
You do have a very good point and I love all your cars. I guess Im just afraid of things turning out to be something like GranTurismo 5, the utterly massive differences between the "standard" cars and "premium" cars. Even the standard cars have difference between each other. Some are from ps2, and some are from psp and other GranTurismo games. Just makes for a "mashed together" gameplay feel as opposed to a polished one.
It won't be that bad Every car will have detailed interiors with working lights, gauges, steering wheel, shifter, etc. "Standard" cars didn't have any of that.
Are you planning on making the cars more detailed (in polygon count)? This can be adjusted for people with slower hardware too. I'm the kind of person who appreaciate details and graphics (an enthusiast). I don't like textures as a substitute to geometry, do you have the same view of that? When it comes to graphics it's very important that stuff is dynamic and "alive", looks truthful. And not being scripted/animated like many games, I love that about BeamNG.drive! I would love an option to challenge my CPU even more, is it possible to increase the resolution of the points simulated in the physics, and the details/precition? I would gladly test this on my superoverclocked CPU. And by the way, Moonhawk look pretty good. Keep up the good work.
Well I see what you're saying, heh I guess we'll just have to wait and see, this game has a long, prosperous and crash filled road ahead of it Im sure I just find adding little details and consistencies to make the game much more polished. Which could be the difference between a side project and a Triple-A game. That and accessibility. Early Garry's Mod is an amazing example. The game heavily relied on user made content. Problem is, everything needed to be installed differently, different folders, different prerequisites, you got them in a mash of different places, and everyone had a different version of what ever it was. So servers weren't compatible, and the game took massive amounts of room with useless folders and such. People found it too difficult to get the game running online, especially if you didn't want to wait for 30 mins for your client to download form the server. They also found it to hard to add things to in general. Both Portal games, counter strike source, RoR are examples of this. So Garry's mod came up with the "toy-box" ( I think it's called) native menu, where you can download and upload creations, then Steam themselves came out with the Workshop. Made Portal sooooo much more amazing. Long and slightly off topic, but Im sure you see where Im getting at.... And polygons devastate GPUs. Personally Im waiting for someone to use Euclideon's tech. Well as for how the game looks, the devs said it wasn't so important them, just they way things moved, and it's amazing. So I doubt they'll ramp up the graphical fidelity anytime soon, but still looks rather good IMO. I'd agree that textures don't make up for lack of detail, but look at Rage or Just Cause 2, heavily texture based and they're awesome. I quadrupled the texture resolution in Boarderlands 2. Looks 200% better easy.
I would like my GPU "devastated". I will allways have high-end hardware. Those who don't have that can still turn the graphics down. Mostly it looks good. The soft body physics make the metal destruction look as good as probably ever. And when drifting into the sand that looks pretty nice. But there's still stuff like bad reflections, low resolution textures, low polygon count some places, collisions, static objects that wont move or deform, smoke and grass goin through objects. I'm sure a lot of this will be improved upon, but otherwise it will look inconsistent and unpolished. If you look at developers like Polyphony Digital and their games you see how it just looks polished. But BeamNG have strenghts that Gran Turismo lack. That game unfortunately have too many invisible walls, static objects, and need a more dynamic feel. My point is, for a game to look good it need to look believable and real. And yes, the way things move is a big part of this. Just look at old robots imitating humans in movies, that is games imitating reality. I just want to say though, comparing these developers and games is not fair. Not remotely fair at all! And on top of that, what BeamNG has done is impressive, all considered. I haven't played Rage, but have played Just Cause 2 a bit. You could increase the texture size by a million, but it still wouldn't look good from the right angle. And also, under the bonnet there's a texture. There's missing geometry that isn't taken into account by the physics engine. That means, it could be more accurate/realistic, right? That reinforcement is there for a reason, and it would change the way that metal bonnet would deform. See where I'm going with this? And still, it wouldn't look as good because you can't deform textures, they will stay the same no matter what. By the way, looked into Euclideon. I've seen it before, but it didn't cought my attention (very skeptical). I watched some videos and an interview and I must say, it looks promising. But there are a lot of unanswered questions. How would this work with physics, animation, lighning, effects, change of forms, memory, capabilities, licence of the engine and so on. And no, it is not "unlimited". Nothing is unlimited. This is just a different and more efficient way of doing it. I'm sure there are a lot of drawbacks to it. But yeah, very interesting indeed. But they haven't proved this yet, their demos is not proof, as they could have cheated in many ways.
Lower graphics won't lower the polygonal count on the cars in the game so it's not an option. BNG doesn't cater to the 0.1% that might be able to run it with ridiculous poly counts but to the majority. The meshes have absolutely nothing to do with deformation. The nodebeam controls everything(press L or K ingame to view it). The meshes are merely visual to make everything look pretty.
But polygon count is scalable. Tesselation is an example, maybe not the best one, but it shows that is very well is quite possible. It would be fairly easy math to lower the polygon count on a car by changing the settings. Oh, my bad. That sounds very logical, I just tried this. Couldn't see how the game took into account that the bonnet were reinforced and thicker underneath, but I'm not an developer. But still, if it takes into account something that's not there being deformed, wouldn't that look wrong on the model?
Not very easy to implement for car models. Most games don't change the amount of vehicle polygons, only the world around it, as it's fairly difficult to do so without making the car look, well, like crap. Crash it into a wall and look for yourself.
Go to BeamNG-vehicles-pickup-engine.beam and search for it. In the transmission part you can change rpm for gear changes. I did this already and with notepad it's easy. Gesendet von meinem HTC One mit Tapatalk 2
I won't claim anything is easy or not, because frankly I don't know, I don't know programming to that extent. But, editing details on a model isn't too much of a hassle, this is a teqnique that many developers use, switching between models, eg. when you're far away to give the GPU a break. And also tesselation is a good teqnique that makes graphics very efficiently rendered. It would make the game run nice on a bigger range of hardware, both slow and fast. I'm not saying it looks bad or anything, I'm just pondering about stuff that could be a potential improvement.
Well there is no substitute for something that isn't there, but looking at Reco's pants, they look very detailed, but really it's just textures, give good illusions They do have a live interview with Intel them selves running their island on a laptop.... I don't think it would work with animation. Which is one of the draw backs I think but then again, we didn't think 3D animations were possible, never mind with calculating every little tiny shape in a single object, then making them completely dynamic. But building a hybrid engine at first could be possible, half Euclideon and half standard Poly. Even just replacing all static meshes with their process. Think of Skyrim for a second. Sometimes over 90% of the polys you see on your screen are static. So now take all of those away, replace them with ridiculously detailed versions, and then quadruple all your non-static meshes poly count, add post effects. While on the same hardware. It does work, and works just fine, problem is nearly everyone who've seen their tech thinks its a scam just like you (and defiantly me at first), and just want to see animations and dynamic meshes. Which is why releasing tech for geographers and other professional applications was a very good idea on their part. Giving that particular industry what it really desperately needed. Once it's proven there, I can guarantee a game company will pick it up and send competitors to annihilation.
I saw the interview with HardOCP, it looked very real. But for all we know it could've been connected to Wi-Fi, could be a cable underneath the laptop (mine even have dock-connectors there). But no, that is not very likely, so I do have to say that I actually believe it. I just don't believe that it's such a sunshine story. But a little revolution? Maybe. And like he said, they are still working with it, this can be good. Is it possible that this has the potential to take over polygons? Of course. I became excited about it, to say at least. What I wonder then is, will hardware development stop? Why should there ever be new hardware coming out (for consumer market that is)? Will this revolutionary engine cost too much for many developers to employ?
Good questions, but most can be answered. Starting with consoles. With the release of the PS3 and xbox 360 came a new realm of gaming. PCs and Consoles sharing the same games. Even with locked in hardware of the consoles and 90% of games coming out for a short while being console games cheaply ported to PC, but PCs still were developing by leaps and bounds. Considering my rig that Im on now has twice the graphical power that a PS4 has, and who knows how much more then a PS3. That being said, consoles were a VERY GOOD thing for PC gamers in a way. Developers leaned TONS of new tricks in making the best looking game on some very limited hardware. Crysis 1 vs Crysis 2 is an amazing example, Crysis 1 still today can get some terrible framerates. The atom tech they use only solves static polys. Not after effects, dynamic meshes, physics ect ect. So now we can spend less resources on geometry, and use the rest up for everything else. So not only will we get detailed worlds, but crazy effects and high quality physics. Developers would still push the hardware limits. Of course they would, they all want to make the best looking game. So would hardware development stop? Of course not. As for cost, I doubt it. They already have working tools and a development kit, plus their full release of Geoverse which companies are already using. They only thing stopping it (that I can see anyways) is skeptical people unfortunately :/
Well, this time it should be easier to port the games at least as they share architecture to a bigger extent (still differences). That is true, but console games are also defining a limit, they pave the way of both console and PC games. But yeah, optimizations are always good. I hope you are right. Stuff like physics, dynamic gameplay, textures all need an overhaul. Haha, about that... ""they all want to make the best looking game". Not all. BeamNG didn't want that. The hardware development has actually stagnated a bit. Especially with CPU's. Intel has practically been the only CPU maker in town (desktop), and AMD cease to challenge them. Intel has put ARM on their radar (with good reason), as the performance and efficiency is starting to get scary for Intel. To secure their future they are now focusing on power consumtion and efficiency rather than pure performance (like they kinda did before). The LGA 1150 platform is now in total better than the LGA 2011 platform, if you ignore pure CPU performance capabilities. LGA 2011 processors has been based on an 2 old year old architecture, and still is. This summer the product line were refreshed with Ivy Bridge-E (still based on Sandy Bridge). The chipset is also old and rusty, the difference between X79 and Z87 is too big they should be ashamed. At least you can now overclock the base LGA 2011 model fully. And Haswell-E isn't going to be around any time soon. Why? Intel don't give us the attention like before. And frankly, I think they are shooting themself. That market is shrinking, and that hurts PC market. And that hurts PC games. And that hurts... well, you see my point. Domino. Game developers today aren't pushing the CPU's enough, but GPU's, yes. I understand it, because people aren't that into PC building anymore and hardware news aren't as exciting as before. But they aren't giving this away for free, that is not a well working business model. They have probably realized they can earn good money from this. And now I just realized, what if they monopolize the game developer like DirectX? Will it support Linux? Questioning is important.
