Renegade and Renegade X: Vehicle physics

[XD_ERROR_XD]

Hey guys! It's been a long time since I've posted here! And i'm back with a couple of videos and feedback. (well, I was never gone, I was just lurking :p). This time, I'm going to talk in further detail on vehicle physics.

Vehicle physics in Renegade X has always been an issue. And to me, their bumpyness and other issues just doesn't make it enjoyable to play. I've made some efforts to pinpoint the issues.

 

I've gone in quite some detail on how C&C Renegade's vehicle physics works, and have finally finished my far overdue video on it. here it is:

Yes, the video is a bit of a mess, but I was a bit too tired of editing it to make it any better than this. Sorry

In case you're not willing to watch a 12 minute video, it explains in detail how parameters like total weight, aerodynamic drag, gravity scale, suspension spring and dampening constant and other factors make vehicles in C&C Renegade work. Don't get me wrong, physics in C&C Renegade is still quite a mess at times, but I do think they nailed it pretty well on the vehile aspect. I also drive a lot in the video, so it'll work as a nice reference to my comparisons of Renegade X's vehicle handling.

 

So yes, I mentioned bumpyness in Renegade X. What causes this? Well, to bring across my point, I have recorded a few short clips. Here's the first one:

From watching this video, it should be clear where all the weight of the Light Tank is placed. The thing wedges very easily, but doesn't move at all. This makes me assume that the weight of the vehicle is barely spread out past the middle. This needs to be worked on. IF this isn't the case, and the weight is spread out perfectly, there is quite a major issue on how objects react to directional forces, because a stationary light tank reacted to a horizontal force by moving vertically.

 

I'll show another clip here to prove my point:

 By giving the rear tracks some air by going up this slope, and then immediately reversing, my vehicle is able to hold up my rear part of the vehicle in the air for far too long to be realistic. This further supports my theory of bad weight distribution amongst vehicles.

However, you can also see another issue appearing in this video: How is my Light Tank able to accelerate at the same speed on two tracks, as it would do on four tracks? This makes no sense. Moving while only having grip on a slight portion of your tracks on a surface, somehow allows that part of your tracks to harness the full power of your tracks to move at the same speed. Yeah no, it really should be less grip = less speed, Having only one or two tracks on the floor should not mean that the other tracks somehow "gives" all of it's power to those tracks.

 

Here's a related issue with tracks and power:

Take a look at how at how fast i'm going at the beginning of the video. Then take a look at how much I slow down later, while making constant turns. My tank just doesn't slow down. Tracked tanks manage to turn by lowering the turn speed of the tracks on either the right or left side, which causes it to slow down significantly during turns. This tank drives as if it has wheels. Maybe this can be explained, by how infantry somehow run faster going uphill, which makes them still go at the same speed in the X(?)-axis. I guess you can put the same idea on tanks aswell. It basically needs work.

 

But, there is one more issue regarding vehicles! (or, at least the Light Tank, i've only tested it on that vehicle so far):

Notice how part of my vehicle clips into the Airstrip. Still remember the issue with tracks I mentioned earlier and how they're able to provide abnormal power? This makes it even easier for my tracks to grip unto surfaces they should not be able to touch, which makes them somehow able to climb near-vertical surfaces, like the base walls in whiteout. Also notice how this clipping causes my vehicle to sometimes stick into the wall? Yeah, this and the previous issue would be mostly mitigated by improvied collision detection (basically a larger collision model).

 

So, in short, the suggested changes to vehicle handling/physics are:

- A better weight distribution amongst tanks), and/or better response to directional forces;

- An actual implementation of the Pythagorean theorem (just google it if you don't know what it means :P)

- Calculation of vehicle torque based on amount of grip, instead of just having grip at all (or just give each track their own amount of power, instead of sharing them across);

- Bigger/better collision models for vehicles.

 

I Hope this helps clearing up some issues, and I'd love to hear your responses! Feedback is always welcome!

 

Edited March 12, 2017 by XD_ERROR_XD

[sterps]

Good job on the videos.

Wouldn't this just mean the corresponding values in RenX need to be adjusted?

Is there a way that players can modify vehicle settings via the console like in Renegade, say during a skirmish game?

Edited March 13, 2017 by sterps

[Fffreak9999]

I don't believe that vehicle physics and other settings like that can be adjusted by the in-game console.

[RypeL]

Underlying UE engine physics. Can not be changed with our UDK license. Only thing we could try was some tricks and workarrounds like cutting your vehicles grip shortly after hitting another vehicle wich is why you couldn't push the ltank in your vid. We tried, a lot, to come up with how it's now. UE vehicle physics are designed for non serious gamemodes and most of it we cannot change as it isent our engine.

Good thread though and maybe it gives ideas to try more like looking at collision boxes again. But don't have much hope. Weird physics came with the engine choice.

[yosh56]

Basically what RypeL said. We're working with the very same  engine that produced the MAKO from Mass Effect 1... And they had a full licence to at least try and curb some of the oddities in the baseline vehicle physics. 

If you don't know what the MAKO is, look it up and you'll understand why it is one of the most hated vehicles ever in a video game. 

[Hohndo]

4 hours ago, yosh56 said:

Basically what RypeL said. We're working with the very same  engine that produced the MAKO from Mass Effect 1... And they had a full licence to at least try and curb some of the oddities in the baseline vehicle physics. 

If you don't know what the MAKO is, look it up and you'll understand why it is one of the most hated vehicles ever in a video game. 

Like Voldemort, you should never invoke its name.

[Havoc89]

Appreciate the whole investigating, but we've done the exact same thing for... well since we first got the humvee ingame in the ut3 version. Vehicle physics, and physics in general cannot be compared, especially in older engines like w3d, and a not as old engine being ue3. Neither use "real" physics, they both use simulated physics that are less computational heavy and thus easier for the engine to handle. And depending on the requirement, simulated physics can be extremely simple (w3d), or a bit more realistic (ue3), granted udk did update vehicle physics a tad bit. Simulated physics are custom mathematically derived equations that are far less complex and have less variables then real world counter parts. Thus the math behind the physics in w3d and ue3 will be completely different. To compare the two is like comparing mario kart vehicle physics to halo 1 vehicle physics. They are VERY different.

You actually can indeed edit vehicle properties in the editor while playing in game, by looking at the vehicle and typing in "editactor trace". This is what we used to dial in vehicle physics. Originally when I brought in the humvee into the ut3 mod, I did exactly what you're doing which is comparing every little detail to from w3d ren. Trust me when I say this, it is not possible to achieve ren's vehicle physics since in udk.

At the same time I would argue that ren's physics aren't all that great. They feel extremely floaty and do not give you the feeling of driving around in a vehicle capable of blowing shit up. I personally like the more grippy more weighty, and more responsive vehicle controls in renx. The only problem with them is that they are entirely server side and thus lag like no tomorrow, and they can climb up 90 degree walls.

I do like the idea of having a more "bouncy" approach to collisions since that forces the vehicle(s) to separate and not adhere to each other like glue. I'm sure we can work that part of it into the physics, but might be at the cost of other elements of the vehicle physics we may not want to remove/edit.

Only way to replicate ren's physics is to create an entirely custom simulated vehicle physics system, which means it wont run natively and hense may be just as expensive to process even if it is less complex. But like I said, I don't agree with replicating ren's vehicle physics 1:1 I think renx vehicles are successful in other means, and the biggest draw back for me is the adhesion and server lag problems. If those two could be addressed then we'd be far more successful with how they operate and feel.

[XD_ERROR_XD]

Thanks for the feedback. My intent was not to specifically make RenX's vehicle combat a 1:1 copy of C&C Renegade's. I've heard of the stories of how UDK's vehicle physics is a mess, but I never knew how bad it was, or how the situation in Renegade X looked like exactly. I wanted to make this video just in case, so comparisons could be made more effectively. Renegade's physics was still also a mess at times but driving vehicles felt more natural to me than in RenX, even with how lightweight they all felt and stuff.

 

If the vehicle adhesion issue could be mitigated, and maybe threaded tanks turned a bit slower when turning, it would already make me a much happier person. I believe optimising the collision boxes of each tank further could already greatly mitigate this, not only on the front of each tank, but also on the sides when tanks hit a wall from an angle. Just so it's more difficult for only the slightest portion of your tracks to hit a surface you don't want. Maybe vehicle blocker properties could be further adjusted too, but I have no knowledge of this.

 

RenX has gone a long way, and it's still going, but my personal opinion is that vehicle physics do need more optimising if a 1.0 release is still on the schedule for this game. Again, thanks for all the responses!

Edited March 15, 2017 by XD_ERROR_XD