axdsilva
Member
SO distracted...
Posts: 403
Registered on: July 2014
|
Post by axdsilva on Jul 14, 2014 22:13:24 GMT
I dunno about tires, but the wheel types definitely have weight, I remember notcing this when I suped up my Ocelot Jackal Coupe fully and added SUV tires for multitasking, and noticed a slight stutter in the initial accel and a split second over-reving sound before I finally started moving. This was eliminated when I switched to high end carbon racer wheels and I accelerated off fantastically.
I'm not sure if it's just SUV wheels, but they definitely give a less torque'd out car a sliver of something at start ups when fitted. after that it jus adds weight, which means it carried itself with a little less control in some bends. However I only noticed this because I was specifically testing SUV rims vs. Carbon rims on my Jackal at the time... in freemode. I never tested this on a track, so I have no idea if it helps or hinders. I just go with high-end or carbon under the universal principal of lighter = less interia
That's what I found
|
|
valleyeve
Member
Posts: 87
Registered on: April 2014
|
Post by valleyeve on Jul 16, 2014 20:42:34 GMT
Broughy just did the video so I'm locking the poll to forever to memorialize our opinions before the testing was done so we can humiliate each other forever. You should be able to comment. Here it is: youtu.be/eVaBBp3Ytb4My dealio: We now know they affect neither braking nor acceleration. The tests also tell us that they have nothing to do with tire traction. I was hoping to see some objective corner testing, but I guess that's something that would take an unreal amount of time to prove anything. I don't know of any other variables that would make a difference in the turns; I'm not a scientist or a tire expert. I'd say this is the best conclusion anyone could definitively make. TY Adam BTW: Chilli, I'd like to get your opinion on this since you seem to know what is going on. Maybe elaborate a bit on Broughy's conclusion, or anyone else for that matter. Heyo, my first post on here after watching Broughy's videos for quite some time. After reading through the spreadsheet that was analyzed and comparing the parameter names and their values with GTA IV I think I'm figuring out part of the puzzle. My take on all of this is that there might be a slight increase in traction once custom wheels are applied. But after that point each wheel type only have differences between "Traction Curve", which is at how many degrees of turning is the maximum grip achieved, and "Traction Spring Delta", which I believe affects how loose the turning will be and lower is better. So why some say certain types are better or not might simply be because of their characteristics involving those two parameters. I've always felt that the muscle type makes my car looser but I still drive across town at about the same speed, I might just brake a little more here and there compared to how I drive with high-end or sports. Judging by that, high-end sports are intended more towards how F-1 is driven as in not turning that sharp. The difference between those two is probably something like "Curve" and "Spring Delta" is tightened up for high-end but "Spring Delta" might be a little looser for sports, but not by much. Basicly what I'm saying is that the amount of traction your car can output is still the same no matter the type, but the driving style changes and that might still be valuable for hot-lappers that want a very specific type of behaviour for a particular track. Some might like a tiny bit more "failure margin" and so they pick something like lowrider or muscle and some might want something as tight as the new non-hipster pants where drifting simply isnt an option. But in the end it all comes down to how early or late you want to reach maximum grip. Muscle and lowrider probably reaches it later and high-end reaches it the earliest. Pick your poision!
|
|
|
Post by CHILLI on Jul 20, 2014 17:05:28 GMT
valleyeve I didnt see the video until now. Broughy's tests are nice but he's also forgetting to double-check the behaviour with the handling lines. I'll give my thoughts on each test in the order it was shown in the video so it's easier to follow what I'm talking about. Keep in mind that I'm not saying wether me or Broughy is the better driver, I'm just stating my observations and take my comparisons between me and him as such. I will never claim I'm better than someone else unless it has been tested and proven. Alright, braking and acceleration. In my initial post I mentioned the "Traction Curve" and "Traction Spring Delta" but I never mentioned the "Low Speed Traction" parameter. When a car is below a certain speed (I dont know what the speed limit is because it isnt in the handling lines) it will always have whatever amount of grip this parameter states. This is why taking off from a standstill is so painful after spinning out and such compared to doing the same thing with a little speed. Personally this parameter can burn in hell but what do you do... But this only covers taking off. So what about braking? Well, it probably wont do much (and Broughy's video proves this). The way the physics model for the wheels work is based on the turning angle. So if you're going in a straight line and braking the game will give you the same results because the wheel types only change the "Curve" and "Spring Delta" (well, I think so anyway), so to test it he would also have to be turning and braking to really see what the wheels change. This is because, again, the physics model is looking at how much the car is turning. That is the only time the grip will increase according to the "Curve" and "Spring Delta". And if I'm not mistaken the "Curve" defines where the "Traction Max" has fully kicked in, stated in degrees. So again, to see the difference there you'd have to turn and brake but that becomes hard to test because of human error. Driving around a track. The laptimes will probably vary slightly (not counting human error) because of what I said above and in my previous post. The minimum and maximum traction remain the same, it's only the uh... transition between those two that change based on wheel type. So to really find out what wheel type does what you'd have to truly push the car, drive it like you stole it (Fast and Furious-style) to really find out how far you can really push the car. If you're taking a safe line all the time the differences will become smaller and smaller between the types. Other peoples' tests. Some say Offroad is faster, some say Lowrider is faster. I wouldnt say that's necessarily wrong because the driving style will play the biggest role here. Depending on how you drive one type will get used more efficiently than another. But give the same car to a different driver and the results will differ because of the driving style of that other person. I know for sure that I've felt the difference when going all-out apeshit with the cars in freemode. And I have mentioned to some that I'm surprised that Broughy experienced so much oversteer in his Entity while mine sits on the ground like glue. I use Sports wheels while I think he used High-end wheels. So that along with how me and him drive is the explanation why things might look the way they do. Other factors. I made a thread a little while ago where I go through some mechanics the game doesnt tell you. Depending on how you combine braking, turning and acceleration (order is based on the situation) the behaviour WILL change slightly. I might overlap my braking and acceleration slightly while Broughy doesnt, and that might also make my car feel a tad grippier than it might for him granted he does no overlapping (explanation is in the Mechanics thread in the GTA Racing section). So because of that brake & acceleration overlap mechanic (heck I probably do it unconciously at this point) the results from person to person might also vary more or less depending on the track and how they handle said inputs. The result... Pick what feels right for you. It really is as simple as that. The maximum and minimum grip will remain the same, taking off will always be the same (because of the "Low Speed Traction" parameter) and braking in a straight line will also be exactly the same no matter the type. It's when you drive in such a way that you're nearly killing yourself the difference becomes more clear. If you're taking it safe then go with whatever I suppose unless you can still tell the difference.
|
|
|
Post by Trowa on Jul 20, 2014 17:36:13 GMT
CHILLI I really enjoy reading these posts. To me this all makes perfect sense and answers a lot of questions if its true which I'm leaning towards believing it is. I wish we could get the game code for you to analyze further. I don't how to even go about that and while Broughy's test was quality and answered the question in its most basic form I was left wanting to know more in regards to your post. After talking to you for several hours about the topic and gta mechanics in general I walked away feeling both enlightened and thirsty for more knowledge. That thirst has not yet been quench. As you said in our chat the only way to truly know for certain is to have a look at the code and look for anomalies. Personally I don't care what others believe or if they want to take Broughy's word and just leave it at that. Thats up to them but that's hardly enough for me at this point. Admittedly I don't use 1 specific wheel type on all my cars. I prefer to pick either tuner, high end or sports on all my cars and usually look for a suitable wheel for my car in that order. So its not like I choose purely based on performance or at least what I believe to be the best performance which is offroad>Tuner>sport>Lowrider>Highend>Muscle However I do believe knowledge is power and I always want to know more. In some ways Chilli's post is plague on my mind because now I am forced to ask myself more questions. On the other hand it makes me feel that there is still more to be learned about the cars in GTA. That is a good thing and could be just one more advantage in races if we figure it out. As far as Im concerned this matter is still wide open for discussion until we get our hands on some data taken from the game code because due to human error there is no suitable test available. All we have to go on for now is feelings more or less.
|
|
|
Post by Hale on Jul 20, 2014 23:40:59 GMT
As a racer i can say that feel is more.important than any statistic. So go with what feels good
|
|
haydugjr
Member
Posts: 359
Registered on: March 2014
|
Post by haydugjr on Jul 21, 2014 3:40:19 GMT
As a racer i can say that feel is more.important than any statistic. So go with what feels good SPINNERS
|
|
|
Post by Hale on Jul 21, 2014 16:11:28 GMT
*Facepalms*
|
|
Twilicane
Member
Posts: 1,953
Registered on: March 2014
PSN ID: Twilicane
Xbox GT: Twilicane
Discord: Twilicane #5303
|
Post by Twilicane on Jul 21, 2014 17:22:47 GMT
I just like it if it looks cool.
|
|
|
Post by Hale on Jul 23, 2014 15:48:15 GMT
Twiliswag
|
|
valleyeve
Member
Posts: 87
Registered on: April 2014
|
Post by valleyeve on Jul 25, 2014 4:34:04 GMT
Chilli you definitely have a point there about the game mechanics. I really have a theory about the actual modeling of the wheels themselves and how Rockstar decided to indicate the parameters of things like weight distribution, geometry, and the focus of each wheel type (sport, SUV, off-road). I actually have a little schooling in 3d modeling and it is completely possible that they didn't trash the geometric and weight-oriented physics for each wheel type because of the ease of keeping that data in the system.
When you do model something using the Euphoria physics engine, I'd have to wonder how each item is calculated in terms of weight. It is possible that has to do with the actual makeup of each wheel. It is also very possible that the geometry influences the weight distribution as the wheels turn as well. This could allow for some small differences in the type of wheel you choose when you go around a track. It is also possible that the wheel types have a focus for each type. SUV may retain some characteristic(s) of SUV vehicles and high-end may retain some high-end vehicle physics.
These differences, however, would need to be completely independent of the set-in-stone stats for the car. Traction, speed, braking, and acceleration would not be affected. It would be based solely on wheel geometry, weight, and maybe leftover physics from each vehicle type which would only affect a car's handling. I don't think were going to get the answer, though, as it is a very subtle difference if any. Personally I believe I can feel that difference but the data that we all get for these wheels in terms of lap times are so near to eachother, I don't think we'll get a definite answer. It would require hundreds of lap times from very consistent unbiased drivers. But there's my opinion.
|
|
|
Post by CHILLI on Jul 29, 2014 18:14:28 GMT
I have never worked with the physics engine GTA uses (assuming it's the same for all RAGE games) but I would assume that it's very close to PhysX which I have worked with quite a bit.
Modding GTA IV was never something I went so far to try myself, but I have seen screenshots of people working with vehicles in various modelling programs, namely 3DS Max and Zmodeler. I dont know if anyone has tried to increase the sides of the collision cylinder the wheel uses (each wheel is an instance of the original unless there are different wheel models, those use seperate originals that are instanced) but I'd assume the engine only reads the size of the cylinder and the rest is taken from the engine itself and not each vehicle.
Mass, center of mass and the resulting density, if the game even cares about that, is most likely calculated the moment a vehicle is spawned, which in turn means that every physical object is treated the same way regardless if it's a car or a fence. The only difference is that all vehicles has parameters that allows the creator to change the total mass aswell as the center of mass "CoM". So as far as my knowledge goes the physics engine (Bullet Physics Library. I'll call it BulletPL in the future) will set the mass for the entire volume to what the handling line states. Wheel mass is another hardcoded variable that I would believe is the same for all vehicles using wheels. Based on how PhysX treats wheel mass, all it really does is counter the torque applied so there's no real need to have two conflicting parameters. Also, this is proven by Broughy's video because his acceleration remained the same with all wheel types. If the mass did change they would differ slightly but that doesnt seem to be the case.
Oh yea, the weight distribution. What V has added to it is additional inertia for each axis of the vehicle. Judging by their numbers they are multipliers of the original expected inertia since they rarely go above 2.0 and if they do it's not by much. Most linger around 1.3 - 1.6 or so. This is probably to give the vehicles a feeling of more weight to them because the anti-roll bars negate that quite a bit (that's probably why so many thought the cars in IV felt like boats. No anti-roll bars were simulated there).
About the "stock" wheel type though, I actually dont know. I just assumed that it would point towards the wheel set that the creator thinks fit it the most so I've never taken my testing that far.
Final notes: The physics engine is Bullet Physics Library which was confirmed in an article that was released soon after the remastered version of V was revealed. I cant remember where I saw it though. Why I'm pointing this out is because Euphoria is actually an AI system which purpose is to let ragdolls stand up instead of just collapsing on the spot. Try pushing some people around in IV and you'll see it doing its magic. It's in V aswell just... toned down to say the least. Look up "Natural Motion" to read more about it, those are the guys that made it. And Bullet Physics Library is free and open-source so if you want to get an idea of what Rockstar is doing with their games, feel free to read through thousands of lines of code. Btw, is this beyond off-topic?
|
|
valleyeve
Member
Posts: 87
Registered on: April 2014
|
Post by valleyeve on Jul 30, 2014 14:33:27 GMT
No I don't think it's off topic. You actually have it nailed down. I thought the Euphoria engine was the physics engine that supported everything, but I never looked into it in particular. I'm not about to read through and decipher lines of code but I agree its probably what the developers deemed fitting when they used the engine. There's no doubt engines are growing more and more impressive as time moves on and they are fully capable of simulating a good deal of realism. I would have to leave any more research to a coder because I'm already overstepping my grounds with questioning the physics let alone coding. However, I would think someone would have looked into it already... So Broughy's tests confirmed that traction isn't affected and the handling differences, if any, are extremely marginal. Even if we compiled tons of lap times, the margin of error of the data would include an overlap of all of the wheel types. I'd say the case is closed unless someone wants to prove it by digging through lines of code, which I think people have tried to do and came up with nothing. I'm going to go with what I like for my wheel types and I'd suggest everyone else do the same. We're not going to find anything else.
|
|
|
Post by CHILLI on Jul 31, 2014 5:27:28 GMT
So Broughy's tests confirmed that traction isn't affected and the handling differences, if any, are extremely marginal. Keep in mind what I said in one of my previous posts though. To get enough data to know for sure we'd need one of two things: 1. Have the AI drive a car for us, pushing the limits of it and let it do one race for each wheel type. or 2. Run and collect lap times of so many tracks and laps that it might aswell become a full time job. That's why I keep saying that it's better to go with what you feel makes said car easier to more comfortable to drive, even if it might actually end up to be proven as placebo. Personally though, as mentioned before, I want to believe it does something but to be really honest I cant think of a good way to explain what changes to look for. Just give your car a few runs around some roads with lots of turns and just have some fun with it. That's what I did about a week ago and I cant help but think something changes between the types. I'll consider to test it on more cars in the future until I'm 110% sure there's a difference and hopefully be able to explain what to look for. When will this happen? I'm not even sure myself but it's tempting to do, so I might do it next time I'm on the game.
|
|
|
Post by Trowa on Aug 29, 2014 20:31:58 GMT
Everyone saying we are crazy may now eat their hats. Good day sir.
|
|
Lord Nemesis420
Member
Many are called, few are chosen correctly from a police lineup
Posts: 75
Registered on: July 2014
|
Post by Lord Nemesis420 on Aug 29, 2014 21:00:13 GMT
Wow.
So, Trowa, in the other thread you were possibly thinking that Muscle and Low Rider were the worst.
Do still feel that way or do you agree with Broughy about the sidewall height be the dominant factor (placing Muscle and Low Rider in the middle of the pack)?
|
|