Tanks - tracked vehicles driving and handling

[Imperator[TFD] 444]

29 minutes ago, xxgetbuck123 said:

 

Allows communities to field more armoured vehicles onto the map, which in turns allows for bigger battles and overall more variety in mission layout + objectives due to the higher number of armoured vehicles. 

 

I love it. 

 

Spot on mate.

[SuicideKing 233]

On 10/15/2017 at 3:42 AM, teabagginpeople said:

 

Hell Alls you gotta do is hop into a wipeout pressing forward.  it will take off as you're still climbing up the ladder.

 

Yeah, was going to make this point too. Heck, I had made a ticket about this back when Jets was released...

[scotg 204]

I've implemented the drivingstickLeft/drivingstickRight animations to my tank. Since it has detailed interior models, I'm thrilled to be able to use this! As awesome as these new animations are, there is a major flaw. As I understand it, IRL, tanks have a gas pedal similar to a car, and toe-heel gear shifter similar to a touring motorcycle (i.e.: these are foot controls). Steering with hands works because pulling on a lever applies a brake to the turning-side drive wheel, thus the tracks. It's like having two E-brakes - one for each track, and there is no pushing motion except to release the brake after pulling. In contrast, the E-brakes in game move forward when forward thrust is given, and backward when reverse thrust is given. To properly animate sticks, they should be brake-input-controlled rather than thrust-input-controlled. I don't know how problematic it might be to change all that, but it's worth pointing out.

There might be some tanks that utilize the sticks for full tank control, but they are the exception to the norm. Also, even if they were the norm, then there's also the issue of partial thrust animation. Unless steering, the sticks currently go either full forward or full back.

Before posting this, I tried changing some values in the animations to limit the movement of the sticks. It was half successful and half worse. Limiting the maxvalue to 0 disables the sticks from moving forward, but also disables any forward-steering animations. It also still animates when reversing without brakes.

To summarize: Most tanks commonly use brakes for steering. A3 steering control animation should only be activated when the brake for the corresponding track is applied, however it is calculated.

[kllrt 154]

We are already aware of this, it will be changed :)

[Peter S 33]

Is it possible to get 'incremental speed control' for Vehicles and tracked vehicles esepcially in terms of this DLC like you have on Fixed wing aircraft and AFM for Helicopters? It is incredibly frustrating to drive and use the vehicle with other vehicles in a formation or in general when you are technically incapable of setting a desired speed for the vehicle without humping the gass or resorting to undesirable mods. This could also make it easier to use pedals and/or XboX controllers where you get only a fraction of 100% speed when only pushing the button/pedal down slightly.

[x3kj 1247]

2 hours ago, Rossmore said:

Is it possible to get 'incremental speed control' for Vehicles and tracked vehicles esepcially in terms of this DLC like you have on Fixed wing aircraft and AFM for Helicopters? It is incredibly frustrating to drive and use the vehicle with other vehicles in a formation or in general when you are technically incapable of setting a desired speed for the vehicle without humping the gass or resorting to undesirable mods. This could also make it easier to use pedals and/or XboX controllers where you get only a fraction of 100% speed when only pushing the button/pedal down slightly.

Speedcontroll is practically not feasible. I think the most practical solution here would be to build a curve into analog throttle for ground vehicles. In 0 to 0.2 range (as first guess) it should be much less sensitive while beeing linear from 0.2 to 1.0

If i want to drive very slowly with my joystick with integrated analog throttle, i have to increase the throttle from zero by less than a millimeter. With a more agressive exponential behaviour i could move the throttle much further forward.

Now the core problem i would attribute to the issue that the engine does not provide brake torque whenever the wheels turn faster than the engine would allow to. This means that no matter how the engine rpm is, it will always provide torque for acceleration. I also have some doubts about the BIS implementation of whatever "dampingrateInAir" parameter is supposed to be in configs

[Beagle 684]

23 hours ago, scotg said:

To summarize: Most tanks commonly use brakes for steering. A3 steering control animation should only be activated when the brake for the corresponding track is applied, however it is calculated.

That's a method of the past. Modern tank drive trains can shift power independly with automatic transmission. The high mobility of current tanks is possible by the use of independent power distribution without the need to brake one side, The effect is a much smoother steerign and the sticks are obsolete....you use a steering wheel and brake + acceleratiion pedals.

[scotg 204]

1 hour ago, Beagle said:

That's a method of the past. Modern tank drive trains can shift power independly with automatic transmission. The high mobility of current tanks is possible by the use of independent power distribution without the need to brake one side, The effect is a much smoother steerign and the sticks are obsolete....you use a steering wheel and brake + acceleratiion pedals.

Ok, but the steering wheel animation is already working, and my mods are not confined to modern tanks. "Older" tanks that use sticks need some BI love, too. You might be stuck on 1995-2035 tanks, but some of us mod developers want to have as much detail in our classic tanks as the modern ones get. In fact, sometimes it requires more detail, and BI have made it clear they aren't forgetting mods. When doing an era mod the word "obsolete" has a different standard, and sticks are reasonable.

[scotg 204]

3 hours ago, x3kj said:

Speedcontroll is practically not feasible. I think the most practical solution here would be to build a curve into analog throttle for ground vehicles. In 0 to 0.2 range (as first guess) it should be much less sensitive while beeing linear from 0.2 to 1.0

If i want to drive very slowly with my joystick with integrated analog throttle, i have to increase the throttle from zero by less than a millimeter. With a more agressive exponential behaviour i could move the throttle much further forward.

Now the core problem i would attribute to the issue that the engine does not provide brake torque whenever the wheels turn faster than the engine would allow to. This means that no matter how the engine rpm is, it will always provide torque for acceleration. I also have some doubts about the BIS implementation of whatever "dampingrateInAir" parameter is supposed to be in configs

If only they could devote this time to creating some sort of ArmA 3 Tool for tank movement. Feed the tool a set of data, via sliders, that your tank needs to perform, and it creates the proper config values in real time*:
Max Climb Slope:  |-----------------------<>------------------|
Engine Power:        |----------------------------<>-------------|
Torque:                    |-----------------------<>------------------|
Track/Wheel Grip:  |-------------------<>----------------------|
Top Speed:              |-----------------------<>------------------|

Turn Speed:             |-------------<>----------------------------|
Mass:                       |-------------------<>----------------------|
other data:               |-----------------------<>------------------|
other data:               |-----------------------<>------------------|
other data:               |-----------------------<>------------------|
other data:               |-----------------------<>------------------|

Maybe the interface has number values on each side of the slider, and a number before it representing where the slider is, much like the RVMAT tool. The range of each slider would depend on the other sliders' values, and change in real time* accordingly. I.E.: You wouldn't have a bunch of extremes like super high speed and no torque, or whatever.

Below the lay-person section above, another section with actual relative config values will also have sliders that change in real time*, for specific or fine adjustment. Wheel values would be distributed evenly, but drop downs (the number depending on a given wheel number slider input) would enable developers to change individual wheel values. True/False values could be given a checkbox instead of sliders. Configs that exceed their limitations somehow can be marked with a "!" or somesuch.

Then a third section would generate a list of possible side effects based on your settings, such as getting hung up on nothing for a few seconds and then suddenly spinning up into an invisible tornado. Others could be track/wheel spin, jerkiness...

When you get it all how you want it, you can have it create config lines to paste into your actual config.

* real time within the tool. Not to be confused with real time in-game, in case anyone was wondering.

I can dream, right?

[x3kj 1247]

Quote

Feed the tool a set of data, via sliders, that your tank needs to perform

If you want better simulation fidelity you also need to put in more time and knowledge when configuring your vehicle. If a simulation was to simulate every aspect of RL cars, you would have to be a team of realworld car engineers to be able to set up the car's simulation to achieve whatever performance you desired...To know what config results are necessary to achieve certain performance set by some slider or condition with so many variables means you already have to know the solution before you entered what you where even looking for...  So the answer is 42.

Same parallel as with AI - if the developer is a poor tactician, his AI will also be poor in tactical decisions (unless it relies evolution via trial&error).

[AegisWolf 16]

2 hours ago, x3kj said:

If you want better simulation fidelity you also need to put in more time and knowledge when configuring your vehicle. If a simulation was to simulate every aspect of RL cars, you would have to be a team of realworld car engineers to be able to set up the car's simulation to achieve whatever performance you desired...To know what config results are necessary to achieve certain performance set by some slider or condition with so many variables means you already have to know the solution before you entered what you where even looking for...  So the answer is 42.

Same parallel as with AI - if the developer is a poor tactician, his AI will also be poor in tactical decisions (unless it relies evolution via trial&error).

I'm not sure how I feel about introducing a learning AI to Arma. xD

[scotg 204]

7 hours ago, x3kj said:

If you want better simulation fidelity you also need to put in more time and knowledge when configuring your vehicle. If a simulation was to simulate every aspect of RL cars, you would have to be a team of realworld car engineers to be able to set up the car's simulation to achieve whatever performance you desired...

Nah, I was talking in simple terms. The configs are not straight forward and pertain to values that seem abstract to the every day guy. I was just saying that this hypothetical tool could translate more common values into values the config file understands. Slide up for higher speed, and the MOI and springdamperrate adjust accordingly but also depending on other values. N3ed a better climbing rate? Slide that up and the tool adjusts lateral stiffness, engine power, torque... whatever else. It doesn't have to be rocket science; it just has to help prevent the user from applying unstable value combinations for basic movement functions.  

[fn_Quiksilver 1636]

6 hours ago, AegisWolf said:

I'm not sure how I feel about introducing a learning AI to Arma. xD

 

not really feasible to have machine learning in a sandbox game. machine learning is better with fixed variables like chess, pac-man, etc. and of course $$$ can help, billion dollar companies can do practical machine learning but we shouldnt expect that sort of thing from a small game studio.

 

what we should expect though is basic AI procedures, such as effective convoy driving, AI infantry being able to use buildings effectively when in settlement areas, AI soldiers to walk along the side of the road instead of the middle ... stuff like this. but this is not the thread for that discussion :)

[bumgie 49]

Hey!

 

 

Can the tack recoil effect be added to cars or even static weapons? Can it be done in config for any vehicle or static weapon? Currently some lighter vehicle weapons and static weapons act like laser weapons due to not having any sort of recoil.

 

 

[Strike_NOR 898]

1 hour ago, bumgie said:

Can the tack recoil effect be added to cars or even static weapons?

 

Yes, please! Many vehicles with tall mounted HMG's should experience some tilting due to the recoil. This is where vehicles with recoil compensation or more stable platform (heavier vehicle) have serious benefits over the more mobile "technicals".

[xxgetbuck123 945]

15 minutes ago, Strike_NOR said:

 

Yes, please! Many vehicles with tall mounted HMG's should experience some tilting due to the recoil. This is where vehicles with recoil compensation or more stable platform (heavier vehicle) have serious benefits over the more mobile "technicals".

 

Cant wait for the Toyotas with Dshkas on the back to throw their rounds all over the place 

[johnnygitarr 7]

possible solution for speedcontrol:

 

what if we had a kind of a cruise control for the speed.

 

with "shift + w" you can set the max.cruise speed up you wanna drive with

(maybe in 5 or 10 km/h steps, to the max. speed the vehicle can reach)

with "shift + s" the opposite way.

 

then with "w" you accelerate only to this max speed and not further.

"s" is for slow down, as normal.

 

"shift + 2x w" you deactivate the cruise control and the vehicle reacts as we know now.

 

... so we can drive easily in formations with the same speed (only difference is

the different acceleration time/engine power the vehicles have)  

[dragon01 902]

I found a bug with the improved recoil. On Kuma (didn't check the other tanks yet) it doesn't work with HE-T rounds. APFSDS are fine, but HE-T recoil doesn't affect the suspension.

[x3kj 1247]

10 hours ago, johnnygitarr said:

possible solution for speedcontrol:

what if we had a kind of a cruise control for the speed.

Solution for wrongly perceived problem A: what if we had a solution for A.

Ingenious...

 

Speedcontroll does not only result in arkward handling, but it will be even more error prone. Speedcontroll is not the problem and not the solution. Try driving a car in the city exclusively with cruise controll and you will know why... or better not, for the sake of everyone else.

[Strike_NOR 898]

1 hour ago, dragon01 said:

it doesn't work with HE-T rounds

 

Hmmm. This could actually be a feature. If there is absolutely no recoil then I think it's a bug. But HE ammunition generally has lower muzzle velocity because it's more sensitive to the heavy acceleration in the barrel. Less acceleration = less recoil. (not factoring in weight here.)

[dragon01 902]

There's completely no recoil. Also, while HE ammo has lower muzzle velocity, it's heavier than APFSDS. Recoil should probably be somewhat lower, but not much. TBH, I'd expect them to be similar, since the point of a sabot round is to provide a small projectile that moves very fast. Both rounds use similar propellant charges, but for HE rounds the charge needs to accelerate a much higher mass, leading to lower acceleration and lower muzzle velocity.

[GordonWeedman 39]

HE does use a much lower charge actually, simply due to its explosive nature. And yes they can be much heavier, but that isn't the sole contributing factor. If it were, I'd like to know just how bloody heavy a modern British HESH shell is for it to go almost 1000 m/s slower than a just as modern APFSDS shell.

[Strike_NOR 898]

23 minutes ago, dragon01 said:

There's completely no recoil.

Sounds fishy....

[danil-ch 165]

2 hours ago, dragon01 said:

On Kuma (didn't check the other tanks yet) it doesn't work with HE-T rounds. APFSDS are fine, but HE-T recoil doesn't affect the suspension.

Same for Slammer

 

[Strike_NOR 898]

1 hour ago, danil-ch said:

Same for Slammer

 

Actually.... there is visible recoil for HE.  Look closely, it's there. Just not as much recoil as with AP.

 

Maybe a dev can comment if it is in any way influenced by the shells kinetic energy or muzzle velocity.