Jump to content
Bohemia Interactive Forums

x3kj

Member
 View Profile  See their activity

  • Content Count

    2605

  • Joined

  • Last visited

  • Medals

  • Medals

Everything posted by x3kj

  1. x3kj
    apply a standard constant rotation. Then add an rpm or thrust based rotation on top of that.
  2. x3kj
    You say like it is easy or minor thing, but i can assure you that it isn't. I have looked at the physx sourcecode. There is a lot of optimisation for tracked vehicles to reduce performance impact of 20 simulated wheels. Doesnt matter if they "just bounce and turn". They still require raytracing and simulation. @reyhard No i haven't yet. It's good to know that it is adjustable - however, i'f i'm correct in my conclusion then it's pretty inconvenient if the "base" for application is the center of bounding box. As soon as you add something somewhere that affects the bounding box you have to re-trial&error the offset. Regarding the BMP jerking on acceleration - the reason IRL why it is jerking is because the sprocket is at the back. Whenever it accelerates, the sprocket pulls the track upwards. Since the track and the vehicle have inertia that needs to be overcome, the same force that is applied to the track is also applied to the sprocket in opposite direction. Means that the rear end is pulled downwards, instead of the front beeing lifted upwards.
  3. x3kj
    Here is the first issue i have (other is still uploading) Note the bounding boxes of both vehicles (as result of having some memory points far below for testing purposes). The small vehicle flips over immediately. The larger lifts up only a bit at times, so the force acts only a short time. But sometimes it goes "all the way" and raises the entire vehicle. This happens reliably when accelerating from stand uphill, but also frequently when driving reverse first and then forward. I conclude that the new force is 1) applied inconsistently - indicated that the large tank does not always flip up, only sometimes 2) applied to center of bounding box, instead of center of gravity- otherwise the tanks would not flip. I would suggest applying the force at a defined point (either memorypoint or coordinate in config), instead of center of bounding box or gravity. It's helpfull to tweak this in case application on center of gravity leads to issues - because otherwise you would have to shift CoG in order to fix some fake force -> leads to a rat tail of other problems. In case there already is a point definition in config then these vehicles would of course miss it - so it might as well be a non issue. its there since quite a while in DEV... They took the standard physx drive model and implemented it. If half-tracks or different track tanks are wanted then they require a nonstandard drive simulation. Considering that there have been several issues with tank simulation already i rather prefer refinement of what's already there over having another semi-working simulation class.
  4. x3kj
    A tank is not sluggish at fast speeds. You have extremely short stopping distance due to the high traction forces (compared to wheeled vehicles). What works for both tracks also works for a single track. The only thing that is a concern for not doing it constantly is safety... If you block one side at full speed, you have a high chance of throwing a track or maybe even flipping over if your track catches something while drifting (reference). The suspension of tanks is not magical - it won't keep the hull perfectly still. In rough terrain it will bounce significantly IRL. That's why stabilizers for guns and sights are needed in the first place. Realistically it should bounce even more off road, but given that Arma terrain is basically "flat" all across (no micro variations in terrain heigth) it's better to keep it that way i guess. Visual wheel animation is not automatically synched to actual simulation - so yeah that's just something they have to re-tweak. Even though tank tracks can create high traction forces, it's still not infinite. Tanks have very high kinetic energy at higher speeds - so changing direction sharply will lead to drifting even with high traction. Example. In games people generally steer extremely sharply - IRL you would never do that (e.g. driving a cargo truck at 70kph 90° around a T-junction ). Maingunner can be primaryGunner and primaryObserver(=commander) at the same time from configuration point. So yeah, that should be doable.
  5. x3kj
    Negative, can't confirm. Check if your controlls for vehicle driver commanding are bound the same as moving turrets/view.
  6. x3kj
    tomorrow i'll try and make a video of the technical issues that i see with the changed physics on my custom vehicles. Another observation: when using the Command-view (comma key on numpad by default when you are squad leader) the sound is attenuated as if you would sit inside. Previously it was always the exterior sound. The RPM goes down in turns in the simulation too. Depending on the gear selected when entering the turn. Not sure we'll be able to solve it. We can't make the vehicle downshift for turns only, on the other hand the sound can go a bit around the simulation. In neutral turns this should be better. The engine is immediately cut off therefore quickly goes back to idle RPM whenever thrustR or thrustL gets negative or 0. So if you turn >= 50% on your controller (or always if you use keyboard) it will be this case. The question is why the engine cutoff in the first place? Also, in PhysX there is an engine moment of inertia parameter (not exposed in arma configs) Increasing it would slow down the deceleration of the engine possibly. Unless the "thrust off" is hardcoded to fore idle rpm and unless the engine can not free wheel: At some point in the past, after one of the physx updates the engine would no longer "free wheel". Since then it was always connected to the drivetrain. Before that it would disconnect whenever you let go of throttle fully. The wheels would turn at their own speed, while engine returned to idle rpm. At current, engine is connected even if no throttle is given. Not sure if that was concience decision of BIS or just a forced side effect of a physics update, or simply a change of physics default behaviour that wasn't adjusted by BIS?
  7. x3kj
    Re: track animation tested Kuma today with latest update - eventually the Track animation to glitched out and stopped moving entirely. I had the impression that it was yet another "jump turning glitch" but it was less extreme than usual. After that i noticed that tracks stopped turning. After respawn it worked all normal again. Yesterday the test with my custom tank, the tracks behaved very weird after a "jump turning glitch" (one pirouette) as well. On my custom tank the tracks are still turning pretty erratically (slow quick slow quick) - for some reason the average wheel speed (epe dialog) is fluctuating a lot while turning/accelerating and also for a while after. When traveling straight for prolonged period it stabilizes.
  8. x3kj
    After driving the Kuma for a couple hundred meters i understand why listen is put in bold... The engine is revving every 5 seconds when driving uphill at full throttle. Seems like the condition for detecting sudden throttle increase does not work properly? Tracks clucking - the clucking ingame at 40km/h fits driving speeds of maybe 5-10km/h at best. I tried to quickly find data on Leo2 but my search foo failed me -> resorting to alternative: Leopard 1 has track segment length of 16 cm Not sure if identical to Leo 2. So 40km/h travel speed on Leo 1 causes a vibration (and therefore noise) with a base frequency of (40/3.6)*0.16 = ~ 70Hz. There is no clacking at this speed anymore. It's already a "humming". I've already mentioned that in sound feedback and tank DLC feedback: Another one i found is here, watch first 18 seconds (rest clips horribly). Second 3 to 18 is all track noise - no engine. Here are a couple of decend ones - the first approach is the clearest. Handling itself and the engine sound samples are a solid step forward. I would love to hear the turbocharger and fan more, especially when you release the throttle. Although this video is terrible quality wise you can pick out both fan and turbo noticeably - esp. when throttle is cut back. Another example. Here's a good sound from another engine, but same principal: Engine + Turboloader, Engine + CoolingFan + Turboloader that i already posted in my big tank DLC feedback post. The "squeak" when stopping the tank that i assume you got inspired by this video at ~32sec is not a metal-on-metal squeak. It's as result of fast hydraulic damper recoil. Currently it sounds like metal squeak ingame.
  9. x3kj
    Unless you contribute actively to the project you have to wait like everybody else.
  10. x3kj
    I hope whatever fake forces are applied to tanks by default now on acceleration can be disabled... my tanks occasionally lift their front up until they fall over backwards when trying to start moving from stand still. But not always, 66-75% they come up with the front a little bit before moving normally. The tank track animations (the track texture that is moved along the track model) behaves completely eratically at times since recently as well. When driving forward with steady speed, the animation speed is not always steady - it can slow down and get faster at random. In general track speed is too slow compared to before. Right and left side move very differently, even when traveling straight forward - very observable at low speeds (e.g. uphill). When driving slowly sometimes one track stops moving. When driving reverse and turning, the wrong side is animated at times. I've also had it glitch out, where the left track would rotate backwards when driving forward, and also continue rotating a short time after the tank was stopped at much faster speed then it should be. I assume the glitch was related to beeing catapulted again on normal terrain during turning. It is not gone. Track texture inconsistency is less noticeable on vanilla tanks because their track textures are pretty poor and too greebly to notice anything and are sometimes barely visible. It's there nontheless.
  11. x3kj
    w00t, more config options for physx Very much looking forward for the documentation of that.
  12. x3kj

    Nassau 1715

    x3kj replied to Delta Hawk's topic in ARMA 3 - ADDONS & MODS: DISCUSSION

    1st rate work mate
  13. x3kj
    logistical challenge - nobody appeared at work because everyone was smart and took the day off ^^
  14. x3kj

    Targeting improvements

    x3kj replied to oukej's topic in ARMA 3 - DEVELOPMENT BRANCH

    What about manual beam riding guidance? The current manual missile guidance we have aims at a specific point in the landscape, which leads to many wonky situations - e.g. if you point at your feet when the missile is in flight - it loops around and if maneuverable enough hits you - which is ridiculous. Similar situation happens if you track the moving target and an object close to you moves in between you and the target - the missile immediately jerkes hard to adjust for the different vector to the much closer point target. If during aiming at the target you "miss" it with your beam occasionally, the point of target will be much further (possibly infinity) leading to very different missile maneuvers.
  15. x3kj

    Editing SAM System

    x3kj replied to gabberxxl's topic in ARMA 3 - ADDONS - CONFIGS & SCRIPTING

    Precisely, because before you can inherit from a subclass (e.g. Turrets:Turrets) you need to declare it as well in the external class reference.
  16. x3kj

    Some troubles with LODS

    x3kj replied to NoctisLucisCaelum's topic in ARMA 3 - MODELLING - (O2)

    dont you understand that visual LOD can look vastly different than geo LOD and still work good?
  17. x3kj
    I have that impression too. Tank DLC= last DLC. I would expect 2 years minimum after tank DLC until Arma 4... Or 5 years if they decide to do another enfusion based game first instead.
  18. x3kj
    Could we please get a documentation how the bipod deployment algorithm picks out what location is suitable for deployment and what not? What conditions have to be met to get the possibility to deploy? What are the limitations of fire geometry shapes - for example how do angled surfaces influence the deploy check? I'm working on more realistic bunkers than we have ingame and it is very unpredictable how the deployment works with the model. In one model of mine, when standing far away from the firing slit, i get a deploy icon shown - but no weapon resting. When i move closer i get the resting icon, but the deployment is gone. In addition, it seems unreliable for otherwise identical geometry. For example i mirrored a firing slit, but one side behaves different than the other (has no deployment, other side has but only at very specific angles and position)
  19. x3kj
    Maybe they save it for tac ops DLC or whatever? A possibility... Or they where not satisfied with the look, or the model technicalities didn't meet their specs, or LODs where not created by the outsourced company and they don't have any artist free to complete it. Though the bipod is questionable, atm deployment does not work for launcher weapons - so they would have to implement that, which is unlikely to happen i think. I know one time a company payed for a model from an external artist that happened to be me and it was in the game for an expansion... they just totally forgot about it for some reason.
  20. x3kj
    Case in point - wall with firing slits and recesses for weapon deployment Left version you can deploy/weapon rest without problem in the wall recess and shoot out of the firing slit. The right version, where the wall is angled, deploying on the recess is not available most of the time, despite using the same position and space dimensions for the recess as the left version. Only on very specific positions, basically standing 1-2 m away from the recess do you get the option to deploy.
  21. x3kj
    progress on Tank physics, yay although that propably means that i have to re-tune my vehicles again *sigh* From 3.4 changelog This sounds like this was the culprit of of Trucks flying in the air when driving over a tiny wall. If that is also the reason why tanks jump on flat goround, that seems strange to me - because that would mean the hit normal sometimes is in the plane of the ground, instead of beeing normal to it. Thats some serious deviation...
  22. x3kj

    New driver/pilot poses

    x3kj replied to scotg's topic in ARMA 3 - MODELLING - (O2)

    the path is irrelevant, at least for the ingame functionality. Its only the name and the number.
  23. x3kj
    I'm on a different playground, because i can implement stuff right away for my total conversion ;)
  24. x3kj
    i think the whole instigator thing requires a revisit. For example for any damage you script, you can't influence the instigator in any easy way. And as such any script relying on kill eventhandlers to do something with the "killer" will not receive correct input.
  25. x3kj
    Thank you so much. The ranges that you could rocket snipe with jets where insane.
×