1. Do not share user accounts! Any account that is shared by another person will be blocked and closed. This means: we will close not only the account that is shared, but also the main account of the user who uses another person's account. We have the ability to detect account sharing, so please do not try to cheat the system. This action will take place on 04/18/2023. Read all forum rules.
    Dismiss Notice
  2. For downloading SimTools plugins you need a Download Package. Get it with virtual coins that you receive for forum activity or Buy Download Package - We have a zero Spam tolerance so read our forum rules first.

    Buy Now a Download Plan!
  3. Do not try to cheat our system and do not post an unnecessary amount of useless posts only to earn credits here. We have a zero spam tolerance policy and this will cause a ban of your user account. Otherwise we wish you a pleasant stay here! Read the forum rules
  4. We have a few rules which you need to read and accept before posting anything here! Following these rules will keep the forum clean and your stay pleasant. Do not follow these rules can lead to permanent exclusion from this website: Read the forum rules.
    Are you a company? Read our company rules

FlyPT Mover

Discussion in 'FlyPt Mover' started by pmvcda, May 30, 2019.

  1. alex928gt

    alex928gt Member

    Joined:
    May 2, 2020
    Messages:
    39
    Balance:
    379Coins
    Ratings:
    +3 / 0 / -0
    The issue is that it's not on one side or the other (i.e. positive acceleration vs negative acceleration). I want to differentiate between acceleration going up in absolute value and acceleration going down in absolute value.

    To continue the analogy I used above :

    - If I get strongly kicked in the back by someone I actually feel it. My acceleration quickly goes from 0 to X. To replicate this my actuator should kick me forward.
    - Right after the kick the acceleration goes back from X to 0. I however don't feel a kick backwards to reflect that. I'm now just travelling forwards at constant speed. So I should not get any kick from the actuator at this point in time.
    - If after been kicked I hit a wall I will feel the shock as a kick backwards. Instantaneously I feel an increase in my acceleration from 0 (travelling at constant speed) to -X as I hit the wall. To replicate this my actuator should kick me backward. It's the same as braking strongly.
    - Right after the shock my acceleration goes back up from -X to 0 but I don't feel a kick forward.

    I therefore believe the horizontal surge actuator should only kick me when acceleration goes away from 0 but not when it's getting back to 0. So the move should not be the same when acceleration goes from X to 0 and when the acceleration goes from 0 to -X.

    Still, in the way it works in Mover using standard filters I did not did find yet a way of making a difference in actuator move between acceleration going from 0 to -X or from X to 0. Both moves get managed the same way.

    Hope this makes sense
    Alex
  2. alex928gt

    alex928gt Member

    Joined:
    May 2, 2020
    Messages:
    39
    Balance:
    379Coins
    Ratings:
    +3 / 0 / -0
    Thanks I'll have a look at the crop function to see if it could help.
  3. Dirty

    Dirty Well-Known Member Gold Contributor

    Joined:
    Oct 15, 2017
    Messages:
    736
    Occupation:
    All the way up front.
    Location:
    Germany
    Balance:
    7,826Coins
    Ratings:
    +859 / 2 / -0
    ehhh,.... I don't know! I have never thought about such a filter.

    I can't really think of any use case right now, but now that I am aware there even is such a filter, I guess it's only a matter of time. Sooner or later someone will run into a problem that this filter will solve, I'm sure.
  4. Dirty

    Dirty Well-Known Member Gold Contributor

    Joined:
    Oct 15, 2017
    Messages:
    736
    Occupation:
    All the way up front.
    Location:
    Germany
    Balance:
    7,826Coins
    Ratings:
    +859 / 2 / -0
    I think what you're saying makes sense! However, Implementing this behaviour without using the brake or gas pedal signal appears challenging, to say the least.

    Let's assume for a minute that I am the motion cueing engine and all I see is a number that represents the longitudinal acceleration of the car. How would I be able to make the distinction between those two cases:
    • Driver just lifted the gas pedal
    • Driver just used the brakes gently.

    I can assure you that when you approach a turn at 290Kph and you just close the throttle, even with an anti-hopping clutch (a motorcycle thing), you will decelerate harder than many street drivers have ever in their whole life. The motion engine only sees the deceleration. It can not infer where that is coming from.

    I can understand that you want to cue only the braking to the driver and not the deceleration due drag, but I wouldn't know how to do that without the brake pedal signal.

    Maybe a solution is just one clever idea away,... if you find it, I'd be truly interested in learning.
    Last edited: Jan 24, 2021
  5. Dirty

    Dirty Well-Known Member Gold Contributor

    Joined:
    Oct 15, 2017
    Messages:
    736
    Occupation:
    All the way up front.
    Location:
    Germany
    Balance:
    7,826Coins
    Ratings:
    +859 / 2 / -0
    Something else came to mind:
    If you could make that distinction, then I guess it would also remove the gear shift jolt, wouldn't it?
  6. hexpod

    hexpod http://heXpod.xyz

    Joined:
    Apr 18, 2016
    Messages:
    1,093
    Location:
    berlin
    Balance:
    7,091Coins
    Ratings:
    +336 / 5 / -0
    My Motion Simulator:
    DC motor, 6DOF
    Not sure we are not misconcepting here

  7. hexpod

    hexpod http://heXpod.xyz

    Joined:
    Apr 18, 2016
    Messages:
    1,093
    Location:
    berlin
    Balance:
    7,091Coins
    Ratings:
    +336 / 5 / -0
    My Motion Simulator:
    DC motor, 6DOF
    In collisions, in real life, you should experience the accels in both directions

    right or wrong?
  8. hexpod

    hexpod http://heXpod.xyz

    Joined:
    Apr 18, 2016
    Messages:
    1,093
    Location:
    berlin
    Balance:
    7,091Coins
    Ratings:
    +336 / 5 / -0
    My Motion Simulator:
    DC motor, 6DOF
    Is the adverse cue on longitudinal axis, (accelerometer returning to 0) a false cue which you want to cancel ?
  9. alex928gt

    alex928gt Member

    Joined:
    May 2, 2020
    Messages:
    39
    Balance:
    379Coins
    Ratings:
    +3 / 0 / -0
    In fact what you mention exactly reverts to what I said earlier but it should not be confused:

    "I could maybe encounter a slight backward kick when releasing the throttle but this should not come from releasing the throttle itself but more from actual engine braking and road/air friction generating an actual braking i.e. an expected backward kick."

    In fact my feeling is that what creates the effect you described earlier when approaching a turn at 290kph is not related to closing the throttle (i.e. having the acceleration getting back to 0) but to the fact that frictions including air frictions, road frictions but mostly internal engine frictions/compressions aka "engine braking" create an actual braking i.e. acceleration actually getting strongly into negative territory. Braking is not getting the acceleration from positive to 0 it's actually getting acceleration into negative values which should not be confused in my view.

    For instance if you could imagine not really closing the throttle but just going to neutral instead (although probably not what you actually want to do at this point in time...) then you would not feel at all the same effect.

    So the effect you described is not actually related to having acceleration getting back to 0 but to other factors that depend on other conditions and should therefore be captured separately. If based on those conditions you are just able to get acceleration back from positive to 0 you should feel no kick. But if together with this effect you actually get into negative acceleration territory then you should actually get a backward kick. In any case it all just depends on the actual behavior of acceleration i.e. just getting back to 0 -> no specific effect, or getting negative -> kick effect.

    Another example: imagine you brake at constant negative acceleration until the car is actually stopped. This means acceleration is negative until the cars stops and the acceleration then gets back to 0 since the speed is now constant and equal to 0.

    I this case you have:
    - constant negative acceleration at -X, decreasing speed
    - acceleration back to 0, speed at 0

    In the current standard modeling of the actuator when the acceleration gets from -X to 0 i.e. when the car is stopped you would get a strong forward kick. But how come, the car is now stopped, why should you feel this forward kick? The only reason why you could to feel a kick at this point in time could come from the compression of shock springs at braking time but 1/ the kick should probably come the other way around i.e. backwards and 2/ spring compression is not the point I'm discussing here anyway since this would be modeled by vertical actuators, not horizontal ones.

    A final example: imagine you are the bullet in a cannon. At ignition time you get a strong acceleration so acceleration starts at 0 before ignition and suddenly goes to X>0. You clearly expect in this case to get a strong forward kick from the actuator. But right afterwards acceleration gets back from X to 0, you are now just travelling into the air at almost constant speed. Should you feel a kick back because your acceleration goes from X to 0? Clearly no, you should only feel the first forward kick and then just nothing with the actuator just washing out nicely towards 0. Until you actually hit a wall or the ground of course, but this is another story...


    At the end of the day I still believe that what I'm describing should be explored for car sims and we actually have the information available since we always know it the acceleration is currently positive or not and if it is going away from 0 or not.

    The question is whether this can actually be managed through standard filters or instead through hard coding as Yobuddy probably did for TL :

    "It's a washout filter that ignores input from the data returning to zero.
    Or you could say, it captures the acceleration data only when it growing either in a positive or negative direction."


    Thx
    Alex

    • Agree Agree x 1
  10. alex928gt

    alex928gt Member

    Joined:
    May 2, 2020
    Messages:
    39
    Balance:
    379Coins
    Ratings:
    +3 / 0 / -0
    @hexpod, I believe the point you are making is right. You have accel going from 0 to X and then back from X to 0 but the accelerometer only received one collision, not two.

    Here in my case it's exactly the same just the other way round (i.e. actuator kick being generated instead of collision being received and acceleration figure being an input source for the actuator instead of accelerometer getting the collision information).

    Accelerometer input says acceleration going from 0 to X and then back from X to 0 but only one shock/kick sent to the actuator when accel goes to X.
  11. hexpod

    hexpod http://heXpod.xyz

    Joined:
    Apr 18, 2016
    Messages:
    1,093
    Location:
    berlin
    Balance:
    7,091Coins
    Ratings:
    +336 / 5 / -0
    My Motion Simulator:
    DC motor, 6DOF
    The question one has to ask is whether the reverse motion of the platform on an adverse accelerometer cue is physically correct and natural or wrong, If that’s wrong, one has to find a way to cancel the false cue.
  12. alex928gt

    alex928gt Member

    Joined:
    May 2, 2020
    Messages:
    39
    Balance:
    379Coins
    Ratings:
    +3 / 0 / -0
    For me it seems very clear from:

    - my example with the cannon: 1 accel change from 0 to X (ignition) + 1 accel change back from X to 0 (bullet back to constant speed) but I instinctively expect only one kick forward at the start
    - confirmed by the youtube video you shared that shows 1 accel change from 0 to X + 1 accel change back from X to 0 for only one collision received

    So clearly the accel change back to 0 is a false cue that should be removed when triggering the kicks from a horizontal surge actuator.

    However both cues should be kept when simulating surge from vertical actuators/tilt of the platform. During acceleration the first cue (accel increase) tilts the platform backwards and if you release the throttle the platform should tilt back forwards simulating the de-compression of the rear shock springs that got compressed during the acceleration phase. So in this case both cues are actually useful. But that's another story and pretty well managed anyway by all sim softwares such as Simtools.

    Alex
  13. hexpod

    hexpod http://heXpod.xyz

    Joined:
    Apr 18, 2016
    Messages:
    1,093
    Location:
    berlin
    Balance:
    7,091Coins
    Ratings:
    +336 / 5 / -0
    My Motion Simulator:
    DC motor, 6DOF
    Well I would be careful about what “our instincts expect”

    Edit :
    ̷C̷o̷u̷l̷d̷ ̷y̷o̷u̷ ̷p̷r̷o̷v̷i̷d̷e̷ ̷a̷ ̷v̷a̷l̷i̷d̷ ̷r̷e̷a̷s̷o̷n̷ ̷w̷h̷y̷ ̷o̷n̷ ̷l̷o̷n̷g̷i̷t̷u̷d̷i̷n̷a̷l̷ ̷a̷x̷i̷s̷ ̷i̷t̷ ̷s̷h̷o̷u̷l̷d̷ ̷b̷e̷ ̷d̷i̷f̷f̷e̷r̷e̷n̷t̷ ̷t̷h̷a̷n̷ ̷o̷n̷ ̷v̷e̷r̷t̷i̷c̷a̷l̷ ̷o̷n̷e̷ ̷?̷

    What would it give, if instead of acceleration, you would use velocity?


    Last edited: Jan 23, 2021
  14. Sverenja

    Sverenja New Member

    Joined:
    Jan 12, 2021
    Messages:
    12
    Location:
    Bavaria
    Balance:
    630Coins
    Ratings:
    +12 / 1 / -0


    I think @alex928gt means a vertically mounted actuator on his rig (tilt/pitch), not a vertical axis.
    And in my eyes his assumptions on what should happen with the actuators makes sense to me, but will be kinda hard to implement considering the physical limitations of a motion rig, which cannot just remain in a changed horizontal position, but has to washout back to the home position.

    On breaking the horizontally mounted actuator kicks the seat backwards simulating the initial jolt of acceleration change. This initial motion is the same in the simulator and in the real world from the relative perspective of the driver. The continued negative acceleration can not be simulated with a horizontally acting actuator of course, or one would simply be sitting in a car with a HMD on their face.

    At the same time the vertically mounted actuators (not axis) tilt (pitch) the platform forward, simulating the direction of acceleration on the driver using gravity and at the same time the compression of the springs of the car. The difference to the purely horizontal movement of the platform this action feels different as the perceived acceleration kinda continues to match the simulated one as does the angle of the platform.

    That's why @alex928gt mentioned that once the negative acceleration ends the jolt of the platform pitch going back to zero is natural, because real world cars do the same thing (decompression of the springs), but at this point the horizontally acting actuator should not kick, but gently go back to home imperceptively (washout). In an ideal world the platform should already be back at the home position as it is not known whether the next event is positive or negative accleration.

    Which means that there should be different actions of the actuators as @Dirty already said and tries to realize using filters.

    The same is true for positive acceleration. In an ideal world (spherical cow in a vacuum) the end of an acceleration does nothing in a purely horizontal axis, but a pitch event simulating the decompression of the springs, but as @Dirty and @alex928gt said in this case a small jolt in horizontal direction is acceptable, because unfortunatley the car is not round nor floating in a vaccuum, so as soon as we release the throttle the car is actually breaking by friction (motor, tires, air, etc.).
    That is the small error in the video above as the little wooden car does not really roll, but due to friction breaks immdiately, exaggerating the negeative acceleration, whereas a real car is kinda optimized for rolling and has a rather favourable amount of friction compared to the weight of the whole thing.

    Physically there is no difference between negative (breaking) and positive acceleration, both result in a change of speed and the end of an acceleration does not result in an additional acceleration in the other direction. Exerted forces are back to zero.



    So after all this drivel that was just repeating what was already said by others

    In my opinion the solution could be to tie the horizontally acting actuator into a pedal position change (joystick mode ?) for normal positive or negative acceleration, but only in the direction of pressing the pedal (releasing a pedal should not really do anything) and mix in the output of the game plugin so heavily filtered that only crash events result in actuator action.
    The pitch events in contrast would only come from the game as their behaviour already matches expected movements (continued accelearation and pitching of a real-world-not-idealized car).

    The problem is what @alex928gt already mentioned in the initial post as [0 -> -x] input should be different from [-x -> 0], but now it would be from a different input, which would not influence the pitch/tilt actuators at the same time.


    A crash event would still not play out correctly as the acceleration spike would still result in two accelerations of the platform, but at least in my opinion this edge-case would be acceptable.
    • Like Like x 1
  15. hexpod

    hexpod http://heXpod.xyz

    Joined:
    Apr 18, 2016
    Messages:
    1,093
    Location:
    berlin
    Balance:
    7,091Coins
    Ratings:
    +336 / 5 / -0
    My Motion Simulator:
    DC motor, 6DOF
    Sure I understand the argumentation which seems logical but I can’t simply imagine how to translate it into physics.

    Physically there is no difference between negative (breaking) and positive acceleration, both result in a change of speed and the end of an acceleration does not result in an additional acceleration in the other direction. Exerted forces are back to zero.

    I have to disagree. If the acceleration is abrupt, it’s perfectly normal to get a “back jolt” because it’s decelerating. The change of speed acts both ways as it stops accelerating.
  16. Sverenja

    Sverenja New Member

    Joined:
    Jan 12, 2021
    Messages:
    12
    Location:
    Bavaria
    Balance:
    630Coins
    Ratings:
    +12 / 1 / -0
    Yes, this is true for a car, as the friction immediately acts as negative acceleration as soon as the force of the engine is removed. I meant that as an idealized concept that's why I added the bit about the spherical cow in vacuum.
    That's why the focus is on breaking (negative acceleration) as this is not followed by a positive acceleration. The car doesn't speed up on release of the pedal. The only positive acceleration on the human body at the end of a breaking event is due to the elasticity of the belts and the body itself "unsquishing".
    • Like Like x 1
  17. hexpod

    hexpod http://heXpod.xyz

    Joined:
    Apr 18, 2016
    Messages:
    1,093
    Location:
    berlin
    Balance:
    7,091Coins
    Ratings:
    +336 / 5 / -0
    My Motion Simulator:
    DC motor, 6DOF
    Oh, maybe in this case we are talking about the washout speed issue ? You can easily adjust the speed of HPF in mover
  18. alex928gt

    alex928gt Member

    Joined:
    May 2, 2020
    Messages:
    39
    Balance:
    379Coins
    Ratings:
    +3 / 0 / -0
    All this maybe becomes clearer if instead of talking about accelerations going from 0 to X and back from X to 0 we just talk about forces.

    When you jump on the throttle, you apply a forward force to the car and the driver. This is equivalent, through the 2nd Newton law F=m.a to applying an acceleration to the car. So a force appears and the acceleration goes from 0 to a. Then when you stop throttling - if there is no other force applied - the force gets back to 0 and still through the 2nd law, the acceleration gets back to 0.

    So instead of talking about acceleration going from 0 to X and then back from X to 0 we should just say that a force is applied and then cut.

    Likewise when braking, a backward force is applied and then cut when releasing the brake.

    And logically what I would expect is that when the throttle force is applied I get a kick forward from the horizontal surge actuator and when the braking force is applied I get a kick backward. But there is not reason to get a kick when either force is cut since a kick needs and represents a force, not the absence thereof.

    Now I'm still back to my original question, i.e. how to implement that simply based on the evolution over time of the acceleration input.
  19. hexpod

    hexpod http://heXpod.xyz

    Joined:
    Apr 18, 2016
    Messages:
    1,093
    Location:
    berlin
    Balance:
    7,091Coins
    Ratings:
    +336 / 5 / -0
    My Motion Simulator:
    DC motor, 6DOF
    But there is not reason to get a kick when either force is cut”

    Nonsense. Nevertheless I respect because after all, everyone does what they want.

    Best.
  20. alex928gt

    alex928gt Member

    Joined:
    May 2, 2020
    Messages:
    39
    Balance:
    379Coins
    Ratings:
    +3 / 0 / -0
    Thanks but could you please elaborate on why is this is a nonsense? Happy to get examples of situations where a force being cut should generate a kick and to discuss.

    This force should of course not be offset by another one because otherwise cutting it is just a way of "activating" the other one.

    Differently said, in my view, a kick requires a force being applied. Otherwise where would the kick come from?