My quaternion math is a bit rusty, and I'm trying to figure out which of the following implementations is more correct...
Looking here you can see Microsoft's version of transforming a vector by a quaternion: https://referencesource.microsoft.com/#System.Numerics/System/Numerics/Vector3.cs,347
Here is the code, and it has clearly been optimized:
public static Vector3 Transform(Vector3 value, Quaternion rotation)
{
float x2 = rotation.X + rotation.X;
float y2 = rotation.Y + rotation.Y;
float z2 = rotation.Z + rotation.Z;
float wx2 = rotation.W * x2;
float wy2 = rotation.W * y2;
float wz2 = rotation.W * z2;
float xx2 = rotation.X * x2;
float xy2 = rotation.X * y2;
float xz2 = rotation.X * z2;
float yy2 = rotation.Y * y2;
float yz2 = rotation.Y * z2;
float zz2 = rotation.Z * z2;
return new Vector3(
value.X * (1.0f - yy2 - zz2) + value.Y * (xy2 - wz2) + value.Z * (xz2 + wy2),
value.X * (xy2 + wz2) + value.Y * (1.0f - xx2 - zz2) + value.Z * (yz2 - wx2),
value.X * (xz2 - wy2) + value.Y * (yz2 + wx2) + value.Z * (1.0f - xx2 - yy2));
}
However, this gives different results to my own (less optimized) implementation:
public static Vector3 Transform(this Vector3 value, Quaternion rotation)
{
var q = new Quaternion(value.X, value.Y, value.Z, 0.0f);
var res = rotation.Conjugate() * q * rotation;
return new Vector3(res.X, res.Y, res.Z);
}
public static Quaternion operator *(Quaternion value1, Quaternion value2)
{
// 9 muls, 27 adds
var tmp_00 = (value1.Z - value1.Y) * (value2.Y - value2.Z);
var tmp_01 = (value1.W + value1.X) * (value2.W + value2.X);
var tmp_02 = (value1.W - value1.X) * (value2.Y + value2.Z);
var tmp_03 = (value1.Y + value1.Z) * (value2.W - value2.X);
var tmp_04 = (value1.Z - value1.X) * (value2.X - value2.Y);
var tmp_05 = (value1.Z + value1.X) * (value2.X + value2.Y);
var tmp_06 = (value1.W + value1.Y) * (value2.W - value2.Z);
var tmp_07 = (value1.W - value1.Y) * (value2.W + value2.Z);
var tmp_08 = tmp_05 + tmp_06 + tmp_07;
var tmp_09 = (tmp_04 + tmp_08) * 0.5f;
return new Quaternion(
tmp_01 + tmp_09 - tmp_08,
tmp_02 + tmp_09 - tmp_07,
tmp_03 + tmp_09 - tmp_06,
tmp_00 + tmp_09 - tmp_05
);
}
Since these do not give the same results, one of them must be wrong, but which one is it and why?
My own implementation seems to work in the cases that I'm trying to use it in, and the MS implementation seems to be broken, but I would be surprised if it were actually incorrect, since I think it is very widely used...
Thanks! :)
There are two different conventions for rotating a vector via a quaternion, left chain and right chain. E.g.,
vnew = q * v * conjugate(q) <-- This is left chain (successive rotations stack up on the left)
vnew = conjugate(q) * v * q <-- This is right chain (successive rotations stack up on the right)
Left chain convention is typically used for active rotations, where the quaternion is being used to rotate a vector within a coordinate frame. I.e., v and vnew are two different vectors coordinatized within the same coordinate frame.
Right chain convention is typically used for passive rotations, e.g. quaternions that represent coordinate system transformations. I.e., v and vnew are actually the same vector, just coordinatized in two different coordinate frames.
The MS code you show above is a left chain convention, but your code is a right chain convention. Hence the different results.
Both conventions are valid and have their uses, but you need to be very careful when pulling code from online sources. You need to verify what the convention used by the code is in order to use it correctly. And you need to ensure the convention matches how you are using the quaternions in your particular application.
Related
So, Dani in his slightly new video -> "Making a Game, But I Only Have 3 Days" (https://youtu.be/S7Dl6ATRK2M) made a enemy which has a bow and arrow (at 5:39). I tried to recreate that but had no luck... I also can't find the website that he used... Today I found this https://physics.stackexchange.com/questions/56265/how-to-get-the-angle-needed-for-a-projectile-to-pass-through-a-given-point-for-t. It worked very well but still had problems if the target was far away and also it wasn't as accurate. The code so far is
float CalculateAngle()
{
float gravity = Physics.gravity.magnitude;
float deltaX = targetPositionMod.x - currentPosition.x;
float deltaY = targetPositionMod.y - currentPosition.y;
float RHSFirstPart = (velocity * velocity) / (gravity * deltaX);
float RHSSecondPart = Mathf.Sqrt(
((velocity * velocity) * ((velocity * velocity) - (2 * gravity * deltaY))
/ (gravity * gravity * deltaX * deltaX))
- 1);
float tanθ = RHSFirstPart - RHSSecondPart;
float angle = Mathf.Atan2(tanθ, 1) * Mathf.Rad2Deg;
if (angle < 0) return angle;
return -angle;
}
The -angle is because the forward axis starts points up when the x-rotation is negative (Unity). Maybe the reason of this not working as intended is that I am not that good at this kind of Physics (Part of that is me being only 14). Maybe the problem is in the code, maybe it is the formula. Any help is appreciated.
Thanks...
Edit:
The Archer class is:
using UnityEngine;
using System;
public class Archer : MonoBehaviour
{
[SerializeField] float velocity = default;
[SerializeField] Transform target = default;
[SerializeField] GameObject arrowPrefab = default;
[SerializeField] float coolDown = default;
Vector3 targetPositionMod;
Vector3 currentPosition;
Vector3 targetPosition;
float countDown = 0f;
void Start()
{
countDown = coolDown;
UpdateVariables();
}
void Update()
{
UpdateVariables();
SetAngle();
ShootBullet();
}
void UpdateVariables()
{
currentPosition = transform.position;
targetPositionMod = Mod(target.position);
targetPosition = target.position;
targetPosition.x /= 10;
targetPosition.y /= 10;
targetPosition.z /= 10;
countDown -= Time.deltaTime;
}
void SetAngle()
{
Vector3 direction = targetPosition - currentPosition;
Quaternion lookRotation = Quaternion.LookRotation(direction);
Vector3 rotation = lookRotation.eulerAngles;
rotation.x = (float) CalculateAngle();
transform.rotation = Quaternion.Euler(rotation.x, rotation.y, 0f);
}
void ShootBullet()
{
if (!(countDown <= 0f)) return;
countDown = coolDown;
GameObject arrow = Instantiate(arrowPrefab, transform.position, transform.rotation);
Rigidbody Rigidbody = arrow.GetComponent<Rigidbody>();
Rigidbody.AddForce(transform.forward * velocity, ForceMode.Impulse);
}
double CalculateAngle()
{
double gravity = Physics.gravity.magnitude;
double deltaX = targetPositionMod.x - currentPosition.x;
double deltaY = targetPositionMod.y - currentPosition.y;
double RHSFirstPart = (velocity * velocity) / (gravity * deltaX);
double RHSSecondPart = Math.Sqrt(
(((velocity * velocity) * ((velocity * velocity) - (2 * gravity * deltaY))
/ (gravity * gravity * deltaX * deltaX))
- 1));
double tanθ = RHSFirstPart - RHSSecondPart;
double angle = Math.Atan2(tanθ, 1) * Mathf.Rad2Deg;
if (angle < 0) return angle;
return -angle;
}
Vector3 Mod(Vector3 Vec)
{
if (Vec.x < 0) Vec.x -= 2 * Vec.x;
if (Vec.y < 0) Vec.y -= 2 * Vec.y;
if (Vec.z < 0) Vec.z -= 2 * Vec.z;
Vec.x /= 10;
Vec.y /= 10;
Vec.z /= 10;
return Vec;
}
}
Ok, as I can see, your implementation of formula from StackExchange is right, but you have to remember two things:
In unity there is a 3D world, so horizontal distance is not just pos1.x - pos2.x, but
Mathf.Sqrt( deltaX * deltaX + deltaZ * deltaZ ), where deltaX = targetPositionMod.x - currentPosition.x and deltaZ = targetPositionMod.z - currentPosition.z
In computer implementation you have no 100% accuracy of math, so some problems can appear because of computational accuracy. And it can have affect on big distances. You can try to use double instead of float or find another implementation for arctangent function (I think, this can really help). But try this (second) advice only if first didn't help. It's harder to implement and it slows computations a bit.
Algorithm:
Step 1: Set up a function that calculates the appropriate solution of a quadratic equation
a*x^2 + b*x + c = 0
double quadratic_root(a,b,c){
D = b^2 - 4*a*c
return ( - b - Math.Sqrt(D) ) / (2 * a)
}
Step 2: Input
current.x
current.y
current.z
target.x
target.y
target.z
velocity
gravity
Step 3: Calculate coefficients of the quadratic polynomial:
dist = Math.Sqrt( (target.x - current.x)^2 + (target.y - current.y)^2 )
a = gravity * dist^2 / (2 * velocity^2)
b = -dist
c = target.z - current.z + a
Step 4:
theta = Math.Atan2( quadratic_root(a,b,c), 1 )
Calculation behind the algorithm. You are in three space. The current position has coordinates
x = current.x
y = current.y
z = current.z
and the target has coordinates
x = target.x
y = target.y
z = target.z
Assume the angle between the initial velocity and the horizontal plane is theta. The magnitude of the projection of the distance between the current position and the target onto the horizontal $x,y-$plane is
dist = sqrt( (target.x - current.x)^2 - (target.y - current.y)^2 )
You are given the velocity magnitude velocity. Then, the speed with which the shadow (i.e. the orthogonal projection) of the arrow moves along the horizontal line between the source and the target is the magnitude of the shadow (i.e. the orthogonal projection) of the actual velocity
velocity * cos(theta)
The vertical speed of the arrow is then
velocity * sin(theta)
So the motion along dist follows the equation
dist = time * velocity * cos(theta)
and hence
time = dist / (velocity * cos(theta))
In the vertical direction, the motions is described by the equation
z = current.z + time * velocity * sin(theta) - time^2 * gravity / 2
You are interested in the time for which the arrow hits the target, which has vertical coordinate target.z, so
target.z = current.z + time * velocity * sin(theta) - time^2 * gravity / 2
The equation can be written as:
0 = - (target.z - current.z) + time * velocity * sin(theta) - time^2 * gravity / 2
We already know that
time = dist / (velocity * cos(theta))
so
0 = - (target.z - current.z) + dist * velocity * sin(theta) / (velocity * cos(theta)) - dist^2 * gravity / ( 2 * (velocity * cos(theta))^2 )
which can be slightly simplified to
0 = - (target.z - current.z) + dist * sin(theta) / cos(theta) - gravity * dist^2 / ( 2 * (velocity * cos(theta))^2 )
Because 1/( cos(theta)^2 ) = 1 + ( tan(theta) )^2 we obtain the quadratic in tan(theta) equation
a * ( tan(theta) )^2 + b * tan(theta) + c = 0
where
a = gravity * dist^2 / (2 * velocity^2)
b = - dist
c = target.z - current.z + a
I am trying to figure math for getting all longitudes and latitudes in a specific distance between each other. I have math for getting distance between geographic coords but I would like to get an array of all geo coords from the standard plane earthmap in the specific distance. Can you help me, please?
Here is example link of earthmap picture. I have earth heightmap in resolution 43200x21600 and I would like to get heights/pixels indexed from bottom left to top right in exact distance 100km. If I can get longitudes and latitudes in distance 100km then I can read these heights.
Here is math for distance between geographic points:
public static float GetGeographicDistance(float latitudeA, float longitudeA, float latitudeB, float longitudeB, float earthRadius = 6371.007f)
{
float phi1 = latitudeA * Mathf.Deg2Rad;
float phi2 = latitudeB * Mathf.Deg2Rad;
float deltaPhi = (latitudeB - latitudeA) * Mathf.Deg2Rad;
float deltaLambda = (longitudeB - longitudeA) * Mathf.Deg2Rad;
float a = Mathf.Sin(deltaPhi / 2) * Mathf.Sin(deltaPhi / 2) +
Mathf.Cos(phi1) * Mathf.Cos(phi2) *
Mathf.Sin(deltaLambda / 2) * Mathf.Sin(deltaLambda / 2);
float c = 2.0f * Mathf.Atan2(Mathf.Sqrt(a), Mathf.Sqrt(1.0f - a));
return earthRadius * c;
}
Got it. If somebody needs it, here is the function in C#.
public static TGeographicCoordinate DestinationPointFromStartPoint(TGeographicCoordinate start, double distance = 100f, double bearing = 90f, double earthRadius = 6371.009d)
{
TGeographicCoordinate result = new TGeographicCoordinate();
double angularDistanceRdn = distance / earthRadius;
double bearingRad = Mathf.Deg2Rad * bearing;
double startLatRad = start.Latitude * Mathf.Deg2Rad;
double startLonRad = start.Longitude * Mathf.Deg2Rad;
result.Latitude = Math.Asin((Math.Sin(startLatRad) * Math.Cos(angularDistanceRdn)) +
(Math.Cos(startLatRad) * Math.Sin(angularDistanceRdn) * Math.Cos(bearingRad)));
result.Longitude = startLonRad + Math.Atan2(Math.Sin(bearingRad) * Math.Sin(angularDistanceRdn) * Math.Cos(startLatRad),
Math.Cos(angularDistanceRdn) - (Math.Sin(startLatRad)) * Math.Sin(result.Latitude));
result.Latitude *= Mathf.Rad2Deg;
result.Longitude *= Mathf.Rad2Deg;
return result;
}
What is the best way to decelerate at a speed of any given value (e.g. accelerationDropOff = 1.5f) before it reaches the end destination?
public bool MoveFromCurrentToPosition(float x, float y, float velocity, float acceleration, float deltaTime)
{
float startX = positionX, startY = positionY;
float endX = x, endY = y;
float deltaX = endX - startX;
float deltaY = endY - startY;
float speed = velocity;
float elapsed = 0.01f;
// On starting movement
float distance = (float)Math.Sqrt(Math.Pow(deltaX, 2) + Math.Pow(deltaY, 2));
float directionX = deltaX / distance;
float directionY = deltaY / distance;
isMoving = true;
// On update
if (isMoving == true)
{
positionX += directionX * speed * elapsed;
positionY += directionY * speed * elapsed;
if (currentAcceleration == 0)
{
currentAcceleration = acceleration;
}
else if (currentAcceleration >= maxAcceleration) // <- Don't accelerate anymore
{
speed *= currentAcceleration;
positionX += (directionX * speed) * deltaTime; positionY += (directionY * speed) * deltaTime;
bounds.X = (int)positionX; bounds.Y = (int)positionY;
}
else
{
currentAcceleration += acceleration;
speed *= currentAcceleration;
positionX += (directionX * speed) * deltaTime; positionY += (directionY * speed) * deltaTime;
bounds.X = (int)positionX; bounds.Y = (int)positionY;
}
float a = x, o = y;
double angle = Math.Atan2(o, a);
angle = angle * 180 / Math.PI;
movementDirection = (float)(180 - angle);
// Decelerate before reaching the end point
if (Math.Sqrt(Math.Pow(positionX - startX, 2) + Math.Pow(positionY - startY, 2)) >= distance)
{
positionX = endX;
positionY = endY;
isMoving = false;
return true;
}
}
return false;
}
I have been stuck on this problem for a hour or two and Math.exe is not responding. Can anyone point me in the correct direction please?
It seems as if you are mixing up speed (velocity) and acceleration. Speed is the change of position with respect to a given time frame. Acceleration is the change of speed with respect to a given time frame. For a constant acceleration, position and velocity change as follows:
v1 = v0 + a * t
x1 = x0 + v0 * t + 1/2 * a * t^2
v0, v1 and x0, x1 are the velocities and positions at the beginning and end of the time frame, respectively, a is the acceleration, t is the time frame length. This is the exact formula if you assume constant acceleration over the period of the time frame. Often, you find approximations like the following, which introduce some integration errors:
v1 = v0 + a * t
x1 = x0 + v1 * t
I would suggest to use the exact formulas.
As far as I understand your question, you want to find an acceleration, such that a body moving at initial velocity v0 stops after travelling d length units.
This gives you the following equations:
0 = v0 + a * t //target velocity of 0
d = 0 + v0 * t + 1/2 * a * t^2 //travel distance of d
The solution is:
a = -1/2 * v0^2 / d
The time needed for this motion is:
t = 2 * d / v0
So calculate the acceleration once at the beginning of the deccelerating movement and then update current position and velocity with the formulas above.
Some additional hints for your code:
If you want to square a variable x, use x * x instead of Math.pow(x, 2). It is easier to read and has a better performance.
If you already use XNA, then use its Vector2 structure. This makes a lot of things much easier. You can just add two vectors and don't need to care about each component separately. There are methods to get the length of a vector, and so on.
I know there are physic plugins for C# or XNA, but I want to create my own, so I can learn about the topic.
My problems are the following:
I try to apply an elastic impulse to my character with the right angle and velocity. The velocity is calculated the right way, the angle is not and distorts the results!
The next problem is, that my character gets into a shaking mode, though it should stand still. I know where the problem comes from, but I don't know how to fix it (edit: do I have to consider the penetration depth for that?)
The IPhysicsObject inherits the most important informations, the Vector2[] has the collisionPoint at index 0 and the penetration depth at index 1.
I have tried to work with this but yeah.. I don't know
public void ElasticImpulse(IPhysicsObject Object, Vector2[] _colPos)
{
//this function is down below
if (checkCollidingObjects(m_cCharacter, Object))
return;
//this List is like this declined:
//public static List<IPhysicsObject[]> CollidingObjects = new List<IPhysicsObject[]>();
//this list contains every pair of objects, that collided this frame, it is cleared after all physics and game logic is done.
CollidingObjects.Add(new IPhysicsObject[] { m_cCharacter, Object });
//deltavelocity is the velocity between two frames
Vector2 v1 = Velocity - DeltaVelocity;
float lv1 = (float)Math.Sqrt(v1.X * v1.X + v1.Y * v1.Y);
float m1 = Mass;
float k1 = Damping;
Vector2 v2 = Object.Physik.Velocity - Object.Physik.DeltaVelocity;
float lv2 = (float)Math.Sqrt(v2.X * v2.X + v2.Y * v2.Y);
float m2 = Object.Mass;
float k2 = Object.Physik.Damping;
Vector2 colDir1 = _colPos[0] - m_cCharacter.Position;
Vector2 colDir2 = _colPos[0] - Object.Position;
colDir1.Normalize();
colDir2.Normalize();
Vector2 colNorm1 = new Vector2(colDir1.Y, -colDir1.X);
Vector2 colNorm2 = new Vector2(colDir2.Y, -colDir2.X);
float ldir1 = (float)Math.Sqrt(colNorm1.X * colNorm1.X + colNorm1.Y * colNorm1.Y);
float ldir2 = (float)Math.Sqrt(colNorm2.X * colNorm2.X + colNorm2.Y * colNorm2.Y);
float pi = MathHelper.Pi;
//float angle1 = pi - ((v1.X * colNorm1.X + v2.Y * colNorm1.Y) / (lv1 * ldir1)) / v1.Length();
float angle1 = pi - (float)Math.Acos(((v1.X * colNorm1.X + v2.Y * colNorm1.Y) / (lv1 * ldir1)) / v1.Length());
angle1 = (float.IsNaN(angle1)) ? 0 : angle1;
//float angle2 = pi - ((v2.X * colNorm2.X + v2.Y * colNorm2.Y) / (lv2 * ldir1)) / v2.Length();
float angle2 = pi - (float)Math.Acos(((v2.X * colNorm2.X + v2.Y * colNorm2.Y) / (lv2 * ldir1)) / v2.Length());
angle2 = (float.IsNaN(angle2)) ? 0 : angle2;
//calculating the new velocities u 1/2. Got this formula out of the wiki link i posted above (took the german wiki version)
Vector2 u1 = (m1 * v1 + m2 * v2 - (m2 * (v1 - v2) * k2)) / (m1 + m2) - v1;
Vector2 u2 = (m1 * v1 + m2 * v2 - (m1 * (v2 - v1) * k1)) / (m1 + m2) - v2;
//transform the new velocities by the correct angle
Vector2 newV1 = new Vector2(
u1.X * (float)Math.Cos(angle1) - u1.Y * (float)Math.Sin(angle1),
u1.X * (float)Math.Sin(angle1) + u1.Y * (float)Math.Cos(angle1));
Vector2 newV2 = new Vector2(
u2.X * (float)Math.Cos(angle2) - u2.Y * (float)Math.Sin(angle2),
u2.X * (float)Math.Sin(angle2) + u2.Y * (float)Math.Cos(angle2));
newV1 = new Vector2(
(float.IsNaN(newV1.X)) ? 0 : newV1.X,
(float.IsNaN(newV1.Y)) ? 0 : newV1.Y);
newV2 = new Vector2(
(float.IsNaN(newV2.X)) ? 0 : newV2.X,
(float.IsNaN(newV2.Y)) ? 0 : newV2.Y);
AddForce(newV1);
Object.Physik.AddForce(newV2);
}
bool checkCollidingObjects(IPhysicsObject obj1, IPhysicsObject obj2)
{
if (CollidingObjects.Count > 0)
{
int a = CollidingObjects.FindIndex(x => (x[0] == obj1 && x[1] == obj2) ||
(x[1] == obj1 && x[0] == obj2));
return a != -1;
}
return false;
}
I am trying to emulate Vector3.TransformNormal without the DirectX library.
Is anyone able to explain how this function works, to allow me to recreate the function?
So far I know the inputs and have seen the description of what it does, but I don't know the calculations.
public static Vector3 TransformNormal(
Vector3 source,
Matrix sourceMatrix
)
This should do it (didn't test)
public Vector3 TransformNormal(Vector3 normal, Matrix matrix)
{
return new Vector3
{
X = normal.X * matrix.M11 + normal.Y * matrix.M21 + normal.Z * matrix.M31,
Y = normal.X * matrix.M12 + normal.Y * matrix.M22 + normal.Z * matrix.M32,
Z = normal.X * matrix.M13 + normal.Y * matrix.M23 + normal.Z * matrix.M33
};
}