• # Quaternion Powers

Math and Physics

Version 1.2, February, 2003

1.2 A minor correction with the formula of converting from Quat to Axis. The scale is missing a square root. Thanks to Shi for pointing that out. From version 1.0 - 1.1 The norm of a quaternion should be the square root of the q.q. The mistake was brought to my attention by several kind readers and upon checking the definition of the Euclidean properties for complex numbers, I realize the norm property [bquote][color=#000080][font=courier new,courier,monospace]

### || u+v || <= || u || + || v ||

[/font][/color][/bquote] is violated for the previous definition of the magnitude. The code in the samples are updated as well.

# Foreword

To me, the term 'Quaternion' sounds out of this world, like some term from quantum theory about dark matter, having dark secret powers. If you, too, are enthralled by this dark power, this article will bring enlightenment (I hope). The article will show you how to do rotations using quaternions, and bring you closer to understanding quaternions (and their powers). If you do spot a mistake please email me at [email="robin@cyberversion.com"]robin@cyberversion.com[/email]. Also if you intend to put this on your site, please send me a mail. I like to know where this ends up.

# Why use Quaternions?

To answer the question, let's first discuss some common orientation implementations.

## Euler representation

This is by far the simplest method to implement orientation. For each axis, there is a value specifying the rotation around the axis. Therefore, we have 3 variables [bquote]

### [font=Courier New][color=#000080]x, y, z <-- angle to rotate around global coordinate axis [/color][/font]

[/bquote] that vary between 0 and 360 degrees (or 0 - 2pi). They are the roll, pitch, and yaw (or pitch, roll, and yaw - whatever) representation. Orientation is obtained by multiplying the 3 rotation matrices generated from the 3 angles together (in a specific order that you define). Note: The rotations are specified with respect to the global coordinate axis frame. This means the first rotation does not change the axis of rotation for the second and third rotations. This causes a situation known as gimbal lock, which I will discuss later.

## Angle Axis representation

This implementation method is better than the Euler representation as it avoids the gimbal lock problem. The representation consists of a unit vector representing an arbitrary axis of rotation, and another variable (0 - 360) representing the rotation around the vector: [bquote][font=Courier New][color=#000080]

### x, y, z <-- unit vector representing arbitrary axis angle <-- angle to rotate around above axis

[/color][/font][/bquote] Why are these representations bad?

## Gimbal Lock

As rotations in the Euler representation are done with respect to the global axis, a rotation in one axis could 'override' a rotation in another, making you lose a degree of freedom. This is gimbal lock. Say, if the rotation in the Y axis rotates a vector (parallel to the x axis) so that the vector is parallel to the z axis. Then, any rotations in the z axis would have no effect on the vector. Later, I will show you an example of gimbal lock and how you can use quaternions to overcome it.

## Interpolation Problems

Though the angle axis representation does not suffer from gimbal lock, there are problems when you need to interpolate between two rotations. The calculated interpolated orientations may not be smooth, so you will get jerky rotation movements. Euler representation suffers from this problem as well.

# Let's get started

Before we begin, let's establish some assumptions I'll be making. I hate the way many articles leave this important section out, causing a great deal of confusion when it comes to the mathematics. Coordinate System - This article assumes a right hand coordinate system, like OpenGL. If you are using a left handed coordinate system like Direct3D, you may need to transpose the matrices. Note that the Direct3D samples have a quaternion library already, though I recommend you check through their implementation before using it. Rotation Order - The sequence of rotations in the Euler representation is X, then Y, and then Z. In matrix form: [bquote][font=Courier New][color=#000080]

### RotX * RotY * RotZ <-- Very Important

[/color][/font][/bquote] Matrix - Matrices are in column major format, like they are in OpenGL. [bquote]

### [font=Courier New][color=#000080]Example[nbsp][[nbsp]0[nbsp][nbsp]4[nbsp][nbsp]8[nbsp][nbsp]12 [nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]1[nbsp][nbsp]5[nbsp][nbsp]9[nbsp][nbsp]13 [nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]2[nbsp][nbsp]6[nbsp][nbsp]10[nbsp]14 [nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]3[nbsp][nbsp]7[nbsp][nbsp]11[nbsp]15[nbsp]][/color][/font]

[/bquote] Vectors and Points - Implemented as a 4x1 matrix so applying a transformation is of the order [bquote]

### [font=Courier New][color=#000080]Rotation[nbsp]Matrix*[nbsp][nbsp][[nbsp]vx [nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]vy [nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]vz [nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]1[nbsp][nbsp]][nbsp][nbsp]<--[nbsp]4x1[nbsp]vector[/color][/font]

[/bquote] This does not imply that I prefer OpenGL over Direct3D. It just happened that I learned OpenGL first, and so my quaternion knowledge was gained in OpenGL. Note: If you specify rotations in another order, certain quaternion functions will be implemented differently, especially those that deal with Euler representation.

# What is a Quaternion?

A complex number is an imaginary number that is defined in terms of i, the imaginary number, which is defined such that i * i = -1. A quaternion is an extension of the complex number. Instead of just i, we have three numbers that are all square roots of -1, denoted by i, j, and k. This means that
[bquote]j * j = -1 k * k = -1[/bquote]
So a quaternion can be represented as
[bquote]q = w + xi + yj + zk[/bquote]
where w is a real number, and x, y, and z are complex numbers. Another common representation is
[bquote]q=[ w,v ][/bquote]
where v = (x, y, z) is called a "vector" and w is called a "scalar". Although the v is called a vector, don't think of it as a typical 3 dimensional vector. It is a vector in 4D space, which is totally unintuitive to visualize.

# Identity Quaternions

Unlike vectors, there are two identity quaternions. The multiplication identity quaternion is
[bquote]q= [1,(0, 0, 0)][/bquote]
So any quaternion multiplied with this identity quaternion will not be changed. The addition identity quaternion (which we do not use) is
[bquote]q= [0,(0, 0, 0)][/bquote]

# Using quaternions as orientations

The first thing I want to point out is that quaternions are not vectors, so please don't use your preconceived vector mathematics on what I'm going to show. This is going to get very mathematical, so please bear with me. We need to first define the magnitude of a quaternion.
[bquote]|| q ||= Norm(q) =sqrt(w[sup]2[/sup] + x[sup]2[/sup] + y[sup]2[/sup] + z[sup]2[/sup])[/bquote]
A unit quaternion has the following property
[bquote]w[sup]2[/sup] + x[sup]2[/sup] + y[sup]2[/sup] + z[sup]2[/sup]=1[/bquote]
So to normalize a quaternion q, we do
[bquote]q = q / || q || = q / sqrt(w[sup]2[/sup] + x[sup]2[/sup] + y[sup]2[/sup] + z[sup]2[/sup])[/bquote]
What is so special about this unit quaternion is that it represents an orientation in 3D space. So you can use a unit quaternion to represent an orientation instead of the two methods discussed previously. To use them as orientations, you will need methods to convert them to other representations (e.g. matrices) and back, which will be discussed soon.

## Visualizing a unit quaternion

You can visualize unit quaternions as a rotation in 4D space where the (x,y,z) components form the arbitrary axis and the w forms the angle of rotation. All the unit quaternions form a sphere of unit length in the 4D space. Again, this is not very intuitive but what I'm getting at is that you can get a 180 degree rotation of a quaternion by simply inverting the scalar (w) component. Note: Only unit quaternions can be used for representing orientations. All discussions from here on will assume unit quaternions.

# Conversion from Quaternions

To be able to use quaternions effectively, you will eventually need to convert them to some other representation. You cannot interpret keyboard presses as quaternions, can you? Well, not yet.

## Quaternion to Matrix

Since OpenGL and Direct3D allow rotations to be specified as matrices, this is probably the most important conversion function, since homogenous matrices are the standard 3D representations. The equivalent rotation matrix representing a quaternion is [bquote]

### [font=Courier New][color=#000080]Matrix[nbsp]=[nbsp][nbsp][[nbsp]w[sup]2[/sup][nbsp]+[nbsp]x[sup]2[/sup][nbsp]-[nbsp]y[sup]2[/sup][nbsp]-[nbsp]z[sup]2[/sup][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]2xy[nbsp]-[nbsp]2wz[nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]2xz[nbsp]+[nbsp]2wy [nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]2xy[nbsp]+[nbsp]2wz[nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]w[sup]2[/sup][nbsp]-[nbsp]x[sup]2[/sup][nbsp]+[nbsp]y[sup]2[/sup][nbsp]-[nbsp]z[sup]2[/sup][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]2yz[nbsp]-[nbsp]2wx [nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]2xz[nbsp]-[nbsp]2wy[nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]2yz[nbsp]+[nbsp]2wx[nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]w[sup]2[/sup][nbsp]-[nbsp]x[sup]2[/sup][nbsp]-[nbsp]y[sup]2[/sup][nbsp]+[nbsp]z[sup]2[/sup][nbsp]][/color][/font]

[/bquote] Using the property of unit quaternions that w[sup]2[/sup] + x[sup]2[/sup] + y[sup]2[/sup] + z[sup]2[/sup] = 1, we can reduce the matrix to [bquote]

[/bquote]

## Quaternion to Axis Angle

To change a quaternion to a rotation around an arbitrary axis in 3D space, we do the following: [bquote]

### [font=Courier New][color=#000080]If[nbsp]the[nbsp]axis[nbsp]of[nbsp]rotation[nbsp]is[nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp](ax,[nbsp]ay,[nbsp]az) and[nbsp]the[nbsp]angle[nbsp]is[nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]theta[nbsp](radians) then[nbsp]the[nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]angle=[nbsp]2[nbsp]*[nbsp]acos(w) [nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]ax=[nbsp]x[nbsp]/[nbsp]scale [nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]ay=[nbsp]y[nbsp]/[nbsp]scale [nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]az=[nbsp]z[nbsp]/[nbsp]scale where[nbsp]scale[nbsp]=[nbsp]sqrt[nbsp](x[sup]2[/sup][nbsp]+[nbsp]y[sup]2[/sup][nbsp]+[nbsp]z[sup]2[/sup])[/color][/font]

[/bquote] Another variation I have seen is that the scale = sin(acos(w)). They may be equivalent, though I didn't try to find the mathematical relationship behind them. Anyway if the scale is 0, it means there is no rotation so unless you do something, the axis will be infinite. So whenever the scale is 0, just set the rotation axis to any unit vector with a rotation angle of 0.

# A Simple Example

In case you are getting confused with what I'm getting at, I will show you a simple example here. Say the camera orientation is represented as Euler angles. Then, in the rendering loop, we position the camera using [bquote][font=Courier New][color=#000080]

### RotateX * RotateY * RotateZ * Translate

[/color][/font][/bquote] where each component is a 4x4 matrix. So if we are using a unit quaternion to represent the camera orientation, we have to convert the quaternion to a matrix first [bquote][font=Courier New][color=#000080]

### Rotate (from Quaternion) * Translate

[/color][/font][/bquote] A more specific example in OpenGL:
EulerQuaternion
[font=courier new,courier,monospace]

### [color=#000080]glRotatef( angleX, 1, 0, 0)[/color]

[/font]
[color=#000080][font=courier new,courier,monospace]

### glRotatef( angleY, 0, 1, 0)

[/font][/color]
[color=#000080]

### [font=courier new,courier,monospace]glRotatef( angleZ, 0, 0, 1)[/font]

[/color]
[color=#000080]

### [font=courier new,courier,monospace]// translate[/font]

[/color]
[color=#000080]

### [font=courier new,courier,monospace]// convert Euler to quaternion[/font]

[/color]
[color=#000080][font=courier new,courier,monospace]

### // convert quaternion to axis angle

[/font][/color]
[color=#000080]

### [font=courier new,courier,monospace]glRotate(theta, ax, ay, az)[/font]

[/color]
[color=#000080]

### [font=courier new,courier,monospace]// translate[/font]

[/color]
The above implementations are equivalent. The point I'm trying to get across is that using quaternions for orientation is the same as using Euler or Axis angle representation and that they can be interchanged through the conversion functions I've described. Note that the above quaternion representation will also incur gimbal lock like the Euler method. Of course, you do not know how to make the rotation to be a quaternion in the first place but we will get to that shortly. Note: If you are using Direct3D or OpenGL, you may not have to deal with matrices directly, but matrix concatenation is something that the API does, so it's worth learning about them.

# Multiplying Quaternions

Since a unit quaternion represents an orientation in 3D space, the multiplication of two unit quaternions will result in another unit quaternion that represents the combined rotation. Amazing, but it's true. Given two unit quaternions
[bquote]Q1=(w1, x1, y1, z1); Q2=(w2, x2, y2, z2);[/bquote]
A combined rotation of unit two quaternions is achieved by
[bquote]Q1 * Q2 =( w1.w2 - v1.v2, w1.v2 + w2.v1 + v1*v2)[/bquote]
where
[bquote]v1= (x1, y1, z1) v2 = (x2, y2, z2)[/bquote]
and both . and * are the standard vector dot and cross product. However an optimization can be made by rearranging the terms to produce
[bquote]w=w1w2 - x1x2 - y1y2 - z1z2 x = w1x2 + x1w2 + y1z2 - z1y2 y = w1y2 + y1w2 + z1x2 - x1z2 z = w1z2 + z1w2 + x1y2 - y1x2[/bquote]
Of course, the resultant unit quaternion can be converted to other representations just like the two original unit quaternions. This is the real beauty of quaternions - the multiplication of two unit quaternions in 4D space solves gimbal lock because the unit quaternions lie on a sphere. Be aware that the order of multiplication is important. Quaternion multiplication is not commutative, meaning
[bquote]q1 * q2 does not equal q2 * q1[/bquote]
Note: Both quaternions must refer to the same coordinate axis. I made the mistake of combining two quaternions from different coordinate axes, and I had a very hard time wondering why the result quaternion fails in certain angles only.

# Conversion To Quaternions

Now we learn how to convert other representations to quaternions. Although I do not use all the conversions in the sample program, there are times when you'll need them when you want to use quaternion orientation for more advanced stuff like inverse kinematics.

## Axis Angle to Quaternion

A rotation around an arbitrary axis in 3D space can be converted to a quaternion as follows [bquote]

### [font=Courier New][color=#000080]If[nbsp]the[nbsp]axis[nbsp]of[nbsp]rotation[nbsp]is[nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp](ax,[nbsp]ay,[nbsp]az)-[nbsp]must[nbsp]be[nbsp]a[nbsp]unit[nbsp]vector and[nbsp]the[nbsp]angle[nbsp]is[nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp][nbsp]theta[nbsp](radians) [nbsp][nbsp][nbsp][nbsp]w[nbsp][nbsp][nbsp]=[nbsp][nbsp][nbsp]cos(theta/2) [nbsp][nbsp][nbsp][nbsp]x[nbsp][nbsp][nbsp]=[nbsp][nbsp][nbsp]ax[nbsp]*[nbsp]sin(theta/2) [nbsp][nbsp][nbsp][nbsp]y[nbsp][nbsp][nbsp]=[nbsp][nbsp][nbsp]ay[nbsp]*[nbsp]sin(theta/2) [nbsp][nbsp][nbsp][nbsp]z[nbsp][nbsp][nbsp]=[nbsp][nbsp][nbsp]az[nbsp]*[nbsp]sin(theta/2)[/color][/font]

[/bquote] The axis must first be normalized. If the axis is a zero vector (meaning there is no rotation), the quaternion should be set to the rotation identity quaternion.

## Euler to Quaternion

Converting from Euler angles to a quaternion is slightly more tricky, as the order of operations must be correct. Since you can convert the Euler angles to three independent quaternions by setting the arbitrary axis to the coordinate axes, you can then multiply the three quaternions together to obtain the final quaternion. So if you have three Euler angles (a, b, c), then you can form three independent quaternions [bquote]Qx = [ cos(a/2), (sin(a/2), 0, 0)] Qy = [ cos(b/2), (0, sin(b/2), 0)] Qz = [ cos(c/2), (0, 0, sin(c/2))][/bquote] And the final quaternion is obtained by Qx * Qy * Qz.

# Demo - Avoiding Gimbal Lock

Finally, we've reached what you all been waiting for: "How can quaternions avoid gimbal lock.?" The basic idea is
1. Use a quaternion to represent the rotation.
2. Generate a temporary quaternion for the change from the current orientation to the new orientation.
3. PostMultiply the temp quaternion with the original quaternion. This results in a new orientation that combines both rotations.
4. Convert the quaternion to a matrix and use matrix multiplication as normal.
Firstly, I want to make a disclaimer regarding the sample code. The code is ugly and very poorly organized. But do remember, this is just a cut down version of my program when I was testing quaternions, and I'm not getting paid for this. There are two executable samples that I have included. The first program, CameraEuler.exe, is an example for camera implementation using Euler angles. The main concern should be the Main_Loop function in main.cpp. The main thing you should take note (in the while loop) is
1. There are 3 angles I keep track of for rotation in the X, Y, and Z axis.
2. With every key press, I adjust the corresponding rotation variable.
3. In the while loop, I translate and then just convert the 3 Euler angles to rotation matrices and multiply them into the final transformation matrix.
Use the up/down keys to rotate around the X axis, left/right to rotate around the Y axis and Insert/PageUp to rotate around the Z axis. This program suffers from gimbal lock. If you want to see it in action, rotate the camera so that the yaw is 90 deg. Then try rotating in the X and Z direction. See what happens. Now for the quaternion solution. The program is CameraQuat.exe and it is a slight modification of the previous program. The main point you should take note (in the while loop) is
1. The orientation of the camera is a quaternion.
2. There are 3 angles corresponding to the keypress. Note the angles are meant to be an on/off switch (not accumulative). I reset them inside the while loop. Of course this is not the best way to do it but as I said, it is a quick job.
3. I convert the 3 angles to a temporary quaternion.
4. I multiply the temporary quaternion to the camera quaternion to obtain the combined orientation. Note the order of multiplication.
5. The camera rotation is then converted to the Axis Angle representation for transforming the final matrix.
When a key is pressed, I generate a temporary quaternion corresponding to the key for a small rotation in that particular axis. I then multiply the temporary quaternion into the camera quaternion. This concatenation of rotations in 4D space will avoid gimbal lock. Try it and see for yourself. The camera quaternion has to be changed into either a matrix form or equivalence form so that you can concatenate it into a final transformation matrix. You have to do this for every quaternion you use, as 4D space and 3D space just don't mix. In the case of OpenGL, I just changed the quaternion to an Axis Angle representation and let the API do the rest. Although I did not use the global Euler angles for rotations in the second program, I have left them there as a guide for you to see the similar Euler rotations in the first program. Note the Euler angles will be incorrect if you rotate more than 1 axis (because it counts the keypress rather than getting the Euler angles from the camera quaternion). It is just a reference for you to see that when you rotate the yaw to 90 deg when the program starts, the gimbal lock problem is no more. Note: I don't recommend you use my math library as it is. Understand the quaternion and write your own. For your information, I am going to throw all of that away and rewrite it too. It is just too messy and ugly for my taste.

# What I did not show

If you notice, I did not show how to convert from a Quaternion to the Euler angle. That's because I have yet to find a conversion that works perfectly. The only way I know is to obtain a matrix from the quaternion and try to extract the Euler angles from the matrix. However, as Euler to matrix conversions are a many-to-one relationship (due to sin and cos), I do not know how to get the reverse even using atan2. If anyone knows how to extract Euler angles from a matrix accurately, please do share with me. The other thing I did not show is the conversion of a matrix to a quaternion. I didn't think I needed this conversion as you can convert the Euler and Axis angle representation to quaternion straight without needing to throw them to a matrix.

# More you can do - SLERP

If you think you are a quaternion master, think again. There is still more to learn about them. Remember I said something about why the Axis Angle representation is bad? Does the word 'interpolation' ring a bell? I don't have the time to write about interpolations using quaternions. This article has taken much longer than I had anticipated. I can give you the basic idea about SLERP (Spherical Linear Interpolation), which is basically generating a series of quaternions between two quaternion orientations (which you specify). The series of quaternions will result in smooth motion between the first and end quaternion (something which both the Euler and Axis Angle representation cannot achieve consistently).

# Final Words

I hope this article can clear up any mystery behind the quaternion theory. A final word of caution again: Don't multiply two quaternions from different coordinate frames. Nothing but pain and hair loss will result from it. With your new found powers, I bid thee farewell. Take care, .. and watch your back.

Report Article

## Create an account

Register a new account

• ### What is your GameDev Story?

In 2019 we are celebrating 20 years of GameDev.net! Share your GameDev Story with us.

• 0
• 4
• 0
• 0
• 0

• 15
• 11
• 9
• 9
• 42
• ### Similar Content

• By _Nyu
Hello,
I'm trying to make a PBR vulkan renderer and I wanted to implement Spherical harmonics for the irradiance part (and maybe PRT in the future but that's another story).
the evaluation on the shader side seems okay (it look good if I hardcode the SH directly in the shader) but when I try to generate it from a .hdr map it output only gray scale.
It's been 3 days I'm trying to debug now I just have no clue why all my colour coefficients are gray.
Here is the generation code:

SH2 ProjectOntoSH9(const glm::vec3& dir) { SH2 sh; // Band 0 sh.coef0.x = 0.282095f; // Band 1 sh.coef1.x = 0.488603f * dir.y; sh.coef2.x = 0.488603f * dir.z; sh.coef3.x = 0.488603f * dir.x; // Band 2 sh.coef4.x = 1.092548f * dir.x * dir.y; sh.coef5.x = 1.092548f * dir.y * dir.z; sh.coef6.x = 0.315392f * (3.0f * dir.z * dir.z - 1.0f); sh.coef7.x = 1.092548f * dir.x * dir.z; sh.coef8.x = 0.546274f * (dir.x * dir.x - dir.y * dir.y); return sh; } SH2 ProjectOntoSH9Color(const glm::vec3& dir, const glm::vec3& color) { SH2 sh = ProjectOntoSH9(dir); SH2 shColor; shColor.coef0 = color * sh.coef0.x; shColor.coef1 = color * sh.coef1.x; shColor.coef2 = color * sh.coef2.x; shColor.coef3 = color * sh.coef3.x; shColor.coef4 = color * sh.coef4.x; shColor.coef5 = color * sh.coef5.x; shColor.coef6 = color * sh.coef6.x; shColor.coef7 = color * sh.coef7.x; shColor.coef8 = color * sh.coef8.x; return shColor; } void SHprojectHDRImage(const float* pixels, glm::ivec3 size, SH2& out) { double pixel_area = (2.0f * M_PI / size.x) * (M_PI / size.y); glm::vec3 color; float weightSum = 0.0f; for (unsigned int t = 0; t < size.y; t++) { float theta = M_PI * (t + 0.5f) / size.y; float weight = pixel_area * sin(theta); for (unsigned int p = 0; p < size.x; p++) { float phi = 2.0 * M_PI * (p + 0.5) / size.x; color = glm::make_vec3(&pixels[t * size.x + p]); glm::vec3 dir(sin(phi) * cos(theta), sin(phi) * sin(theta), cos(theta)); out += ProjectOntoSH9Color(dir, color) * weight; weightSum += weight; } } out.print(); out *= (4.0f * M_PI) / weightSum; }
outside of the SHProjectHDRImage function that's pretty much the code from MJP that you can check here:
https://github.com/TheRealMJP/LowResRendering/blob/2f5742f04ab869fef5783a7c6837c38aefe008c3/SampleFramework11/v1.01/Graphics/SH.cpp
I'm not doing anything fancy in term of math or code but I that's my first time with those so I feel like I forgot something important.
basically for every pixel on my equi-rectangular hdr map I generate a direction, get the colour and project it on the SH
but strangely I endup with a SH looking like this:
coef0: 1.42326 1.42326 1.42326
coef1: -0.0727784 -0.0727848 -0.0727895
coef2: -0.154357 -0.154357 -0.154356
coef3: 0.0538229 0.0537928 0.0537615
coef4: -0.0914876 -0.0914385 -0.0913899
coef5: 0.0482638 0.0482385 0.0482151
coef6: 0.0531449 0.0531443 0.0531443
coef7: -0.134459 -0.134402 -0.134345
coef8: -0.413949 -0.413989 -0.414021
with the HDR map "Ditch River" from this web page http://www.hdrlabs.com/sibl/archive.html
but I also get grayscale on the 6 other hdr maps I tried from hdr heaven, it's just different gray.
If anyone have any clue that would be really welcome.
• By pedro87
I am trying to understand calculating normal vectors in my game engine book. I am having issues understanding a specific section about using gradient to calculate the normal vector. I have attached an image of the page, basically i am unsure where the gradient is coming into play, or just really anything thats going on, on the page. Any help would be appreciated.

• Hello.
I'm trying to implement normal mapping. I've been following this: http://ogldev.atspace.co.uk/www/tutorial26/tutorial26.html
The problem is that my tangent vectors appear rather obviously wrong. But only one of them, never both. Here's my code for calculating the tangents:
this.makeTriangle = function(a, b, c) { var edge1 = VectorSub(b.pos, a.pos); var edge2 = VectorSub(c.pos, a.pos); var deltaU1 = b.texCoords[0] - a.texCoords[0]; var deltaV1 = b.texCoords[1] - a.texCoords[1]; var deltaU2 = c.texCoords[0] - a.texCoords[0]; var deltaV2 = c.texCoords[1] - a.texCoords[1]; var f = 1.0 / (deltaU1 * deltaV2 - deltaU2 * deltaV1); var vvec = VectorNormal([ f * (deltaV2 * edge1[0] - deltaV1 * edge2[0]), f * (deltaV2 * edge1[1] - deltaV1 * edge2[1]), f * (deltaV2 * edge1[2] - deltaV1 * edge2[2]), 0.0 ]); var uvec = VectorNormal([ f * (-deltaU2 * edge1[0] - deltaU1 * edge2[0]), f * (-deltaU2 * edge1[1] - deltaU1 * edge2[1]), f * (-deltaU2 * edge1[2] - deltaU1 * edge2[2]), 0.0 ]); if (VectorDot(VectorCross(a.normal, uvec), vvec) < 0.0) { uvec = VectorScale(uvec, -1.0); }; /* console.log("Normal: "); console.log(a.normal); console.log("UVec: "); console.log(uvec); console.log("VVec: "); console.log(vvec); */ this.emitVertex(a, uvec, vvec); this.emitVertex(b, uvec, vvec); this.emitVertex(c, uvec, vvec); }; My vertex shader:
precision mediump float; uniform mat4 matProj; uniform mat4 matView; uniform mat4 matModel; in vec4 attrVertex; in vec2 attrTexCoords; in vec3 attrNormal; in vec3 attrUVec; in vec3 attrVVec; out vec2 fTexCoords; out vec4 fNormalCamera; out vec4 fWorldPos; out vec4 fWorldNormal; out vec4 fWorldUVec; out vec4 fWorldVVec; void main() { fTexCoords = attrTexCoords; fNormalCamera = matView * matModel * vec4(attrNormal, 0.0); vec3 uvec = attrUVec; vec3 vvec = attrVVec; fWorldPos = matModel * attrVertex; fWorldNormal = matModel * vec4(attrNormal, 0.0); fWorldUVec = matModel * vec4(uvec, 0.0); fWorldVVec = matModel * vec4(vvec, 0.0); gl_Position = matProj * matView * matModel * attrVertex; } And finally the fragment shader:
precision mediump float; uniform sampler2D texImage; uniform sampler2D texNormal; uniform float sunFactor; uniform mat4 matView; in vec2 fTexCoords; in vec4 fNormalCamera; in vec4 fWorldPos; in vec4 fWorldNormal; in vec4 fWorldUVec; in vec4 fWorldVVec; out vec4 outColor; vec4 calcPointLight(in vec4 normal, in vec4 source, in vec4 color, in float intensity) { vec4 lightVec = source - fWorldPos; float sqdist = dot(lightVec, lightVec); vec4 lightDir = normalize(lightVec); return color * dot(normal, lightDir) * (1.0 / sqdist) * intensity; } vec4 calcLights(vec4 pNormal) { vec4 result = vec4(0.0, 0.0, 0.0, 0.0); \${CALC_LIGHTS} return result; } void main() { vec4 surfNormal = vec4(cross(vec3(fWorldUVec), vec3(fWorldVVec)), 0.0); vec2 bumpCoords = fTexCoords; vec4 bumpNormal = texture(texNormal, bumpCoords); bumpNormal = (2.0 * bumpNormal - vec4(1.0, 1.0, 1.0, 0.0)) * vec4(1.0, 1.0, 1.0, 1.0); bumpNormal.w = 0.0; mat4 bumpMat = mat4(fWorldUVec, fWorldVVec, fWorldNormal, vec4(0.0, 0.0, 0.0, 1.0)); vec4 realNormal = normalize(bumpMat * bumpNormal); vec4 realCameraNormal = matView * realNormal; float intensitySun = clamp(dot(normalize(realCameraNormal.xyz), normalize(vec3(0.0, 0.0, 1.0))), 0.0, 1.0) * sunFactor; float intensity = clamp(intensitySun + 0.2, 0.0, 1.0); outColor = texture(texImage, fTexCoords) * (vec4(intensity, intensity, intensity, 1.0) + calcLights(realNormal)); //outColor = texture(texNormal, fTexCoords); //outColor = 0.5 * (fWorldUVec + vec4(1.0, 1.0, 1.0, 0.0)); //outColor = vec4(fTexCoords, 1.0, 1.0); outColor.w = 1.0; } Here is the result of rendering an object, showing its normal render, the uvec, vvec, and texture coordinates (each commented out in the fragment shader code):

Normal map itself:

The uvec, as far as I can tell, should not be all over the place like it is; either this, or some other mistake, causes the normal vectors to be all wrong, so you can see on the normal render that for example there is a random dent on the left side which should not be there. As far as I can tell, my code follows the math from that tutorial. I use right-handed corodinates.
So what could be wrong?

• I need a measure for how close a collection of points is to forming a straight line in 2d.
Currently I use Principle Component Analysis in 2d and take the mean of the of square on the second axis.
It works fine, but I feel it is overkill.

I would like something simpler I could express in a li rary like Tensorflow.

Is there a simpler way?

• So I have hundreds of moving objects that need to check there speed. One of the reasons they need to check there speed is so they don't accelerate into oblivion, as more and more force is added to each object.
At first I was just using the Unity vector3.magnitude. However this is actually very slow; when used hundreds of times.
Next I tried the dot-product check:  vector3.dot(this.transform.foward, ShipBody.velocity) The performance boost was fantastic. However this only measures speed in the forward direction. Resulting in bouncing objects accelerating way past the allowed limit.

I am hoping someone else knows a good way for me to check the speed with accuracy, that is fast on the CPU. Or just any magnitude calculations that I can test when I get home later.

What if I used  vector3.dot(ShipBody.velocity.normalized, ShipBody.velocity)?
How slow is it to normalize a vector, compared to asking it's magnitude?
×