# OpenGL OpenGL Procedural Planet Generation - Quadtrees and Geomipmapping

This topic is 1812 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

## Recommended Posts

Recently, I've been looking into procedural generation in the form of a planet. After reading many many articles i managed to create my own 'planet' though some work still needs to be done for it to look like a planet and yes... The repetition of many was intended. Before i can really start making my planet generate nice terrain i will need to implement some form of LOD. After reading through pretty much all of the articles on LOD that are on http://vterrain.org/LOD/Implementations/ i believe it would be best to use a Quadtree due to how i create my planet.

So first off here is a picture of my planet:

That picture might be a bit big but you get the idea... So the planet has it's heights generated on the shader using perlin noise and so it can change in real time. However, it is clear that the scale is all wrong. I need an adaptive quad tree to achieve good looking terrain at space, atmosphere and ground levels. The only problem is... I don't know how to fill the quad tree with the data.

How is the vertex data stored in a quad tree and split? Splitting the quad tree is one thing but splitting the data is another.

I create the sphere from a cube by normalizing it. So this brings me on to another problem: I'll need 6 quad trees to split my planet up. One for each face. How will i extract the planes to subdivide them from vertex data?

My last question is how do i actually fill the quadtree with the data. Most examples on the internet all use things such as Ogre/Irrlicht/D3D and i haven't seen any detail on how to do this in OpenGL. Some source code would be a great help. If you require any of my source code to answer just comment and i will attatch it. For example - How could i change this to hold vertex data - https://github.com/veeableful/Adaptive_Quadtree_Minimal Thanks in advance.

A video of the shader noise generation can be seen here:
">
Edited by Vangoule

##### Share on other sites

I have also been working on/off on a planet generator.  While I make no claims to the 'idealness' of my method, it does seem to work.  You're on the right track for an adaptive quadtree, because that is what I'm using.  I looked into a bunch of terrain algorithms, and (ignoring older, more dynamic algorithms like ROAM), I've seen two main strategies:

• Fixed number of patches, dynamic detail levels:  You need a 'small' outdoor environment.  The ground is divided into patches which can be preloaded into GL (or DX if you prefer).  Each patch can be drawn at different detail levels.  So, a given 'square' of ground might be drawn at 1x1 all the way to 32x32, or even 64x64 quads.  This seems fine for environments where the draw distance will always be limited.  If this is a game 'on foot' or a flying game with a low maximum altitude, this works fine.  You can even 'scroll' over a larger world by loadeding in extra patches where you need them.  The basic render loop for this just needs to decide which patches are visible (frustum cull) and then at what detail level.  I started with this one, and it seems to work great until you get too far up from the planet surface.

• Fixed patch detail, dynamic number of patches:  This is what I'm using because it makes more sense for planets, where the scale and total amount drawn can change a bit.  I chose I fixed patch detail amount.  Suppose its 32x32 quads.  All patches are drawn at the fixed patch detail level.  As detail is needed, patches are split into 4 children subpatches.  Detail is taken away by recombining the 4 children back into their parent patch.

I use the same basic approach you are taking: I start with a cube of quads that I have squished into a sphere.

I do not understand this question:

How will i extract the planes to subdivide them from vertex data?

Are you squishing a cube into a sphere, and they trying to decide which quads in the sphere belong to which plane?  I did it this way:

• I am using opengl , so I assume the coordinate system of positive Y is 'up', positive X is 'right' and positive Z sticks out of the monitor poking me in the eye.
• Decide the center of the planet is <0,0,0>.  And that the planet will fit in a box from <-1,-1,-1> to <1,1,1>.  That makes all the math easy.  You can move the planet around and resize it later.
• Start with a working quadtree algorithm for a single plane, at Y=1.  The plane should cover from <-1,1,-1> to < 1,1,1>.  A nice square.  No deformations yet, just draw a quadtree in wireframe.  As you get closer and farther than the plane, it should split/recombine.
• After that is working, 'bend' the square around the top of the sphere.  For each point, simply normalize it, so it is a distance of 1 from the center point of <0,0,0>.  This will create the top 'lid' of the planet.  You should have the top part of the ball.
• Displace the points on the top of the ball.  Not in the 'y' direction, but in the 'up' direction relative to the inhabitants of the ball.  You will find this direction is the same direction as the 'normal' of the ball, and since the radius of the ball is 1, and it is centered at <0,0,0>, the position of your point IS also the direction you need to displace in.  Very convenient.
• You should have a working 'top' of a planet, where you can fly around and get more/less detail where you need it.
• Add in the other 5 planes.

That was the basic idea.  Do you have any specific questions?

##### Share on other sites

I'll also add a video of it mostly working.  Notice there are seams where the detail levels between adjacent areas don't match.  My next task (when I have time, that is) is to fill those in.

">

##### Share on other sites

Thanks for the reply. The bit on the sorting out which vertices belong to which face helps a lot. My main problem though is actually storing data in a Quad Tree. Most information from articles simply say - Store it in a quadtree in patches size n*n. How would i adapt a Quad Tree that is created like this - https://github.com/veeableful/Adaptive_Quadtree_Minimal to be able to hold data? Some source code would be much appreciated.

##### Share on other sites

For my planet generator (in the planning stages right now), I'm going to use an icosahedron for the basic spherical subdivision to get a set of equilateral triangles which will each represent flat-ish terrain cells. These cells can then be subdivided recursively into 4 smaller equilateral triangles, then extruded to lie on the surface of the sphere (or something else, depending on elevation). You could theoretically use a quad tree for the triangle subdivisions though the math for the spatial subdivision will be a little weird (because you're dealing with triangles and not rectangles).

I'm not sure how it's going to work out in practice, but it's another idea for you to think about.

For rendering a planet at a distance (i.e. you can see the whole thing), it's probably best to just build a big vertex buffer for the whole thing because that won't take more than a few tens of thousands of triangles to achieve pixel-perfect curvature, then only switch to LOD terrain when the view gets really close to the surface (within 10-20 miles for earth).

##### Share on other sites

For rendering a planet at a distance (i.e. you can see the whole thing), it's probably best to just build a big vertex buffer for the whole thing because that won't take more than a few tens of thousands of triangles to achieve pixel-perfect curvature, then only switch to LOD terrain when the view gets really close to the surface (within 10-20 miles for earth).

I agree with this.  Right now, each quadtree patch lives in its own VBO.  I want to add two things 1. have neighboring patches share VBOs (for instance, the initial 6 patches used for the starting 'cube' planet could share one VBO) and 2. let 'root' level patches be drawn at different mipmap levels.  This way, I could do this:  1) extreme distance: planet is a sprite 2) far away, planet is all in one VBO, drawn at different tesselations 3) when close up, start dividing into the quadtrees

Later tonight I'll post the structs I'm using.  Yes, right, structs, I'm working in plain C because this is a hobby project.  But you can adapt the info to these C++ classes easily.

Are you trying to generate the planet on the GPU or on the CPU?  What I'm doing is creating the geometry on the CPU (because I need it for reasons other than rendering), and I send VBOs representing the geometry in each patch to the graphics card.

This is how I'm spliting patches right now:

Currently, I'm using random midpoint displacement.  Suppose I have a patch I want to subdivide, and it's 33x33.  I will create four patches, but the geometry has to be based on this parent patch.   So, what I do is 'scale up' just I would do for an image.  I take a 17x17 vertex region of the parent patch, and linearly interpolate it to a 33x33 patch.  As I copy, the in-between vertices (vertices that exist in the child but not the parent) are given a displacement.  You can see, there's a problem here.  Suppose you have two neighboring patches sharing an edge.  When I divide one patch, that edge will be expanded in detail by creating more points.  Later, when I subdivide the other patch, the same edge will be expanded again, except different values will be created.  This will create a mismatch gap between two neighboring patches even though they are both at the same detail level.

There are three ways to deal with this.

1. (the way I currently am now): each patch knows it's neighbors.  (Not just which patch is it's neighbor, but which edge of that neighbor connects to it.  Sure, on a flat plane like graph paper, each squares's 'right' edge lines up easy with the 'left' edge of another square.  Cut up the graph paper and make various shapes out of it(like a cube), and you'll find that the 'left' edge of one square might meet the 'bottom' edge of it's neighbor.  You need to match edge-to-edge, not just neighbor to neighbor.)  When a patch divides, if it's neighbor already expanded out detail for an edge, the same points are copied into the new patch.  This way, when both patches are at the same detail level, they match up perfectly.

2. The way to I want to do it:  The random number generator will always generate the same displacement values for the same edges no matter which patch is doing it (Seed the RNG with quick hash of the edge points before generating the numbers that detail that edge.)  If I can get this working, patch edge fix-up logic can go away.

3. The way a real game would deal with it:  All the data is pre-stored on disk.  When dividing a patch, stream in child patches from disc, the authoring tools already seamed everything up.  Patches don't need to know their neighbors

The above does not address the gaps created with neighboring patches are at different detail levels.  This is something I'm still working on.  I will probably settle with 'skirts'.

##### Share on other sites

Currently i've got my normalized cube on the CPU i then send this through a VBO to the shaders. Here i apply perlin noise functions to create my height. I then choose what texture to use based on height with a change value made with sin cos and tan to make sure it's not a straight line of height.

If i can just implement a Quad Tree i'll be able to continue to create my procedural planet. This is my main goal right now.

So top priority: Implement Quad Tree. I can show source code of my project if necessary.

##### Share on other sites

So currently my cube has thousands of vertices when i export it from
time a lot.

Oh, now I know what you mean when you said that you were having trouble deciding which vertices go to which of the 6 planes.  I think it would be a lot easier if instead you were to build the sphere inside the application, instead of loading a sphere made somewhere else.  Before I show you how I build the sphere in-program, let me show you the struct I use first.  I removed a few things specific to my app, so it's pretty basic:

#define EDGE_LEFT    0
#define EDGE_RIGHT   1
#define EDGE_TOP     2
#define EDGE_BOTTOM  3

{
gx_vbuffer_t* vb;      //vertex buffer storing geometry for this patch
vindex startindex;     //start and end draw index for this patch
vindex endindex;

zint32 w;  		//2d array dimension.  w and h can both be 33 for example, for 33x33 patch
zint32 h;

struct quadarray_s* children[4];   //4 children patches to subdivide
int self; 			   //which one of my parent's children am i?

vec3 center;  //center of the quadarray
float size;   //area metric
float dsize;  //displacement metric

//neighbors
int adjacent_edge[4];       //inverse relationship from neighbor



Probably the most confusing part is tracking neighbors, but you can probably ignore that at the moment.  You can get started without it, there will just be cracks and seems between patches.

I create 6 quadarray patches, one for each face of the cube: (warning, messing C code!)


for (face = 0; face < 6; face++)  //6 sides of cube...
{

qa->size = .3;    //I set the 'size' of this patch to .3.  it's rather arbitray, I just declared this patch has area metric of .3
qa->dsize = .01;   //I set the detail size of this patch to .1.  It's also arbitray, it just affects how 'big' the displacements are

// a and b will visit each vertex in the quadarray.
for (a=0;a < qa->w;a++)			//0 to 'width'
{
for (b=0;b<qa->h;b++)		//0 to 'height'
{
vec3* vv = quadarray_get_vertex(qa, a, b);

//convert each of the integer vertex coordinates to float -1 to 1

fa = ((a-qa->w/2)/  (float) (qa->w-1)) *2 ;
fb  =  ((b-qa->h/2)/ (float) (qa->h-1)) *2 ;

//now decide on a plane depending on what face we are on

switch (face)
{
case 0: //top Y=1
vec3set(*vv, fa, 1, fb);
break;

case 1: //bottom Y=-1
vec3set(*vv, -fa, -1, fb);
break;

case 2: //front Z=1
vec3set(*vv, -fa, fb, 1);
break;

case 3:  //back Z=-1
vec3set(*vv, fa, fb, -1);
break;

case 4:  //left  X=-1
vec3set(*vv, -1, fb, -fa);
break;

case 5:  //right X=1
vec3set(*vv, 1, fb, fa);
break;

}

//normalize vv to unit length
d = vec3dot(*vv,*vv);
d = sqrt(d);
vec3scale(*vv, 1/d);
}
}

}



The above creates a perfect ball radius 1 centered around <0,0,0>. The quadarrays vector stores the current list of active quadarrays.  'Active' means current detail level.

The main algorithm I use is in the pseudocode here:

//once per render/update loop:

//test for visibility and draw
if (  qa is in camera frustum )
draw (qa);

if ( patch_too_small( qa, camera_position))
split_patch(qa);

if (patch_too_big(qa, camera_position))
combine_patches(qa);

}

{

if (!qa->children[0])
{

//patch has bot been split before, need to generate or load children

qa->children[0] = ...
qa->children[1] =...;
qa->children[2] =...
qa->children[3] = ...

}

}

remove qa->parent->child[0]
remove qa->parent->child[1]
remove qa->parent->child[2]
remove qa->parent->child[3]

}



I currently alpha-blend details in and out, so I am doing something more complicated than this.  I have a 'fade in' and 'fade out' list.  I hope to find some time to fix the cracks between detail levels and then cleanup the code.  After that, I'll put it on github, and everyone can rip it apart and flame my archaic C style :)

##### Share on other sites
Ahh that simplifies it a lot. By building it in the application it should theoretically subdivide infinitely. I'll read more later.

Edit: So after reading the code. I understand that making the cube in the application will be easier. However, the style of your quad tree looks different to most the stuff i've found online. Lets say i have these faces and have a quadtree/quadarray setup for them. When i split the face. It creates the 4 children. However, how do i actually split the face and not just the quadtree. I remove the current geometry and create 4 new patches the same W and H but at a different scale as it's children? How do i scale/move/generate these new vertices to be in the right place. Edited by Vangoule

##### Share on other sites

Ok, I see what you mean.  When I get home tonight I can put up the function that creates subpatches.  For now, lets separate geometry from topology.  The topology is how vertices are connected.  The topology of every single patch is the same.  It is a rectangular array of 33x33 vertices, which have been pieced together in 32x32 quads.  In GL , the index buffer defines topology because it says which points go together to make triangles.  It doesn't matter that the patches have been scaled, rotated distorted, whatever, the topology is still just a grid like 'graph paper.'  By setting a fixed patch size, we don't have to worry about different tesselation levels or anything like that.

Now to the geometry.  To make things easy, each point in the patch is identified by two integer coordinates, 'a' and 'b', both varying from 0-32.  You can think of it as a 33x33 pixel image.  Except each 'pixel' contains float x,y,z coordinates instead of a color. Now what you want to do is three things:

a. Define the source regions.  If I'm cutting a 33x33 pixel image up into quarters, each quarter will be 17x17.  yes, the quarters will overlap their neighbors by 1 pixel.  This is correct, since these 'pixels' in this map are actually points, and we want to share an edge of points with the other quarters so there are no gaps.

b. upscale each 17x17 region to a 33x33 map.  This is easy.  You can see if you go along a row 17 original source points, and average every two adjacent values you get 16 in-between values. 17+16 is 33 again.  You have reused 17 points from the previous generation, and created 16 new points.  You can do this for all 17 source rows.  Now you have a 33x17 map.  Going vertically, there are 16 in-between rows you can average, creating a final map of 33 in height.  I recommend NOT trying to literally reuse the same points (as in same VBO) as the parent, but instead copy the positions into the child's VBO.  1 VBO per patch is reasonable to start with; maybe later adjacent quadtrees can share VBOs or something.

c. Displace.  When averaging and creating in-between values, this is the place to add displacement.

I'll post that bit of code up tonight.

I haven't worked on this in a few weeks, but this thread is inspiring me to find the time to fix the cracks between different detail levels!  Thanks!

##### Share on other sites

There is so much information about quad tree's. Everyone seems to do it differently. I'm currently experimenting in a separate project to try and get atleast one quad. I'll see how it goes. I'm not sure if i have enough information yet. I'll try get it working if not i'm hoping your code will save me from my problems. Thanks a lot for all the help.

##### Share on other sites
Oh and as for Gaps. My noise function takes in a vec3 to calculate the height of a vertex. If a vertex is in the same place it will go to the same height. Therefore i should get no gaps. 'Hopefully' T-Junctions however will still exist. When creating the new patches - How are you 'linearly interpolating' them to scale them back up?

Edit: I assume i just average between two vertices. However, my vertices are in an std::vector<glm::vec3> so it's only 1 dimensional. So i'm trying to get my head round it atm. Though i'm still wondering how to position my patches. I'll upload some pictures when i'm finished. Currently not using indices so i have well over 20 buffers. (Bad FPS incoming) I'll sort that out when it works. Edited by Vangoule

##### Share on other sites

Ok, So basically i kinda got it working... Half way atleast. I make the quads i add my camera as an object and the first subdivide is fine. Howeevr, the next thing that i subdivide leaves a big hole. I'll try upload a pic in a sec.

##### Share on other sites

Ok, So basically i kinda got it working... Half way atleast. I make the quads i add my camera as an object and the first subdivide is fine. However, the next thing that i subdivide leaves a big hole. It's probably due to positioning but i can't seem to find it. The source responsible for the terrain has been attached

Edited by Vangoule

##### Share on other sites

I took a look at your source.  What does 'Object.cpp/hpp' represent?  Since you're storing the vertices inside the quadtree itself, I think you can drop Object completely.  There will be a point where you will want to place things on your quadtree grid (characters, portals into buildings or caves), but you should have your own 'entity' class defined somewhere else.

I notice when you create the child nodes, you are using a fixed offset like center.z+34.  Is it really fixed, shouldn't the next level down use half the offset and so forth?

I also notice you set the center values for the 4 children as only being (center.x, center.z)  +   (0,0), (34,0), (34,34) and (0,34).  Shouldn't some of them be '-'.   None of the children should be 0,0: The center of the child will not be the center of the current node.  Look at this:

-------------

|  ,  |  ,  |

|     |     |

|- - -* - - |

|  ,  |  ,  |

|     |     |

-------------

The * is the center of the current node, and the comma's are the center of the children nodes.

------,-----,

|     |     |

|     |     |

|- - -*,- - ,

|     |     |

|     |     |

-------------

I hope this helps!

Edited by DracoLacertae

##### Share on other sites

The object class is just an object that can be added to the quadtree. In this case it's the camera. This just handles the subdividing depending on where it is. I could probably get rid of this when i switch to distance. As for the centers, I'm rendering using normal triangles and the rendering starts at the top left. So my 'centers' really ended up as corners. The problem is i don't know what numbers to be using in the positions. Through trial and error i got the NW quad to work properly but none of the others seem to.

##### Share on other sites

What is the significance of the value 34 you're using?

##### Share on other sites
The value was found by trial and error but it's the point at which the two planes overlayer eachother. Though the value came to be that number because of the scaling and things. I changed the order in which i scale and translate everything and now the value is 168 which is the equivelant of the 'right' value from 0. Here's the source code now. I'll try and upload the actual application aswell. However, i'm not sure if it'll run on your machine. This should allow you too see what's happening. The planes have different height values so you can tell them apart. Moving the camera on to a plane should subdivide it once. It can be downloaded here: http://www.mediafire.com/?lk4zn5lmga3h99c Edited by Vangoule

##### Share on other sites

Your zip is missing glew32.dll.  I tried using my local copy, but it crashed immediately; I think I might have a different version.  I'm downloading your source right now.

##### Share on other sites

Hmm, Here's my version of glew. However, the program was compiled on Windows 8 and a lot of the DLL's are from Windows 8 so it may be that which is crashing it.

##### Share on other sites

It still hates my computer, but I'm also running Windows 8.  Do you have visual C project files or something?

##### Share on other sites

I didn't give the whole project because the size is rather big but if it'll allow you to compile the whole thing too see it there's no problem. It was going to be OpenSource anyway. The problem is it relies on several libraries such as DevIL and SFML 2.0

Edited by Vangoule

##### Share on other sites

I still can't get it to build on my PC.  I tried remaking project files for my version of VS, but I think I might as well upgrade to 2012 express sometime soon anyway.  Is there any chance you can post a video of it running?  Sorry if I'm not being very helpful, I'm getting cramped for time with other things.

##### Share on other sites

I'll try get a video up soon. I've not been working on the project for a little while over frustration. This positioning stuff is really annoying. It should be simple but there's something screwing it up somewhere.

##### Share on other sites

have you considered using the OpenGL 4.0 Tessallator?

• 11
• 9
• 11
• 9
• 12
• ### Similar Content

• By EddieK
Hello. I'm trying to make an android game and I have come across a problem. I want to draw different map layers at different Z depths so that some of the tiles are drawn above the player while others are drawn under him. But there's an issue where the pixels with alpha drawn above the player. This is the code i'm using:
int setup(){ GLES20.glEnable(GLES20.GL_DEPTH_TEST); GLES20.glEnable(GL10.GL_ALPHA_TEST); GLES20.glEnable(GLES20.GL_TEXTURE_2D); } int render(){ GLES20.glClearColor(0, 0, 0, 0); GLES20.glClear(GLES20.GL_ALPHA_BITS); GLES20.glClear(GLES20.GL_COLOR_BUFFER_BIT); GLES20.glClear(GLES20.GL_DEPTH_BUFFER_BIT); GLES20.glBlendFunc(GLES20.GL_ONE, GL10.GL_ONE_MINUS_SRC_ALPHA); // do the binding of textures and drawing vertices } My vertex shader:
uniform mat4 MVPMatrix; // model-view-projection matrix uniform mat4 projectionMatrix; attribute vec4 position; attribute vec2 textureCoords; attribute vec4 color; attribute vec3 normal; varying vec4 outColor; varying vec2 outTexCoords; varying vec3 outNormal; void main() { outNormal = normal; outTexCoords = textureCoords; outColor = color; gl_Position = MVPMatrix * position; } My fragment shader:
precision highp float; uniform sampler2D texture; varying vec4 outColor; varying vec2 outTexCoords; varying vec3 outNormal; void main() { vec4 color = texture2D(texture, outTexCoords) * outColor; gl_FragColor = vec4(color.r,color.g,color.b,color.a);//color.a); } I have attached a picture of how it looks. You can see the black squares near the tree. These squares should be transparent as they are in the png image:

Its strange that in this picture instead of alpha or just black color it displays the grass texture beneath the player and the tree:

Any ideas on how to fix this?

• This article uses material originally posted on Diligent Graphics web site.
Introduction
Graphics APIs have come a long way from small set of basic commands allowing limited control of configurable stages of early 3D accelerators to very low-level programming interfaces exposing almost every aspect of the underlying graphics hardware. Next-generation APIs, Direct3D12 by Microsoft and Vulkan by Khronos are relatively new and have only started getting widespread adoption and support from hardware vendors, while Direct3D11 and OpenGL are still considered industry standard. New APIs can provide substantial performance and functional improvements, but may not be supported by older hardware. An application targeting wide range of platforms needs to support Direct3D11 and OpenGL. New APIs will not give any advantage when used with old paradigms. It is totally possible to add Direct3D12 support to an existing renderer by implementing Direct3D11 interface through Direct3D12, but this will give zero benefits. Instead, new approaches and rendering architectures that leverage flexibility provided by the next-generation APIs are expected to be developed.
There are at least four APIs (Direct3D11, Direct3D12, OpenGL/GLES, Vulkan, plus Apple's Metal for iOS and osX platforms) that a cross-platform 3D application may need to support. Writing separate code paths for all APIs is clearly not an option for any real-world application and the need for a cross-platform graphics abstraction layer is evident. The following is the list of requirements that I believe such layer needs to satisfy:
Lightweight abstractions: the API should be as close to the underlying native APIs as possible to allow an application leverage all available low-level functionality. In many cases this requirement is difficult to achieve because specific features exposed by different APIs may vary considerably. Low performance overhead: the abstraction layer needs to be efficient from performance point of view. If it introduces considerable amount of overhead, there is no point in using it. Convenience: the API needs to be convenient to use. It needs to assist developers in achieving their goals not limiting their control of the graphics hardware. Multithreading: ability to efficiently parallelize work is in the core of Direct3D12 and Vulkan and one of the main selling points of the new APIs. Support for multithreading in a cross-platform layer is a must. Extensibility: no matter how well the API is designed, it still introduces some level of abstraction. In some cases the most efficient way to implement certain functionality is to directly use native API. The abstraction layer needs to provide seamless interoperability with the underlying native APIs to provide a way for the app to add features that may be missing. Diligent Engine is designed to solve these problems. Its main goal is to take advantages of the next-generation APIs such as Direct3D12 and Vulkan, but at the same time provide support for older platforms via Direct3D11, OpenGL and OpenGLES. Diligent Engine exposes common C++ front-end for all supported platforms and provides interoperability with underlying native APIs. It also supports integration with Unity and is designed to be used as graphics subsystem in a standalone game engine, Unity native plugin or any other 3D application. Full source code is available for download at GitHub and is free to use.
Overview
Diligent Engine API takes some features from Direct3D11 and Direct3D12 as well as introduces new concepts to hide certain platform-specific details and make the system easy to use. It contains the following main components:
Render device (IRenderDevice  interface) is responsible for creating all other objects (textures, buffers, shaders, pipeline states, etc.).
Device context (IDeviceContext interface) is the main interface for recording rendering commands. Similar to Direct3D11, there are immediate context and deferred contexts (which in Direct3D11 implementation map directly to the corresponding context types). Immediate context combines command queue and command list recording functionality. It records commands and submits the command list for execution when it contains sufficient number of commands. Deferred contexts are designed to only record command lists that can be submitted for execution through the immediate context.
An alternative way to design the API would be to expose command queue and command lists directly. This approach however does not map well to Direct3D11 and OpenGL. Besides, some functionality (such as dynamic descriptor allocation) can be much more efficiently implemented when it is known that a command list is recorded by a certain deferred context from some thread.
The approach taken in the engine does not limit scalability as the application is expected to create one deferred context per thread, and internally every deferred context records a command list in lock-free fashion. At the same time this approach maps well to older APIs.
In current implementation, only one immediate context that uses default graphics command queue is created. To support multiple GPUs or multiple command queue types (compute, copy, etc.), it is natural to have one immediate contexts per queue. Cross-context synchronization utilities will be necessary.
Swap Chain (ISwapChain interface). Swap chain interface represents a chain of back buffers and is responsible for showing the final rendered image on the screen.
Render device, device contexts and swap chain are created during the engine initialization.
Resources (ITexture and IBuffer interfaces). There are two types of resources - textures and buffers. There are many different texture types (2D textures, 3D textures, texture array, cubmepas, etc.) that can all be represented by ITexture interface.
Resources Views (ITextureView and IBufferView interfaces). While textures and buffers are mere data containers, texture views and buffer views describe how the data should be interpreted. For instance, a 2D texture can be used as a render target for rendering commands or as a shader resource.
Pipeline State (IPipelineState interface). GPU pipeline contains many configurable stages (depth-stencil, rasterizer and blend states, different shader stage, etc.). Direct3D11 uses coarse-grain objects to set all stage parameters at once (for instance, a rasterizer object encompasses all rasterizer attributes), while OpenGL contains myriad functions to fine-grain control every individual attribute of every stage. Both methods do not map very well to modern graphics hardware that combines all states into one monolithic state under the hood. Direct3D12 directly exposes pipeline state object in the API, and Diligent Engine uses the same approach.
Shader Resource Binding (IShaderResourceBinding interface). Shaders are programs that run on the GPU. Shaders may access various resources (textures and buffers), and setting correspondence between shader variables and actual resources is called resource binding. Resource binding implementation varies considerably between different API. Diligent Engine introduces a new object called shader resource binding that encompasses all resources needed by all shaders in a certain pipeline state.
API Basics
Creating Resources
Device resources are created by the render device. The two main resource types are buffers, which represent linear memory, and textures, which use memory layouts optimized for fast filtering. Graphics APIs usually have a native object that represents linear buffer. Diligent Engine uses IBuffer interface as an abstraction for a native buffer. To create a buffer, one needs to populate BufferDesc structure and call IRenderDevice::CreateBuffer() method as in the following example:
BufferDesc BuffDesc; BufferDesc.Name = "Uniform buffer"; BuffDesc.BindFlags = BIND_UNIFORM_BUFFER; BuffDesc.Usage = USAGE_DYNAMIC; BuffDesc.uiSizeInBytes = sizeof(ShaderConstants); BuffDesc.CPUAccessFlags = CPU_ACCESS_WRITE; m_pDevice->CreateBuffer( BuffDesc, BufferData(), &m_pConstantBuffer ); While there is usually just one buffer object, different APIs use very different approaches to represent textures. For instance, in Direct3D11, there are ID3D11Texture1D, ID3D11Texture2D, and ID3D11Texture3D objects. In OpenGL, there is individual object for every texture dimension (1D, 2D, 3D, Cube), which may be a texture array, which may also be multisampled (i.e. GL_TEXTURE_2D_MULTISAMPLE_ARRAY). As a result there are nine different GL texture types that Diligent Engine may create under the hood. In Direct3D12, there is only one resource interface. Diligent Engine hides all these details in ITexture interface. There is only one  IRenderDevice::CreateTexture() method that is capable of creating all texture types. Dimension, format, array size and all other parameters are specified by the members of the TextureDesc structure:
TextureDesc TexDesc; TexDesc.Name = "My texture 2D"; TexDesc.Type = TEXTURE_TYPE_2D; TexDesc.Width = 1024; TexDesc.Height = 1024; TexDesc.Format = TEX_FORMAT_RGBA8_UNORM; TexDesc.Usage = USAGE_DEFAULT; TexDesc.BindFlags = BIND_SHADER_RESOURCE | BIND_RENDER_TARGET | BIND_UNORDERED_ACCESS; TexDesc.Name = "Sample 2D Texture"; m_pRenderDevice->CreateTexture( TexDesc, TextureData(), &m_pTestTex ); If native API supports multithreaded resource creation, textures and buffers can be created by multiple threads simultaneously.
Interoperability with native API provides access to the native buffer/texture objects and also allows creating Diligent Engine objects from native handles. It allows applications seamlessly integrate native API-specific code with Diligent Engine.
Next-generation APIs allow fine level-control over how resources are allocated. Diligent Engine does not currently expose this functionality, but it can be added by implementing IResourceAllocator interface that encapsulates specifics of resource allocation and providing this interface to CreateBuffer() or CreateTexture() methods. If null is provided, default allocator should be used.
Initializing the Pipeline State
As it was mentioned earlier, Diligent Engine follows next-gen APIs to configure the graphics/compute pipeline. One big Pipelines State Object (PSO) encompasses all required states (all shader stages, input layout description, depth stencil, rasterizer and blend state descriptions etc.). This approach maps directly to Direct3D12/Vulkan, but is also beneficial for older APIs as it eliminates pipeline misconfiguration errors. With many individual calls tweaking various GPU pipeline settings it is very easy to forget to set one of the states or assume the stage is already properly configured when in fact it is not. Using pipeline state object helps avoid these problems as all stages are configured at once.
While in earlier APIs shaders were bound separately, in the next-generation APIs as well as in Diligent Engine shaders are part of the pipeline state object. The biggest challenge when authoring shaders is that Direct3D and OpenGL/Vulkan use different shader languages (while Apple uses yet another language in their Metal API). Maintaining two versions of every shader is not an option for real applications and Diligent Engine implements shader source code converter that allows shaders authored in HLSL to be translated to GLSL. To create a shader, one needs to populate ShaderCreationAttribs structure. SourceLanguage member of this structure tells the system which language the shader is authored in:
When sampling a texture in a shader, the texture sampler was traditionally specified as separate object that was bound to the pipeline at run time or set as part of the texture object itself. However, in most cases it is known beforehand what kind of sampler will be used in the shader. Next-generation APIs expose new type of sampler called static sampler that can be initialized directly in the pipeline state. Diligent Engine exposes this functionality: when creating a shader, textures can be assigned static samplers. If static sampler is assigned, it will always be used instead of the one initialized in the texture shader resource view. To initialize static samplers, prepare an array of StaticSamplerDesc structures and initialize StaticSamplers and NumStaticSamplers members. Static samplers are more efficient and it is highly recommended to use them whenever possible. On older APIs, static samplers are emulated via generic sampler objects.
The following is an example of shader initialization:
Creating the Pipeline State Object
After all required shaders are created, the rest of the fields of the PipelineStateDesc structure provide depth-stencil, rasterizer, and blend state descriptions, the number and format of render targets, input layout format, etc. For instance, rasterizer state can be described as follows:
PipelineStateDesc PSODesc; RasterizerStateDesc &RasterizerDesc = PSODesc.GraphicsPipeline.RasterizerDesc; RasterizerDesc.FillMode = FILL_MODE_SOLID; RasterizerDesc.CullMode = CULL_MODE_NONE; RasterizerDesc.FrontCounterClockwise = True; RasterizerDesc.ScissorEnable = True; RasterizerDesc.AntialiasedLineEnable = False; Depth-stencil and blend states are defined in a similar fashion.
Another important thing that pipeline state object encompasses is the input layout description that defines how inputs to the vertex shader, which is the very first shader stage, should be read from the memory. Input layout may define several vertex streams that contain values of different formats and sizes:
// Define input layout InputLayoutDesc &Layout = PSODesc.GraphicsPipeline.InputLayout; LayoutElement TextLayoutElems[] = {     LayoutElement( 0, 0, 3, VT_FLOAT32, False ),     LayoutElement( 1, 0, 4, VT_UINT8, True ),     LayoutElement( 2, 0, 2, VT_FLOAT32, False ), }; Layout.LayoutElements = TextLayoutElems; Layout.NumElements = _countof( TextLayoutElems ); Finally, pipeline state defines primitive topology type. When all required members are initialized, a pipeline state object can be created by IRenderDevice::CreatePipelineState() method:
// Define shader and primitive topology PSODesc.GraphicsPipeline.PrimitiveTopologyType = PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE; PSODesc.GraphicsPipeline.pVS = pVertexShader; PSODesc.GraphicsPipeline.pPS = pPixelShader; PSODesc.Name = "My pipeline state"; m_pDev->CreatePipelineState(PSODesc, &m_pPSO); When PSO object is bound to the pipeline, the engine invokes all API-specific commands to set all states specified by the object. In case of Direct3D12 this maps directly to setting the D3D12 PSO object. In case of Direct3D11, this involves setting individual state objects (such as rasterizer and blend states), shaders, input layout etc. In case of OpenGL, this requires a number of fine-grain state tweaking calls. Diligent Engine keeps track of currently bound states and only calls functions to update these states that have actually changed.
Direct3D11 and OpenGL utilize fine-grain resource binding models, where an application binds individual buffers and textures to certain shader or program resource binding slots. Direct3D12 uses a very different approach, where resource descriptors are grouped into tables, and an application can bind all resources in the table at once by setting the table in the command list. Resource binding model in Diligent Engine is designed to leverage this new method. It introduces a new object called shader resource binding that encapsulates all resource bindings required for all shaders in a certain pipeline state. It also introduces the classification of shader variables based on the frequency of expected change that helps the engine group them into tables under the hood:
Static variables (SHADER_VARIABLE_TYPE_STATIC) are variables that are expected to be set only once. They may not be changed once a resource is bound to the variable. Such variables are intended to hold global constants such as camera attributes or global light attributes constant buffers. Mutable variables (SHADER_VARIABLE_TYPE_MUTABLE) define resources that are expected to change on a per-material frequency. Examples may include diffuse textures, normal maps etc. Dynamic variables (SHADER_VARIABLE_TYPE_DYNAMIC) are expected to change frequently and randomly. Shader variable type must be specified during shader creation by populating an array of ShaderVariableDesc structures and initializing ShaderCreationAttribs::Desc::VariableDesc and ShaderCreationAttribs::Desc::NumVariables members (see example of shader creation above).
Static variables cannot be changed once a resource is bound to the variable. They are bound directly to the shader object. For instance, a shadow map texture is not expected to change after it is created, so it can be bound directly to the shader:
m_pPSO->CreateShaderResourceBinding(&m_pSRB); Note that an SRB is only compatible with the pipeline state it was created from. SRB object inherits all static bindings from shaders in the pipeline, but is not allowed to change them.
Mutable resources can only be set once for every instance of a shader resource binding. Such resources are intended to define specific material properties. For instance, a diffuse texture for a specific material is not expected to change once the material is defined and can be set right after the SRB object has been created:
m_pSRB->GetVariable(SHADER_TYPE_PIXEL, "tex2DDiffuse")->Set(pDiffuseTexSRV); In some cases it is necessary to bind a new resource to a variable every time a draw command is invoked. Such variables should be labeled as dynamic, which will allow setting them multiple times through the same SRB object:
m_pSRB->GetVariable(SHADER_TYPE_VERTEX, "cbRandomAttribs")->Set(pRandomAttrsCB); Under the hood, the engine pre-allocates descriptor tables for static and mutable resources when an SRB objcet is created. Space for dynamic resources is dynamically allocated at run time. Static and mutable resources are thus more efficient and should be used whenever possible.
As you can see, Diligent Engine does not expose low-level details of how resources are bound to shader variables. One reason for this is that these details are very different for various APIs. The other reason is that using low-level binding methods is extremely error-prone: it is very easy to forget to bind some resource, or bind incorrect resource such as bind a buffer to the variable that is in fact a texture, especially during shader development when everything changes fast. Diligent Engine instead relies on shader reflection system to automatically query the list of all shader variables. Grouping variables based on three types mentioned above allows the engine to create optimized layout and take heavy lifting of matching resources to API-specific resource location, register or descriptor in the table.
This post gives more details about the resource binding model in Diligent Engine.
Setting the Pipeline State and Committing Shader Resources
Before any draw or compute command can be invoked, the pipeline state needs to be bound to the context:
m_pContext->SetPipelineState(m_pPSO); Under the hood, the engine sets the internal PSO object in the command list or calls all the required native API functions to properly configure all pipeline stages.
The next step is to bind all required shader resources to the GPU pipeline, which is accomplished by IDeviceContext::CommitShaderResources() method:
m_pContext->CommitShaderResources(m_pSRB, COMMIT_SHADER_RESOURCES_FLAG_TRANSITION_RESOURCES); The method takes a pointer to the shader resource binding object and makes all resources the object holds available for the shaders. In the case of D3D12, this only requires setting appropriate descriptor tables in the command list. For older APIs, this typically requires setting all resources individually.
Next-generation APIs require the application to track the state of every resource and explicitly inform the system about all state transitions. For instance, if a texture was used as render target before, while the next draw command is going to use it as shader resource, a transition barrier needs to be executed. Diligent Engine does the heavy lifting of state tracking.  When CommitShaderResources() method is called with COMMIT_SHADER_RESOURCES_FLAG_TRANSITION_RESOURCES flag, the engine commits and transitions resources to correct states at the same time. Note that transitioning resources does introduce some overhead. The engine tracks state of every resource and it will not issue the barrier if the state is already correct. But checking resource state is an overhead that can sometimes be avoided. The engine provides IDeviceContext::TransitionShaderResources() method that only transitions resources:
m_pContext->TransitionShaderResources(m_pPSO, m_pSRB); In some scenarios it is more efficient to transition resources once and then only commit them.
Invoking Draw Command
The final step is to set states that are not part of the PSO, such as render targets, vertex and index buffers. Diligent Engine uses Direct3D11-syle API that is translated to other native API calls under the hood:
ITextureView *pRTVs[] = {m_pRTV}; m_pContext->SetRenderTargets(_countof( pRTVs ), pRTVs, m_pDSV); // Clear render target and depth buffer const float zero[4] = {0, 0, 0, 0}; m_pContext->ClearRenderTarget(nullptr, zero); m_pContext->ClearDepthStencil(nullptr, CLEAR_DEPTH_FLAG, 1.f); // Set vertex and index buffers IBuffer *buffer[] = {m_pVertexBuffer}; Uint32 offsets[] = {0}; Uint32 strides[] = {sizeof(MyVertex)}; m_pContext->SetVertexBuffers(0, 1, buffer, strides, offsets, SET_VERTEX_BUFFERS_FLAG_RESET); m_pContext->SetIndexBuffer(m_pIndexBuffer, 0); Different native APIs use various set of function to execute draw commands depending on command details (if the command is indexed, instanced or both, what offsets in the source buffers are used etc.). For instance, there are 5 draw commands in Direct3D11 and more than 9 commands in OpenGL with something like glDrawElementsInstancedBaseVertexBaseInstance not uncommon. Diligent Engine hides all details with single IDeviceContext::Draw() method that takes takes DrawAttribs structure as an argument. The structure members define all attributes required to perform the command (primitive topology, number of vertices or indices, if draw call is indexed or not, if draw call is instanced or not, if draw call is indirect or not, etc.). For example:
DrawAttribs attrs; attrs.IsIndexed = true; attrs.IndexType = VT_UINT16; attrs.NumIndices = 36; attrs.Topology = PRIMITIVE_TOPOLOGY_TRIANGLE_LIST; pContext->Draw(attrs); For compute commands, there is IDeviceContext::DispatchCompute() method that takes DispatchComputeAttribs structure that defines compute grid dimension.
Source Code
Full engine source code is available on GitHub and is free to use. The repository contains two samples, asteroids performance benchmark and example Unity project that uses Diligent Engine in native plugin.
AntTweakBar sample is Diligent Engine’s “Hello World” example.

Atmospheric scattering sample is a more advanced example. It demonstrates how Diligent Engine can be used to implement various rendering tasks: loading textures from files, using complex shaders, rendering to multiple render targets, using compute shaders and unordered access views, etc.

Asteroids performance benchmark is based on this demo developed by Intel. It renders 50,000 unique textured asteroids and allows comparing performance of Direct3D11 and Direct3D12 implementations. Every asteroid is a combination of one of 1000 unique meshes and one of 10 unique textures.

Finally, there is an example project that shows how Diligent Engine can be integrated with Unity.

Future Work
The engine is under active development. It currently supports Windows desktop, Universal Windows and Android platforms. Direct3D11, Direct3D12, OpenGL/GLES backends are now feature complete. Vulkan backend is coming next, and support for more platforms is planned.
• By reenigne
For those that don't know me. I am the individual who's two videos are listed here under setup for https://wiki.libsdl.org/Tutorials
I also run grhmedia.com where I host the projects and code for the tutorials I have online.
Recently, I received a notice from youtube they will be implementing their new policy in protecting video content as of which I won't be monetized till I meat there required number of viewers and views each month.

Frankly, I'm pretty sick of youtube. I put up a video and someone else learns from it and puts up another video and because of the way youtube does their placement they end up with more views.
Even guys that clearly post false information such as one individual who said GLEW 2.0 was broken because he didn't know how to compile it. He in short didn't know how to modify the script he used because he didn't understand make files and how the requirements of the compiler and library changes needed some different flags.

At the end of the month when they implement this I will take down the content and host on my own server purely and it will be a paid system and or patreon.

I get my videos may be a bit dry, I generally figure people are there to learn how to do something and I rather not waste their time.
I used to also help people for free even those coming from the other videos. That won't be the case any more. I used to just take anyone emails and work with them my email is posted on the site.

I don't expect to get the required number of subscribers in that time or increased views. Even if I did well it wouldn't take care of each reoccurring month.
I figure this is simpler and I don't plan on putting some sort of exorbitant fee for a monthly subscription or the like.
I was thinking on the lines of a few dollars 1,2, and 3 and the larger subscription gets you assistance with the content in the tutorials if needed that month.
Maybe another fee if it is related but not directly in the content.
The fees would serve to cut down on the number of people who ask for help and maybe encourage some of the people to actually pay attention to what is said rather than do their own thing. That actually turns out to be 90% of the issues. I spent 6 hours helping one individual last week I must have asked him 20 times did you do exactly like I said in the video even pointed directly to the section. When he finally sent me a copy of the what he entered I knew then and there he had not. I circled it and I pointed out that wasn't what I said to do in the video. I didn't tell him what was wrong and how I knew that way he would go back and actually follow what it said to do. He then reported it worked. Yea, no kidding following directions works. But hey isn't alone and well its part of the learning process.

So the point of this isn't to be a gripe session. I'm just looking for a bit of feed back. Do you think the fees are unreasonable?
Should I keep the youtube channel and do just the fees with patreon or do you think locking the content to my site and require a subscription is an idea.

I'm just looking at the fact it is unrealistic to think youtube/google will actually get stuff right or that youtube viewers will actually bother to start looking for more accurate videos.

• i got error 1282 in my code.