• Announcements

    • khawk

      Download the Game Design and Indie Game Marketing Freebook   07/19/17

      GameDev.net and CRC Press have teamed up to bring a free ebook of content curated from top titles published by CRC Press. The freebook, Practices of Game Design & Indie Game Marketing, includes chapters from The Art of Game Design: A Book of Lenses, A Practical Guide to Indie Game Marketing, and An Architectural Approach to Level Design. The GameDev.net FreeBook is relevant to game designers, developers, and those interested in learning more about the challenges in game development. We know game development can be a tough discipline and business, so we picked several chapters from CRC Press titles that we thought would be of interest to you, the GameDev.net audience, in your journey to design, develop, and market your next game. The free ebook is available through CRC Press by clicking here. The Curated Books The Art of Game Design: A Book of Lenses, Second Edition, by Jesse Schell Presents 100+ sets of questions, or different lenses, for viewing a game’s design, encompassing diverse fields such as psychology, architecture, music, film, software engineering, theme park design, mathematics, anthropology, and more. Written by one of the world's top game designers, this book describes the deepest and most fundamental principles of game design, demonstrating how tactics used in board, card, and athletic games also work in video games. It provides practical instruction on creating world-class games that will be played again and again. View it here. A Practical Guide to Indie Game Marketing, by Joel Dreskin Marketing is an essential but too frequently overlooked or minimized component of the release plan for indie games. A Practical Guide to Indie Game Marketing provides you with the tools needed to build visibility and sell your indie games. With special focus on those developers with small budgets and limited staff and resources, this book is packed with tangible recommendations and techniques that you can put to use immediately. As a seasoned professional of the indie game arena, author Joel Dreskin gives you insight into practical, real-world experiences of marketing numerous successful games and also provides stories of the failures. View it here. An Architectural Approach to Level Design This is one of the first books to integrate architectural and spatial design theory with the field of level design. The book presents architectural techniques and theories for level designers to use in their own work. It connects architecture and level design in different ways that address the practical elements of how designers construct space and the experiential elements of how and why humans interact with this space. Throughout the text, readers learn skills for spatial layout, evoking emotion through gamespaces, and creating better levels through architectural theory. View it here. Learn more and download the ebook by clicking here. Did you know? GameDev.net and CRC Press also recently teamed up to bring GDNet+ Members up to a 20% discount on all CRC Press books. Learn more about this and other benefits here.
  • entries
    316
  • comments
    485
  • views
    321044

Volume Rendering

Sign in to follow this  
Followers 0
Jason Z

1630 views

Lately I have been preparing Hieroglyph 3 for the switch to VS2012, and also expanding it to multiple platforms as well (i.e. Win7 desktop, Win8 desktop, Windows Store Apps, etc...). These are important tasks, but aren't very exciting at all... So as a side project, I implemented the first component of a new algorithm that I have been investigating - volume rendering. I have found the process to be really enjoyable, so I wanted to share what I have accomplished so far.

Volume Rendering
When I say volume rendering, what I am referring to is a small scale ray tracing through a volume data structure. More specifically, since I am using Direct3D 11, I trace rays through a Texture3D resource on the GPU via the Pixel Shader. The basic setup is as follows:

  1. Load the desired volume data into a 3D Texture.
  2. Render triangle geometry to cover the area that the volume should appear in.
  3. In the pixel shader, use the view position to trace through the 3D texture until your ray intersects something.
  4. Shade that pixel according to the properties you find at that instersection.


In general, this concept is very reminiscent of one of the first complex rendering operations I can remember investigating - Parallax Occlusion Mapping. The difference here is that POM uses a 2D texture that is mapped to the surface of a primitive, while true volume rendering uses a 3D texture that is mapped to the 3D space occupied by the surface primitives. So in many ways the volume rendering is a simpler concept, since you don't have to do a basis calculation at every vertex - but of course at the expense of exponentially more memory consumption.

As with all things in life, the choice of 2D or 3D is a matter of tradeoff...

Implicit Surfaces
There are lots of variants to this algorithm, and you can pretty much tailor it to your specific needs. For example, the volume representation in your 3D texture can be customized to provide a unique data representation. For the algorithm that I am working on, I needed to be able to visualize an implicit surface. If you haven't had the pleasure of working with these yet, the basic concept is that each data point within the 3D texture is assigned a floating point value. These values can range from positive to negative, and the implicit 'surface' of the object is defined as the point where the interpolated values cross the '0' boundary.

In traditional CPU algorithms, this surface is usually extracted using marching cubes or something similar. But in the GPU, we don't want to get a mesh based representation - we just want to determine where each pixel view ray intersects the surface, and then shade it accordingly. This let's us have a constant volume representation, and we only query the data based on our current view. It works out pretty nice, and is surprisingly easy to get working right.

Texture Space
One small complication is that the ray tracing has to be done in a common 'space' as the volume data. This means we either need to convert the volume data from texture space to world space (or even view space), or we could convert the view position to texture space. Since the latter can easily be carried out with a single matrix transformation on the CPU and applied as a constant in the pixel shader, I decided to go with converting to texture space.

As I alluded to earlier, working with 3D data is simpler than a quasi-2D solution like POM. You can use regular transforms to convert your world space view position first to object space (with the inverse of the world transform) and then to texture space. That conversion to texture space will vary depending on how you map your vertices with texture coordinates, but if you choose them appropriately then you may not even need to apply a transform here (i.e. texture space can be equivalent to object space...).

Shading the Surface
After converting the view position to texture space, it is pretty simple to step through the volume and check for a transition. This is basically done by looking for any values that are greater than '0', as shown here (note: snippet is simplified for clarity!):


float3 direction = normalize( input.tex - CameraPositionTS.xyz );

// Perform the sampling along the view vector until you max out the number of iterations
// or you hit the surface.

float stepLength = 1.0f/80.0f;
float3 stepSize = direction * stepLength;
float3 location = input.tex;
float surfaceHit = 0.0f;
float currTSDF = 0.0f;
float prevTSDF = 0.0f;
float3 surfacePosition = input.tex;

for ( int i = 0; i < 80; i++ ) {

prevTSDF = currTSDF;
currTSDF = VolumeTexture.SampleGrad( LinearSampler, location, 1.0f, 1.0f ).x;
if ( currTSDF > 0.0f ) {
float TSDFdelta = currTSDF - prevTSDF;
float TSDFintersect = currTSDF / TSDFdelta;
surfacePosition = location - stepSize * TSDFintersect;

surfaceHit = 1.0f;
break;
}
location = location + stepSize;
}



Once the intersection point is found, we have to calculate a normal vector to use in the lighting calculations. One might think that we could just store a normal vector in the 3D texture, but if you do the math on how much memory it takes to store 4 values in a 3D texture you will quickly want a way to calculate the normal vector smile.png

Fortunately, this isn't too hard to do. Since we are already working in texture space, we can very easily calculate the gradient of our 3D texture value by sampling in each of the dimensional directions (+/- X, Y, and Z), and finding the delta along each axis. If we then normalize the resulting values as a vector, we get a pretty good representation of the surface normal vector. Here is the code that I used, which worked quite well:


// Calculate the gradient based normal vector using central differences
float right = VolumeTexture.SampleGrad( LinearSampler, surfacePosition + float3( 1.0f, 0.0f, 0.0f ) * stepLength, 1.0f, 1.0f ).x;
float left = VolumeTexture.SampleGrad( LinearSampler, surfacePosition + float3( -1.0f, 0.0f, 0.0f ) * stepLength, 1.0f, 1.0f ).x;
float up = VolumeTexture.SampleGrad( LinearSampler, surfacePosition + float3( 0.0f, 1.0f, 0.0f ) * stepLength, 1.0f, 1.0f ).x;
float down = VolumeTexture.SampleGrad( LinearSampler, surfacePosition + float3( 0.0f, -1.0f, 0.0f ) * stepLength, 1.0f, 1.0f ).x;
float front = VolumeTexture.SampleGrad( LinearSampler, surfacePosition + float3( 0.0f, 0.0f, 1.0f ) * stepLength, 1.0f, 1.0f ).x;
float back = VolumeTexture.SampleGrad( LinearSampler, surfacePosition + float3( 0.0f, 0.0f, -1.0f ) * stepLength, 1.0f, 1.0f ).x;

float3 surfaceNormal = normalize( float3( left-right, down-up, back-front ) );


Once the surface normal is available, you can apply your favorite shading model as needed. I chose to use a simple Phong lighting model, similar to what I was using for my immediate mode geometry rendering.

Some Visual Results
After getting the initial algorithm hammered out with simple squares in my volume, I decided that I wanted to really see how the volume looked with a complex shape inside. Instead of using a modelling tool, and trying to import a pre-crafted volume model, I decided to go back to one of my favorites - Perlin noise. For generating something interesting in 3D, there is not many more useful algorithms than Perlin noise, and I haven't used it in a while either so it was a natural choice. I generated the 3D texture to have values ranging from -1 to +1, and then produced the following images which show two cubes rendering volumes:

VolumeRenderer100001.png
VolumeRenderer100002.png
VolumeRenderer100003.png

I am quite happy with how it turned out, and if you are interested to see the whole implementation, then head over the Hieroglyph 3 Codeplex site and grab the latest trunk!

2
Sign in to follow this  
Followers 0


2 Comments


The Volume Rendering stuff looks very interesting (Although I'm not sure it's actually usable in a game, but then I'm guessing you didn't intend it to be).

For me I'm most interested in the port to VS2012. I notice you said "expanding it to multiple platforms as well (i.e. Win7 desktop, Win8 desktop, Windows Store Apps, etc...)".

Is there a difference between Win7 Desktop and Win8 Desktop? Especially since they've just announced that most of the DirectX 11.1 functionality is now included in the latest SP for Win7: [url="http://blogs.msdn.com/b/chuckw/archive/2012/11/14/directx-11-1-and-windows-7.aspx"]http://blogs.msdn.com/b/chuckw/archive/2012/11/14/directx-11-1-and-windows-7.aspx[/url]

Obviously there's a big difference between Desktop and Windows store Apps.

Keep up the great work.

Thanks
Ben
0

Share this comment


Link to comment
You are right on the money regarding the intended application - this is aimed directly at a particular engineering problem that I'll be talking more about in the coming weeks/months, and isn't really intended for use in games. However, you never know when it could come in handy :)

Regarding the Win7/Win8 desktop variants - I wanted to leave the VS2010 projects in place for users who wanted to utilize the older SDK instead of the Win8 SDK. I was referring to this as the Win7 variant, although that isn't exactly a correct name I guess. However, I am now leaning towards just converting over to VS2012 completely and providing support for several different platforms from there - in particular the desktop, WinRT, and WP8. It should be interesting, and it is my next priority for Hieroglyph3. I want to have a modernized solution setup within the next week or so, so that I can jump back into the real development work.

Thanks for the comment, and I'll be writing a journal post about where things end up!
0

Share this comment


Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now