Shadow Mapping with DirectX 11

Started by
1 comment, last by Vexal 11 years ago

I am having a hard time finding a tutorial for implementing shadow mapping with non-deprecated DirectX 11.

I've found two guides:

http://takinginitiative.net/2011/05/15/directx10-tutorial-10-shadow-mapping/

and

http://www.rastertek.com/dx11tut40.html

Both of these guides seem to work completely differently, and I can't figure out why. So far, I've just been trying to get it to correctly render the depth information to the texture, and then make the pixel shader do something with that texture just to convince me the first pass is doing anything at all.

In the first guide, the source download doesn't work. Also, apparently this person is uses techniques or effects, which are deprecated. I can't figure out how to translate this part into modern DirectX:

pEffect->GetVariableByName("shadowMap")->AsShaderResource()->SetResource( pShadowMapSRView );
//-------------------------------------------------
//RENDER SCENE
//-------------------------------------------------
//unbind shadow map as SRV and call apply on scene rendering technique
pEffect->GetVariableByName("shadowMap")->AsShaderResource()->SetResource( 0 );
pShadowMapTechnique->GetPassByIndex(0)->Apply( 0 );

The two tutorials seem to have completely different methods for setting up the render targets, and I have no idea which to actually use. The second tutorial I linked (rastertek) seems to not have most of the stuff the first tutorial has.


The guide that comes with the MSDN DirectX SDK uses deprecated effects API commands as well.

I've found many articles on state-of-the-art shadowing techniques, but I just want it to do *something* before I move onto more complicated stuff.


Advertisement

If you want to write it from scratch, then here are the basic steps:

  1. Do normal application initialization (create device and output swap chain).
  2. Create shadow map render target (probably good to start out with the R32 DXGI format texture)
  3. For frame rendering, do the shadow pass and write only depth into the shadow map. The view/projection matrices are determined by the location and nature of your light.
  4. Then bind your output render target, and bind the shadow map as a shader resource view to your pixel shader. Sample it and do the depth comparison to determine if it is in shadow or not.

If you understand these general steps, then try to implement them on your own instead of relying on the tutorials. In general, I find that most tutorials like that are too superficial to show you really how the algorithm works. Once you implement the thing yourself, you will have lots more experience and understand it better too.

And of course, if any of those steps don't work for you, you can always ask a specific question here to get some further guidance.

I understand the steps and the math. I don't know the exact DirectX 11 commands to use.

I have a working DirectX 11 game engine I've built completely from scratch, but this is my first time trying to render to do something like this. I don't see the point of figuring out this part myself, because that equates to just guessing the DirectX API.

I actually finally just got it to render the shadow map directly as a texture to the ground, just to see if the shadow map was even being built in the first place, which was what my goal was in the first place, and which was what I was stumped on because NONE of the tutorials are updated with modern DirectX commands.

Just throwing a random crappy model from 3DS max I threw together in five minutes to test it gives this, http://i.imgur.com/ltbdWNE.jpg

which is what I wanted.

I'm not sure what exactly the issue was, but something I did fixed it.

This topic is closed to new replies.

Advertisement