Couple of questions about effect variables: 1) If I write a default value in an HLSL effect, when is that variable set to that value? When the effect is created? Everytime a pass from a technique within that effect is bound? Immediatly after a technique is rendered? (this is what would make sense along with at creation!) 2) Suppose I ...

I've made progress with DXGI after struggling for 2 months now. However, I cannot get DXGI::SwapChain::ResizeTarget to do its job! Calling this method seems to have no effect on my application. I have been trying for months on figuring out a way to tell DXGI to use a given fullscreen resolution when switching from windowed back to ...

The docs clearly state that releasing a swapchain in fullscreen mode will cause an exception. You must switch to windowed mode and then release using SetFullscreenState.

In reading the MSDN about DXGI and DX10, it informs me that I must set the refresh rate to exactly match an enumerated refresh rate or suffer severe performance penalties. Well, the enumerated refresh rates, I assume are for when you are going to a full screen mode. However, even when you create a swapchain for windowed mode, you must still set ...

I have a slew of questions after working through enumerating adapters and display modes. I am working in DX10, but I do not know if that matters. 1) When enumerating display modes you get them by monitor. 1.1) Should I even make it an option to go fullscreen on a different monitor than the primary? 1.1.1) How do you tell which monitor IS ...

Ok, since noone seems to know how to do this, I figure I will share my findings. After a full week of pooring through the docs and asking on several forums to no avail, I have finally devised a way to query a pass for the input layout it expects. So, when I create my renderable object it is going to create one vertex buffer for each vertex ...

I have a need to get a description of what input layout a pass expects. Everything I've seen in examples and in the SDK gives no hint to this. However, advanced techniques are arising where I may have multistreamed data and would like to provide only the array of vertex buffers that a pass requires instead of being inefficient and passing all ...

Another way I thought of was to request it be loaded in R32G32B32A32 format, check that it got loaded in R32G32B32A32, iterate through the data, check each pixel for alpha != 1 That is also pretty computationally expensive, even if it is at init time. Anyone have better ideas?

Is there an easy way to check if a teture includes alpha information? Something cleaner than: if( texDesc.Format == DXGI_FORMAT_R32G32B32A32_TYPELESS ||       texDesc.Format == DXGI_FORMAT_R32G32B32A32_FLOAT    ||   ...

Ok, thanks. I am using D3DX to load the image, which means that I must change the V coords because of how D3DX loads it.