Hi Guys,

Previously when I used DX9c I used to draw to a surface of a preset size (usually screen dimensions 1600x900) and when I went into windowed mode I would simply draw the surface to the window using...

(FAILED(d3dDevice->StretchRect(surfaceApplication,NULL,surfaceBackbuffer,&DstRect,D3DTEXF_LINEAR )))

So, this way the scene would resize gracefully and I could still move objects as if they were still in full screen dimensions.

Is there a way to do similar in DirectX 11?

Thanks in advance

Not with a simple function call, no. You render to a texture and then draw a full screen quad (or whatever dimension you like) and a pixel shader simply sampling said texture. The texture needs to have both D3D11_BIND_SHADER_RESOURCE and D3D11_BIND_RENDER_TARGET for BindFlags and you create ID3D11ShaderResourceView and ID3D11RenderTargetView to use for this scenario. Make sure you unbind one view first before using the other, the pipeline disallows simultanous read/write and will nullify such an attempt.

On the other hand, you can achieve the same more directly with a proper camera/viewport/coordinate system which accounts for the output dimension.

Thanks for the reply unbird

Makes sense what you are saying.

The camera system suggestion is an awesome one, I didn't even think of adjusting that.

I allready have an OrthoGraphic camera setup in my project, I tried adjusting the camera's window width and height, and what do you know 'it just works!'.

I was ready to have a hell of a time to get this working and it was easy in the end. So, it just goes to show some decent advance planning does pay off sometimes ;)

[edit]
This will allow for all sorts of awesome zoom effects also! You have made my day

Indeed. A flexible 2D camera is not only a good exercise in transformations, but since it results in changing only one matrix (usually) for the shader - which you do need anyway - it is practically free on the GPU side. Drop those magic numbers!

Glad you got it already