Apologies for the super long delay in responding.

I spent a lot of time working with some of the advice here and managed to improve things a little. There is still some poor quality artifacts left, but it turned out the rendertarget was slightly larger than the window.

Thanks for the time responding.

sp.

Get a GPU debugging apps, I believe PIX is available for DX11? This will allow you to spectate GPU memory such as frame/render buffers etc.

What artifacts are you still experiencing after render target is not down/up sampled when presented?

Hi JonnyCode-

I think I am ok on this - I think I am expecting to see something that requires other things like better graphics processing or something.

I render the wireframe and I have antialiasing on, but the lines arent quite like what I was expecting - I am sure this is a misunderstanding on my part. The sizes of the viewport/rendertarget are legit. I have decided to move on and come back to this later.

Thanks for your response,

sp.

The issue you are seeing is a 100% correct when rasterization is in play, and not entirely related to antialiasing at least not in the sense you think. What you are seeing is depth-related aliasing/ z-fighting. Unless you are rendering your line using special logic such as with depth disabled or some other means, the fragments being generated for each line base on a given viewpoint will have depth values that will pass/fail the depth test in relation to the fragments of the underlying triangle. A search for wireframe rendering or solid wireframe rendering should give you ample resources to explain further what is going on as well as proposed solution.

Check that the window resolution is the actual client size of the window. Typically, with Win32 at least, when you create a window, say at 1280x720, that includes the window chrome (the title bar, close minimize buttons, etc.). So when you render to the 720p buffer in DirectX, it is really mapping it to like a 1272x696 visible window size (numbers made up, but you get the idea). You have to account for the chrome when you do your window creation, in Win32 or whatever library you are using. I'm pretty sure that is the problem.

@cgrant Thanks for the information. A few terms there I havent come across so for, so this is helpful.

sp.

@SibylSystem Yeh, that was part of my problem originally - Ihad to adjust the window rect to line up with what I was expecting.

Thanks,

sp.