Hi! this is my first post and I want to begin to tell you about my experiences and experiments regard .NET world.
There is a feature that lots of applications and sites miss. I'm talking about the audio spectrum, sometime built by fake bars that don't follow the music. So I developed a mixed assembly using C++ CLI that process audio samples and computes peaks grouping them by frequency range. Through DirectShow I load audio track and I grab 44100 samples per second and I transform them with FFT (I found this good guide) . FFT is heavy calculation to repeat many times per second, thus I chosen C++ CLI to reduce interop, limiting it only to the peaks exposed as a property.
The WPF side contains a custom element named AudioVisualization that 12 times per second loads peaks and shows bars using animations. My sample player contains a simple skin, but obviuosly we can customize bars exploiting WPF capabilities.
Here a video demo:
Reyalp
Full source code is available here.