Drizzle

I like looking at the frequency spectrum of sounds.
Sound is generally created with a waveform, but human ears and brains convert it back to their frequency components (with some extra postprocessing). Seeing a sound displayed in the same way that I experience it with my ears is really satisfying to me.

I couldn't easily find a program that displayed what I wanted to see, so I made one myself! I took Waterfall, and completely replaced every single component of it (aside from some of the FFT code, there is nothing left of Waterfall in Drizzle).

Simply drop in .wav files from your file explorer, and you can play, pause, adjust volume, and skip between tracks. Works on both Linux and Windows. There is a GitHub page as well, but it needs some work.

Technical Details

For this, I created my own GUI library, Soggy, which I've used for many things since. The FFT has a windowing function (so less artifacting than in Waterfall), and I also wrote my own .wav file parser (adapted from the one I wrote in Vanadis). The sound playback is done with MiniAudio.

Drizzle chooses a framerate based on the sample rate and length of a song, and does FFT's during playback, continually drawing them to the screen while scrolling down. The FFT is sampled in such a way that it will never fall behind, even on an extremely underpowered machine. All past FFT's are not stored anywhere else other than rendered to the screenbuffer, so currently, there's no scrubbing through the song.

Future plans

I started realizing later in development that I want Drizzle to be more of a music player than a visualizer, but the current architecture is not aligned with that. So, I should build in support for more than just .wav files, add a more elaborate file exploring/directory playing function, and add scrubbing. With that last one comes the fact that I need to save all FFT'd sections somehow, and I'm currently considering doing a wavelet transform, and rendering that to a texture, that a user could then scroll through at runtime. A user might even be able to export the full song's frequency data.