Reading the "Doom Engine Black Book" has changed my perspectives on software portability. I used to always think "Oh, C is super portable, look at DOOM it runs on everything". In reality, there were a bunch of caveats and technical hurdles that were needed to overcome each and every port. It didn't "just work". The thing that was truly portable? The WADs. The data. It's making me rethink my approach in my own creative software ecosystems. Up to this point, my work has been represented as code and executable programs. These are more brittle than I expected. It turns out, I can't even rely on C compilers to add 2 floating point numbers with consistent behavior. So, I'm trying to think about designs that are more data-oriented and implementation-agnostic.
I've attempted to build an empirical model of the classic loon warble. It's just made from well tuned sine waves and envelope generators.
The first is the actual call I was analyzing. The second is my synthesized attempt. The third is my attempt put through a forest reverb impulse response.
There's a beautiful lilt happening in that initial onset that I'm missing from mine. But, I think it captures the vibe well enough.
More looning around this morning. This time I'm trying to build a crude model around the "cry" of the loon. (Recording: reference, synthesized (wet), synthesized (dry)) The pitch curve of the cry is both jumping and rising. The jumps (which sound like voice cracks) remind me of yodeling. Since my loon call yesterday, I've added a bit more grit to the sine oscillator spectrum. I've also added a bit of ambience noise using filtered white noise and sparse noise for pops.
"Because of concurrent programming errors (also known as race conditions), it sometimes gave its patients radiation doses that were hundreds of times greater than normal, resulting in death or serious injury."
Saw Oppenheimer last night. Opp was great as to be expected. Heimer ended up only making a brief cameo, which was disappointing. I guess they don't get along like they used to.
Sound artist exploring the human voice, simulacrum, and mechanisms for lyricism.These are some of my favorite things: vocal synthesis, generative music, Palestrina, audio DSP, baking, croissants, tea, dogs, grids, retrogaming, computer graphics, powers of 2, 1-bit art/sound.Over the years, I've developed several audio engines and music software ecosystems, and like to talk about them here. I also have a quieter account where I mainly talk about vocal synthesis: @patchlore