Like kernel or compiler development, audio and DSP development is seen by some as a black art that’s subject to mysterious incantations and arcane rituals.

Everybody knows how to write a for loop and wire up a CRUD app these days, but satisfying a low-latency audio buffer request while decoding media and applying effects feels fraught and often (genuinely) looks like a garble of translated Matlab with countless inscrutable variable names and seemingly random constants peppered everywhere.

From that sense of awe (traditional meaning), it’s only a small leap to start wondering if even “known” textbook truths about things like memcpy matter differently than normal. And then the cultist gossip begins, you get the echo chamber effect on casual forums and meetups, and this is where you end up.