Скачать или смотреть Jeff Emtman & Martin Austwick: Neutrinowatch - the podcast that plays itself

Neutrinowatch is a podcast where every episode changes every day. Python code in the background overwrites the audio based on procedural generation and real world events and conditions to create a new version of each episode. Examples include: - An astronomy show which uses the current positions of the stars visible from a different location - A song where the lyrics and instruments change - An episode where two “robots” talk about real-world news headlines - A “Dear John” letter that gets rewritten every day - Monologues that fade in and out of focus - Daily pieces of oblique advice for the creatively stuck Every episode is alive, a new version appearing each day - which presents unique challenges and opportunities for making audio. In our talk, we’ll dive into a couple of episodes, sharing the techniques and lessons we’ve learned as we created the show - and the roles audio effects, ambience, music, and text-to-speech models have in the sound design. How can we make an episode that’s interesting to listen to once? Twice? Ten times? How often have we succeeded in making something replayable? Join us, and help us to answer these questions, and more…