This Week in Sound: Audio Deepfakes vs. Basso Profondo Fakes

A lightly annotated clipping service

These sound-studies highlights of the week originally appeared in the September 27, 2022, issue of the free weekly email newsletter This Week in Sound:

NPR profiles Jim Metzner, who “spent nearly five decades documenting and sharing the sounds of the world, from immersive portraits of American cities to indelible moments with people and wildlife in places as varied as Alaska, Australia, Japan, Greece, Cuba, Nepal and Morocco.” A collection of his work from the 1970s to 2019, amounting to roughly 28,000 items, has been added to the Library of Congress: “Sometimes you hear people say, ‘You know, I captured this sound’ and ‘I captured that sound,’ ” he adds. “Right from the get-go, I never felt that I was capturing anything. I felt like these things were gifts. You receive something extraordinary, the first thing you want to do is say, ‘Oh my God, listen to this! Let me share this with somebody!'” ➔,

The fight against computer-produced audio deepfakes is taking place on computers. One question is whether this is a solution, or another phase in an arms race: “To detect audio deepfakes, we and our research colleagues at the University of Florida have developed a technique that measures the acoustic and fluid dynamic differences between voice samples created organically by human speakers and those generated synthetically by computers.” ➔

There’s audio deepfakes — and then there’s basso profondo fakes: “After 45 years of voicing one of the most iconic characters in cinema history, James Earl Jones has said goodbye to Darth Vader. At 91, the legendary actor recently told Disney he was ‘looking into winding down this particular character.’ That forced the company to ask itself how do you even replace Jones? The answer Disney eventually settled on, with the actor’s consent, involved an AI program.” ➔

“One of the unique safety features built into Apple Watch Ultra is an 86-decibel siren. It uses two different sound patterns to attract help with the alert being heard up to 600 feet away.” ➔

Why, historically, have the stages at Nashville’s Louder Than Life stage “faced toward the city instead of away from it”? Because the event occurs on the site of a demolished hotel, and until this year, “dirt took up space where concert stages and crowds would be.” ➔

“Even the sound of the engine changes the way food tastes. Exposure to the background noise of an aeroplane, which can reach 80-85dB, dulls your sensitivity to salty and sugary flavours, while enhancing your perception of the proteinous fifth taste, umami. This explains the enduring love affair between air passengers and tomato juice, which is ordered as much as beer in flight. If you drink it in the sky, it will taste richer, more savoury, and less acidic.” Note the dissenting voices in the replies to the Twitter thread that helped popularize this particular article. ➔, (Thanks, George Kelly!)

“Welcome to the cult of brown noise, a sometimes hazily-defined category of neutral, dense sound that contains every frequency our ears can detect. Brown noise is like white noise but has a lower, deeper quality. It gained a fervent following over the summer, picking up speed in online A.D.H.D. communities, where people made videos of their reactions to hearing it for the first time.” ➔

Sensors and microphones are being deployed off San Francisco to deter “oceanic roadkill” by steering ships away from whales: “Whale Safe uses three data streams: the buoy listens for and identifies the songs of blue, fin and humpback whales with an algorithm and beams its findings to a satellite; a mathematical model informed by present and past oceanographic and biological data predicts where blue whales are most likely to be; and citizen scientists and trained observers report whale sightings via an app called Whale Alert.” ➔

It’s intriguing when voice AIs add a feature that allows you to access the service without the use of your voice. A new update for Amazon’s Alexa means you “can not only control the assistant exclusively using touch, but pair with a supporting Bluetooth switch or use text-to-speech to have commands spoken for you.” There’s an Apple feature along these lines called Type to Siri. ➔,

Per earlier reporting, Spotify has launched an audiobook service. “It also wants to ‘innovate the format’ so that listeners, authors and publishers benefit more from what the format can offer.” ➔

Bucharest, Romania, fights noise pollution by banning “the nighttime circulation of cars and motorcycles that produce noise above 95 decibels.” I’d like this to happen in my neighborhood. ➔

Leave a Reply

Your email address will not be published. Required fields are marked *