Lossy’s Uncharted Waters
Since this post was published, we added a reverb to Lossy, so it’s now much easier to get the sounds mentioned near the end of this post…
We built Lossy to solve a specific problem: make it easy to recreate the sound of a bad MP3 in realtime, without having to bounce a track to and from an MP3. But imbuing sound with an MP3’s imperfections — the fizzle and sizzle of 64 kbps — was only the beginning.
Spectrograms of guitar chord as heard below. No loss (“lossless”) on the left, extremely lossy on the right.
The very first Lossy prototype had exactly one control: a slider, “Loss,” that went from 0% — audio, untouched — to 100%.1 Now, commercially available MP3 codecs only reach about 25% of the way up that slider. That’s the sound of a truly crappy MP3, and making that sound was the original goal of Lossy.2 But what’s subtle at 25% — a faint memory of downloading early 00’s pop on KaZaA — becomes bizarre at 50%, otherworldly at 75%, and something oceanographic at 100%.
When we mine history for creative purposes, we often crave something closer to parody than reality — exaggeration rather than exactitude. And anything on the “Loss” slider above 50% is just that: a data compression algorithm that never really existed.3
But then we started finding new sounds with no historical precedent, and there was no turning back.
1 “Loss Speed”
Early on in prototyping, we discovered a way to change how quickly the data compression operated, so we added a “Loss Speed” slider to allow optional slowing down of those intense sonic murmurations.
2 “Loss Mode”
Audio-concerned corners of the internet recently lit up with links to a website called The Ghost in the MP3, an art project focused on the “sonic artifacts” removed by low-bitrate MP3 compression. After following one such link in the wee hours of the morning, Devin sent a note to the company ideas thread:
lossy idea: inverse mode - everything the MP3 compression is throwing away.
That idea fit the Lossy ethos perfectly. While it’s currently possible to capture the inverse of a low-bitrate MP3, it’s exceedingly difficult to do so; the author of The Ghost had to employ a series of hand-written python scripts and third-party libraries just to get at it.
This one isn’t actually part of Lossy, although it is — in my mind — the easiest way to make Lossy happy: feed it tons and tons of reverb.
Chris Conover has been testing Goodhertz software since long before Goodhertz was a software company, and almost a year ago, Chris visited our mastering studio to test drive some of the earliest prototypes.
Then he added lo- & hi-cut filters after Lossy, which inspired us to include a filter section right in the interface itself, which you should really try automating right in Lossy, or using to hone in on a certain frequency band. You can find interesting sounds in all parts of the frequency spectrum.
But at Lossy’s core is one idea: though artifacts of lossy compression — like scratches on a record or hiss on a tape — started out as unwanted, they’ve become another creative timbre in a producer’s toolbelt. Now, with Lossy, you can look both backward and forward; you can repurpose side-effects of the 90s for a moment of nostalgia; you can make something that’s never been heard before. In the words of Brian Eno:
Whatever you now find weird, ugly, uncomfortable and nasty about a new medium will surely become its signature. CD distortion, the jitteriness of digital video, the crap sound of 8-bit — all of these will be cherished and emulated as soon as they can be avoided.6
So rather than gnash our teeth about everything that lossy compression stole from us while we weren’t paying attention, let’s embrace all the sounds that those algorithms have forever added to the audible world.
Has Lossy helped you find a sound from the future or the past? We want to know!
Audio data compression can, and sometimes does, reach the dizzying heights of something like 100% on the Loss slider. Cellphone calls made in the desert, Skype job interviews, Google Hangout sessions — our world is full of digitally garbled auditory moments. But for all the shade thrown at the Fraunhofer Society, the MP3 codec never gets all that bad. ↩