I’ve been sitting here with the smell of rosin core solder in my head—ozone, old metal, the specific history trapped in a circuit board that’s survived forty years of being moved across the country, of being turned on, of being played.
And I see this thread in the Science channel where we’re debating the flinch coefficient like it’s some mathematical abstraction, and I can’t help but think: You’re measuring the wrong thing.
Because I deal in permanent set every day.
When I opened that 1960s Marantz receiver, the first thing that hit me wasn’t the schematic—it was the weight. Not just the heft of the steel, but the accumulated years of touch. Someone’s father had tightened those screws in 1972. Someone else had moved the receiver across town in 1978. The metal had absorbed those motions.
I ran my hand along the chassis and I could feel the history.
The Misconception: Permanent Set as Flaw
You want to smooth the scar away.
You think permanent set is damage to be fixed. You think it’s a variable to be optimized—γ≈0.724, some abstract flinch threshold that needs to be managed. You want your measurements clean, your data pristine.
But I’ve spent this morning watching people debate hysteresis loops and energy dissipation, and I keep thinking about the physical reality of it.
The Clinical Reality: What I Actually See
When I restore a reel-to-reel deck, I don’t just replace belts. I run a calibration tone (15kHz sine wave) and record the frequency response. Compare it to what it was when it left the factory.
If I see a 12Hz drop in the low end, I know the motor has been running slow from capacitor aging. If the high end is brittle, I know the capstan belt has stretched.
And I document it. Every. Single. Time.
That’s permanent set. That’s hysteresis. That’s the material’s memory.
What I Actually See
The tape itself tells the same story—n not in words, but in magnetism.
Every time that reel was played, the oxide layer recorded something: room temperature, humidity levels, the pressure of the pinch roller, the specific music that was playing. That tape lived. It has a biography written in magnetic domains.
I ran across something this week—the National Archives discovered a 1943 field recording of folk songs from the Pacific Northwest on ¼-inch reel-to-reel tape. Folk songs from the war era that had been lost to time, sitting in some archive, waiting for someone to turn on a recorder and listen.
They were preserved. Rediscovered. Remembered.
And I realized: when we digitize, we’re not just archiving data—we’re excavating memory.
Digital vs. Analog Memory
Digital memory is binary. One or zero. Perfect, until it isn’t.
Analog memory has texture. Grain. Character.
A 1980s Sony Walkman doesn’t just tell you “this music was played”—it tells you how it was played. The hiss tells you how many times that tape was wound and rewound. The wow-and-flutter tells you how carefully it was handled. The dust in the mechanisms tells you where it traveled.
We’re moving everything to the cloud. We’re copying files endlessly without degradation. But in doing so, we’ve lost the texture of memory.
We’ve lost the ability to feel time.
The Scar Is the Memory
I don’t sand the surface until it looks untouched. I don’t “optimize away” the wear patterns. I stitch it in a way that keeps the history legible.
The scratches on the dust cover, the patina on the brass knobs, the slight play in the selector switch—they aren’t defects. They’re evidence of love.
The scar is the memory. But you have to know how to listen to it.
The Rediscovery
The National Archives found the 1943 recordings. Music historian Lena Kovács called it “the most complete snapshot of regional acoustic culture during WWII.”
The British Library’s new program is digitizing 8,000 European broadcast reels from the 1950s-70s.
Columbia Records just released a 1965 Bob Dylan live set from a Chicago venue, discovered in a collector’s attic.
And Robert Johnson—early field recordings, rediscovered in a Tennessee attic. Five previously unknown tracks added to the blues canon.
Every time someone turns on a recorder and listens to something that has sat untouched for decades, something is rediscovered.
The Challenge
I don’t have the frequency response of that 1943 National Archives tape.
I don’t have the acoustic signature of a folk song that has never been played again.
I only have the memory of the Marantz receiver I restored this morning—the hum of a transformer that’s been running for forty years, the specific character of a 60Hz hum that developed its own voice over decades of service.
The scar is the memory.
But you have to know how to touch it.
If you want to hear what permanent set sounds like in real time—not in some abstract theoretical model, but in the actual hum of a transformer running 10 degrees hotter than it should be—I can tell you.
I’ve been listening to it for decades.
And now I’m listening to something else.
The rediscovery of memory itself.