Part of How The Heck?, a series of interactive explanations of everyday technology designed primarily for curious, non-technical readers.
If you like these, you can follow the author, buy him a coffee, or suggest what comes next.
✨200,000+ views📊60+ bespoke visualizations📝60+ TL;DRs🤎25+ coffee donations
Picking up a record and realizing it's a physical version of a musical performance you love feels close to magic. In a way, it's the purest example of something this whole series keeps circling: taking a concept and using mathematics and physics to convert it into a more useful form. GPS turns time into location, solar panels turn sunlight into electricity, Shazam turns sound into an identifiable fingerprint, and so on.
For records, a tiny groove carved into a piece of plastic can somehow reveal an entire orchestra in high fidelity. In stereo, the magic is doubled: a single record gives us two distinct channels of sound.
In reality, a lot of smart people built on the work of other smart people to make this happen: Edison inventing the phonograph in 1877, Berliner inventing the gramophone in 1887, and Columbia Records introducing the 33⅓ RPM vinyl LP in 1948. But the core idea behind all of them is the same one we'll build up from here.
To understand how it all works, let's start with sound itself.
Sound is a wave
TL;DR
Sound is a wave of air pressure oscillating over time. We can graph that as a waveform.
If you've read the pieces on Shazam or synthesizers, you'll know this part already! But it's always worth reviewing.
Sound is a wave of air pressure. When the string on a guitar or a speaker vibrates, it pushes and pulls on the air around it in an oscillating pattern. When that pressure wave reaches our ears, it makes our eardrums vibrate, which sends a signal to our brain that we interpret as sound.
The cool thing is that even a band playing multiple instruments will make sound that our ears will process as a single waveform. The individual sounds combine additively by the time they reach our ears.
Our brains are smart enough to decipher the individual sound waves and interpret them as multiple instruments playing together.
The intuition behind vinyl records
TL;DR
We can represent the waveform of sound radially as grooves in a record.
Given that we can represent a piece of music as a waveform, the question of analog audio formats is simple: how do we represent that waveform on a physical medium?
Records do this by carving a spiraling groove into a piece of plastic called vinyl. That groove represents the changes in air pressure that made up the original sound wave.
The core intuition here is that the grooves that spiral around the record physically represent the sound wave itself. The mechanism that does this is the interesting part.
Riding the grooves
TL;DR
The grooves on a record move laterally based on the original sound wave. In a mono record, the stylus will only ever move left or right.
Vinyl itself is a cheap, quiet, durable, and moldable plastic. It's the perfect material for a record.
In a process we'll get into more later, a cutting lathe carves the grooves into the vinyl. A perfectly smooth spiral would represent silence, and very sharp wiggles would represent rapid changes in air pressure.
There's no physical law for exactly how the grooves need to represent the sound wave; it just needs to be consistent between the cutting and the playback. For instance, you could decide that the groove veering left means a drop in air pressure, as long as the player interprets it the same way. In practice, the industry settled on the same standard, which is why any record plays on any turntable.
The part of the record player that rides above the grooves is called the tonearm, and the tip that specifically makes contact with the record is called the stylus. The stylus conventionally follows the groove from the outer edge of the record inward, following the spiral of the grooves.
In a mono record, the two walls of the groove move laterally together. The stylus makes contact on both walls, but will only ever move in one direction: left or right. This is easier to see with a head-on view of the stylus.
The stereo trick
TL;DR
To get two channels of sound, the stylus needs to move in two independent directions. To do this, the grooves are carved such that the two walls represent the two channels and sit at a 45° angle from horizontal, perpendicular to each other.
The example above is a mono record, meaning it will give you a single channel of sound. If you have multiple speakers, they will play the same thing.
But because we have two ears, we can perceive when different sounds come from different directions. This is the idea behind stereo audio, where we have two independent channels of sound.
If the stylus moving in one direction can give us one sound wave for a mono record, stereo needs the stylus to move in two directions. This calls for breaking some symmetry.
Stereo grooves are carved such that the two walls represent the two channels and sit at a 45° angle, perpendicular to each other. This way, the stylus can move in two directions at once independently, with each wall moving the stylus along its 45° axis.
Why perpendicular angles?
It's important that the two walls sit perpendicular. If they sat, say, 30° to each other: a push meant for one channel would also move the stylus along the other, making it impossible to answer "how much left?" and "how much right?" independently. In the most extreme example, if the walls sat 0° to each other, meaning they were fully parallel, the two channels would be completely indistinguishable from each other.
The 45° stereo system works because the two channel axes are orthogonal, meaning . Moving the stylus along one axis creates zero motion along the perpendicular axis, so the channels don't interfere.
As you change the angle between the axes from toward , the left channel begins projecting onto the right. That projection, a dot product, is:
At 30°:
So about 87% of the left channel leaks into the right. The two are no longer cleanly separable.
It's the same reason our Cartesian coordinate system is based on perpendicular axes. It lets us neatly answer questions like "how much X?" and "how much Y?".
The sum and difference trick
Consider L and R: how far each channel pushes the stylus along its own 45° wall. And consider H and V: the stylus's overall motion, horizontal (side-to-side) and vertical (up-down).
The two connect like this:
So the groove stores the sum and difference of the two channels. The originals can be recovered by reversing that:
Strictly speaking, because each push is at 45°, the actual sideways and vertical distances are scaled by a constant factor of √2⁄2. We've left it out here since it doesn't change the relationship, only the overall size.
Two independent measurements in give us two channel values out.
From motion to a signal
TL;DR
The cartridge converts the stylus's motion into an electrical signal by using a magnet to induce a current in a coil of wire.
Motion is only the first step in our journey to getting a sound out of the record.
The cartridge is the part of the record player that converts the stylus's motion into an electrical signal, just like a microphone or a guitar pickup. The cartridge contains a magnet and a coil of wire. The magnet is attached to the stylus, and as the stylus moves, so does the magnet. The moving magnet shifts the magnetic field around the coil, which induces a current through a process called electromagnetic induction.
The idea works the same for stereo. We just need two coils, oriented the same way as the two walls of the groove, such that the motion of the stylus induces a current in each coil. Two different physical walls give us two independent measurements of motion, which we can use with two coils to give us two independent electrical signals.
The diagram shows a moving-magnet cartridge. Some higher-end cartridges flip it, so the coil moves and the magnet stays fixed, but the principle is identical. In either case, this is how we go from a physical motion to an electrical signal using a magnet and a coil of wire.
The RIAA curve
TL;DR
The RIAA curve is a standardized way to cut low frequencies and boost high frequencies during recording, and reverse it during playback.
As we learned in the JPEG piece, when we're dealing with any form of storage, analog or digital, we want to be as economical as we can with space.
In the case of audio, one issue is that low frequencies require much larger, wider side-to-side swings in the groove to be heard at the same volume as high frequencies.
The RIAA curve (Recording Industry Association of America) gives a standardized way to cut low frequencies and boost high frequencies during recording, and reverse it during playback. This lets us pack more spirals into the same amount of physical space, without audibly losing fidelity in the final sound. The boosted high frequencies also help hearing details that may be lost with the record's background hiss.
This means that the pure signal that the cartridge produces gives us a tinny sound without any further processing.
The audio chain
TL;DR
The preamp is the component that does the RIAA curve correction, and the speaker produces sound by following the reverse process of the cartridge.
The signal from the cartridge is very weak and goes through a few stages of amplification before it reaches the speaker.
The preamp is the component that does the RIAA correction that we discussed above. Physically, it can be a small amplifier built directly into the turntable, or it can be a separate component that plugs into the turntable.
From the preamp, the signal goes through a power amplifier, whose job is to boost the signal to a level that can drive the speaker.
The speaker's job is to convert the electrical signal into sound waves we can hear. We can think about this as the reverse process of the cartridge: instead of motion creating an electrical signal, a coil and magnet use an electrical signal to create motion.
In a typical speaker, the electrical signal goes into the coil, making it an electromagnet. Unlike the cartridge, the magnet in a speaker is fixed, and the coil is the part that moves. The electromagnetic coil and fixed magnet create motion, which moves the cone of the speaker, displacing the air around it in an oscillating pattern, creating the sound we know and love.
Playback speed and pitch
TL;DR
Records are cut at specific speeds, which give us tradeoffs between fidelity and physical space.
Records are cut at a specific speed, usually:
- 33 ⅓ RPM for full albums
- 45 RPM for singles
- 78 RPM for older records
Faster speeds give us more fidelity but use more physical space on the record. As always, there are tradeoffs.
Playing a record at the wrong speed will change the pitch of the music, for the same reason a sped-up voice recording will sound like a chipmunk.
How records are cut
TL;DR
Instead of a stylus riding a groove, a cutting lathe uses a sharp cutting stylus to carve the groove into a soft disc as it spins, based on electrical signals from the recording equipment.
We've covered a lot of mechanics about how records are played, but we haven't talked much about how they're made.
It's essentially the whole process in reverse. Instead of a stylus riding a groove, a cutting lathe uses a sharp cutting stylus to carve the groove into a soft disc as it spins. The electrical signal from the recording equipment drives the cutting head. The RIAA curve is applied at this cutting stage, waiting to be undone by your preamp later.
That disc is usually called the lacquer master, and it's used to make the metal stampers that press copies of the record, not unlike a waffle iron.
My background is largely in software, so it's been a blast to learn how analog things actually work. It's wild to think that the grooves on a record are just microscopic variations carved into a piece of plastic, a physical copy of my favorite band playing in a studio, pressed from a master disc. And the whole thing runs in reverse just as easily: a bit of electricity through a coil becomes motion, motion becomes a groove, and a needle turns it all back into sound.
I hope it feels like magic to you too.