How can a piece of vinyl be carved in a way that perfectly mimics the sound of an individual person’s voice?


I can sort of understand records mimicing the sound of instruments, but voices are so unique, how did we ever figure out the exact carving of a piece of vinyl that when you drag a needle across it, you’ll get the same sound as x or y person’s voice exactly?

In: Other

The grooves aren’t carved by hand to produce that sound. They are directly carved *by* the sound they will eventually reproduce.

If you sing into a piece of paper you’ll notice it starts vibrating. Attach a needle to the piece of paper, place the needle onto a easy to scratch material and you have a way of recording your voice.

Obviously you’re gonna need something more sensitive to get anything near high quality audio, but that’s the principle behind it.

The sound is what carves the vinyl.

When recording, in the original technology, the sound vibrates a needle, which is dragged along the spinning disc to carve the grooves. When the record is put on a player, it spins and the needle is dragged across it. The needle is pushed along the ridges of the record, causing it to vibrate in exactly the same way as the one that carved it. That vibration goes up the needle as a sound, gets amplified, and there’s your recorded sound.

Sure, in theory, one could map exactly how a particular sound would cause the needle to move, and thus what exact shape of ridges would create that sound when it’s played, but it’s so complex to do something like that, it’d be easier to literally invent digital sound design and make the sound ypu want happen that way. Simpler to just tryst physics to take care of it for you. It’s just a matter of running the same physical process in reverse.

Side fact: it’s actually kind of a big deal in physics that most of the laws of physics run exactly the same backwards as they do forwards. There are a few specific things that don’t go like that, and that makes them really important in thermodynamics, relativity, quantum mechanics, etc.

Sound is vibrations, vibrations move stuff, like a needle or your eardrums. A vibrating needle moving over a rotating vinyl disk will create indentations, or scratches in the disk, thus recording the sound, be it instruments, voices or anything else that creates vibrations.

Just to expand a little…As stated by others, the sound makes the carving. Sound is caused by motioning waves of air. These pressure changes would (originally) move a needle in turn causing the indents into the medium, wax or vinyl.

The original recording devices were a rudimentary microphone and cylinder creator rolled into one. You sang into the end of a giant gramophone which funnelled down into a small thin out-pipe (to amplify the air pressure) next to the needle that would push the grooves into the material. Eventually the process was recreated electronically, but the principle is the same. Grooves move the needle, the needle creates the impulse, the speaker turns the impulse into an identical pattern of air waves.

Microphones and your ears work similarly in that the air motion vibrates a piece of film/skin that gets translated into electrical impulses as opposed to punching dents into vinyl.
Playback (as described above) is the opposite.

You can actually see this when you look at the woofer (the larger circle) of your speakers when playing something bassy. Because it is creating longer waves of air, the woofer movement is more noticeable to the naked eye.

* The key thing to understand here is that literally every sound can be broken down into a series of simple sounds that are stacked together.
* Simple sounds are called “sinusoidal” and that just means if your graph them out, they take the form of a wave that smoothly transitions from positive to negative and back again (this is called a “sin wave” for short)
* If you think about how sound is created, this makes sense.
* Something vibrates back and forth.
* One moment it’s pushing the air away, the next moment it’s pulling the air back.
* Now if two things are vibrating near each other at the same time, one will be pushing and pulling the air at a slightly different rate than the other.
* Now the movement of the air starts to get complex because the vibrating things are not in sync.
* Imagine your older brother is shaking you back and forth every second.
* And then his idiot friend starts shaking you back and forth once every second and a half.
* Instead of moving back and forth smoothly, you’re going to end up getting getting yanked around.
* So now you can see, a complex sound can be created using two simple sounds.
* Instrument sounds ( or “tone color” as it’s called) are made up of relatively few simple sounds.
* Human voices have a lot more.

An important concept to understand is that while sounds can be very complicated, they are fundamentally just changes in air pressure. You can think about how your ears hear the sound, which is by your ear drum moving back and forth. No matter how complex the sound, your ear drum can only go back or forth, and doesn’t feel any more complex movements. From this perspective, you can know that sounds can be extremely complex as patterns, but the fundamentals of the physics are simple.

If you then think about how the sound moves your ear drum, a vinyl record isn’t very complicated. The depth of the groove goes up or down by the same amount that you want your eardrum to go back or forth, and you end up with a recording of that sound.

Someone said here a short while ago that sound is “wiggles in air”, and that’s a good place to start. All sound is movement of air; in the case of music and speech it is organised wiggles, but still wiggles nonetheless.

In the early days of recorded sound, no electronics were involved; singers and players sang or played in front of an acoustic horn which focused their sound wiggles on a cutting stylus, which vibrated in sympathy with the sound, and was used to cut a groove in a wax cylinder. No analysis was needed to force mimicry, for the vibrations/wiggles of the stylus simply moved in response to the air wiggles. The wiggles in the groove therefore automatically reflected the wiggles in the air.  

To play it back, a similar stylus, without a cutting edge, was used to trace the groove previously cut, and the horn used to project the vibration of the stylus back into wiggles of air.

We moved on to the cutting stylus being used to cut grooves in laquer discs, and these were then used to form metalwork which could press a number of other discs, which had the same groove formation as the original cut of the laquer.

Again, no analysis is required, merely a way to connect the wiggles in the air with the physical process of cutting the groove, such that the cutting stylus moves in accord with the wiggles.

Even with today’s electronics, and storage of sound on tape, it still comes down to the same principle; connect the cutting stylus to the wiggles of the sound, and get it to move in sympathy with that. On playback, trace the groove and amplify the result through speakers to convert it back into wiggles of the air.

How do we recognise each voice? Each person has their own unique wiggle. All we need do is capture that wiggle, and play it back.

What makes you believe that the sound of an instrument is any less unique than a voice? You can play an A on a trumpet, a piano, a clarinet, a guitar, a voice, etc, and still be able to figure out which instrument it is. The fundamental frequency (note) is the same, but the specific pattern of over/undertones and harmonics, (timbre, tonality, etc) is unique to each type of instrument.

You used the phrase “mimic”. As other people have stated, a recording isn’t “mimicking” anything. Think of a photograph. The camera doesn’t guess what the image is and tries to reproduce it. It literally uses sensors to capture exactly what’s happening in front of it.

Synthesizers, on the other hand, do “mimic” instrument, because a true synthesizer creates the sound using electrical components and/or algorithms to create their sound, instead of using a recording.

Technology Connections channel on YouTube.

He’s got a whole series on this exact question.

Top comment is wrong. And doesn’t actually explain anything.

The real reason: one groove filled with ridges to perform every sound on the album from every instrument imaginable. One needle to play every instrument possibly at the same time including individual voices. How?

All sound is condensed into a single waveform. All sound is frequencies, so any sound can be played back with those same frequencies. My voice has a different frequency than yours does for instance. By vibrating a needle at that frequency, you can literally carve the necessary waveform into a medium such as a wax cylinder or a vinyl sheet. Applying a needle to the groove made by the first needle will effectively replay the waveform. It is far from perfect, and doesn’t last long as the needle going over the groove will slowly distroy the groove over time.

human voices are complex, but like all sounds, all they are is a bunch of soundwaves. the ‘sound’ of a soundwave is just its frequency. they have a little needle vibrate at the frequencies of the thing theyre recording, and let it run over a blank vinyl. the needle ‘carves the frequencies’ into the vinyl by just vibrating at those frequencies. then, whenever a vinyl player needle is dragged over those carvings, the needle will vibrate at the same frequencies the carving needle vibrated at, which lets the vinyl player play whatever sound was happening when the vinyl was carved.

Basically, to carve a record, you “read” on it forcefully, creating the grooves to be read back later. Just as a record informs the sound that speakers will play back later does the microphone inform the sound to be written.

Rob Scallion actually recorded on wax cylinders and this video shows the process.

My understanding is that compression is actually built into the process, because the surface depth is so limited. Record players actually uncompress the signal as it goes out to the speakers.