This site is only a doorway

Everything here is gathered from things other people actually researched, thought through, and built: books, papers, historical records, and the people behind them. If something here catches your interest, go look at the original sources listed at the bottom of the page. They carry more depth than a page like this ever can.

How Did Carrying Music Go From Cassette Players to MP3 Players?

Category: Technology

In 1979, the Japanese company Sony put a small cassette player on sale. It could not record. It had no built-in speaker, either. Some people at the time doubted that anyone would want such a machine.

It did catch on. And over the next 22 years, carrying music went through two kinds of shrinking. First the player got small enough to wear while walking. Then the music itself got small enough that one pocket-sized device could hold about 1,000 songs. This article follows both steps and asks one question along the way: how can a music file lose about nine-tenths of its size and still sound like music?

The short answer is that MP3 does not throw away sound evenly. It saves its effort for the sounds you notice and handles the sounds that are harder to hear more roughly.

1. Why the First Walkman Left Out Recording

Before the Walkman, Sony already sold a small cassette recorder called the Pressman, made for reporters. According to Sony's own company history, Sony's co-founder Masaru Ibuka had taken a stereo cassette recorder and headphones on a trip abroad. He enjoyed the stereo sound on the plane, but the machine was too heavy. Before another trip to the United States he asked for a Pressman with the stereo circuits added. His words, as Sony tells it: "I can do without recording." [1]

That one sentence explains the design. Taking out recording and the speaker left a machine that did only one job: play stereo music into headphones. Ibuka tried the result himself and then showed it to Sony's other co-founder, Akio Morita. Morita described it as a stereo cassette player you could listen to while walking. He wanted it in stores before the summer holidays, aimed at younger people, for under 40,000 yen. [1]

The headphones were part of the puzzle too. Ordinary headphones of the day weighed 300 to 400 grams (about 0.7 to 0.9 pounds). Sony's audio team happened to be working on a pair that weighed only 50 grams, and the two teams joined forces in March 1979. [1] Only about four months were left before the planned launch. [1]

2. Why the Walkman Took Off Slowly, and Why It Got Three Different Names

The Walkman went on sale in Japan on July 1, 1979, at 33,000 yen. The first batch was 30,000 units. [2] A music player you wore on the street was a strange idea, and the start was quiet. Sony's history says most newspapers did not even report the launch, and only about 3,000 units sold in July. [1]

Sony's staff did not wait. They rode Tokyo's circular Yamanote train line all day with Walkmans on, so people could see one in use. They also handed out demonstration tapes in shops and ran magazine ads with young pop stars. By late July and August the first 30,000 had sold out. [1] The two sources agree on the pattern, a weak first month and a sell-out within about two months, though they word it a little differently. [2]

Why did a player with fewer features win? The sources do not give one reason, so this is a reading of the evidence, not a statement from Sony. The machine was built for young people who wanted music all day, and it gave up recording to do that one thing well. [1]

The name was a surprise to English speakers. When Sony prepared to sell abroad, people worried that "Walkman" was not proper English. So the company used other names at first: "Soundabout" in the United States, "Stowaway" in the United Kingdom, and "Freestyle" in Australia. But on business trips, Morita kept hearing people ask for "the Walkman you're selling in Japan." The name had already traveled. He kept it and announced the decision in April 1980. [1]

3. Why a Cassette Player Could Not Hold a Whole Music Collection

The Walkman made the player small. But a cassette holds only what is on that one tape. To hear something different, you swapped tapes. A simple way to picture it: if you want ten albums with you, you carry ten cassettes.

For a pocket device to hold hundreds of songs, the music had to be stored as files, and the files had to be small. Music on a compact disc (CD) is large. A CD stores sound as 44,100 measurements per second, in 16 bits each, for each of two stereo channels. Multiplying those out gives about 1.4 million bits for every second of music. So a computer or a small player would fill up fast unless the music could be squeezed.

4. How MP3 Makes Music Files About Ten Times Smaller

A course from the UK's Open University says MP3 files are typically about one-tenth the size of the equivalent uncompressed music files, such as CD files. [3] That is a rule of thumb. The exact ratio depends on the settings used.

The trick is a feature of human hearing called masking. When one sound is loud, it can hide a quieter sound that is close to it in pitch. The Open University material also describes a time version: a loud sound can hide quieter sounds that come right after it. [3] Karlheinz Brandenburg, one of MP3's inventors, explains that MP3 relies on how the ear works, along with statistics about the signal, and that masking is stronger when music is complex and covers a wide range of pitches. [4]

So the encoder, the program that makes the MP3 file, spends fewer bits on parts that masking hides and more on parts you would notice. A simple way to picture it: in a noisy cafeteria you cannot hear a friend whispering across the room, so it would be a waste to record the whisper carefully. One caution. This does not mean MP3 deletes every sound you cannot hear. It handles hard-to-hear parts more roughly, which is why very low settings can start to sound off.

5. Why a Song With Only a Voice Was So Hard for the MP3 Team

MP3 began as research in Germany. In 1987 the work grew into a larger Fraunhofer project, partly to support a planned European digital radio service. [5] Brandenburg and his team had a hard time convincing people that music needed compression at all, since the new CD seemed to have plenty of room. [5]

In early 1988 the team hit a wall. Brandenburg tried Suzanne Vega's song "Tom's Diner," which is sung with no instruments. The test came back mostly as scratchy distortion. His model of what the ear can and cannot hear was not accurate enough, so the team went back and improved it. [5]

Brandenburg later explained why a lone voice is hard. A voice has a basic pitch plus a set of higher overtones, and the algorithm codes them separately. A voice also spreads over a wide range of pitches. For this song, the program reported that it would need four times the bit rate that was available. [4] Without other instruments to hide behind, mistakes in a single voice are easy to hear. That is a sensible reading of his explanation, though he did not put it in exactly those words.

The format got its name on July 14, 1995, when the team settled on the file extension ".mp3," a shorter form of "MPEG-1 Audio Layer 3." [5] That date is when it got its name, not the year MP3 was invented. The research had been going for many years.

6. What the iPod Added in 2001: 1,000 Songs in One Device

On October 23, 2001, the American company Apple announced the iPod. Its press release described a music player holding up to 1,000 songs at what Apple called "CD-quality" on a 5 GB hard drive, weighing 6.5 ounces (about 184 grams). A built-in FireWire port could copy a whole CD into it in under 10 seconds. It played for up to 10 hours, and the suggested price in the United States was $399. [6]

Put the two steps side by side. The Walkman made the player small, and the cassette still limited what you could carry. MP3 made the music small, and a hard drive gave it room. Apple's own headline for the product was "1,000 songs in your pocket." [6]

7. Is It Allowed to Copy Your Own CDs onto a Computer?

This question comes up as soon as music becomes files. There is no single answer; it depends on your country. In the United States, the doctrine of fair use lets people use copyrighted works without permission in some situations. Courts weigh factors such as the purpose of the use and its effect on the market, case by case. [7] The Copyright Office itself says its index is not a substitute for legal advice. [7] Other countries have their own rules. Giving copies to friends or posting music online is a different matter. Even a copy just for yourself can be treated differently depending on the country and the situation. If you are unsure, ask an adult and read your own country's official copyright guidance.

8. Testing Masking at Home With a Running Tap or a Fan

With an adult nearby, play a song at a low volume, quiet enough that it would not bother anyone. Keep the player well away from water. Now turn on a fan, or run a kitchen tap some distance from the speaker, and listen. Quiet details in the music, such as soft cymbals or the breath before a note, may become harder to hear. Turn the noise off and they may come back.

This is only a rough version of masking. Real MP3 encoders use detailed models of hearing, and a fan does not hide sound in exactly the same way as MP3 encoding does. Please do not turn the volume up to test it, because loud sound can hurt your ears. If you want to read more, the interview with Brandenburg [4] and the Internet Hall of Fame story [5] are both short and readable.

Sources

Where I could read only part of a source, I say so below. The Walkman launch details come from Sony's own history and a Japanese news feature. The iPod facts come from Apple's own press release.

  1. Sony Group Portal, "Sony Group History, Volume 2 Chapter 3" (80th Anniversary). https://www.sony.com/en/SonyInfo/CorporateInfo/History/SonyHistory/2-03.html (Company history by Sony itself. Used for the Pressman origin, Ibuka's request, the headphones, the July 1979 launch, the first-month sales, and the product names. I read the passages of the Walkman chapter that cover these points, not every line of it. It is the company's own account, so it tells the story the way Sony remembers it.)
  2. nippon.com, "How the First Walkman Was Born" (Japan Topics, in Japanese). https://www.nippon.com/ja/japan-topics/g00726/ (Secondary source based on newspapers and catalogs of the time. Used for the launch date, the 33,000 yen price, the 30,000-unit first batch, and the weak first month. Read in full.)
  3. The Open University, OpenLearn, "Exploring communications technology" (MP3 sections). https://www.open.edu/openlearn/digital-computing/exploring-communications-technology/content-section-3.6 (University course material. Used for the one-tenth size rule of thumb and for frequency and temporal masking. I read the nearby sections of the page, not the whole course. The 1.4 million bits per second figure is my own arithmetic from the standard CD format, not a number quoted from this source.)
  4. Fraunhofer-Gesellschaft, "Beyond mp3: Karlheinz Brandenburg on his new project" (interview). https://www.fraunhofer.de/en/quick-links/alumni/news/mp3-inventor-brandenburg-interview.html (Interview with an MP3 inventor. Used for masking, signal statistics, and the "Tom's Diner" difficulty. The visible page was cut off partway through, so I read the first part of the interview, which covers these points.)
  5. Internet Society, Internet Hall of Fame, "How MP3 Got Its Groove" (2014). https://www.internethalloffame.org/2014/08/13/how-mp3-got-its-groove/ (Short history of MP3. Used for 1987, early 1988, and the July 14, 1995 file extension. The page was cut off near the end, but the dates used here came from the part I read. It counts the work from 1982, when Brandenburg's doctoral research began; I use 1987 only for the start of the larger Fraunhofer project.)
  6. Apple, "Apple Presents iPod" (press release, October 23, 2001). https://www.apple.com/newsroom/2001/10/23Apple-Presents-iPod/ (Primary source. The current Apple page came back nearly empty when I opened it, so I read an archived copy of the same release from 2002. Used for capacity, weight, transfer speed, battery time, and price.)

Revision history

  • First published.