Get best of the week

Weekend Reading: A History of Audio Formats - The Age of Cassettes and the Development of Speech Synthesis Technologies

Jokes about millennials and the compatibility of a pencil with a cassette have long passed into the category of “button accordions,” but videos in which synthesized speech is played instead of the speaker’s voice can still be surprising.

In this collection, you will find historical analysis of all stages of the development of compact cassettes, and in the second part - a series of materials on the development of speech synthesis technologies.


Photo by Chris Benson / Unsplash

Beginning of the era of compact audio or “first shift” for reels. We tell how compact cassettes replaced bobbins. This process was gradual. So, in the early fifties, technology “absorbed” the principle of an endless movie tape. It turned out something remotely resembling a modern cassette. The project was called Orrtronic Tapette and developed a special player for the media, but the device was quite bulky, and the technology itself had a significant number of disadvantages. Fortunately, similar formats were developing in parallel - for example, Fidelipac. We pay a little more attention to him. By the way, this format has taken root very well on radio stations, where it was used until the early nineties.

How cassettes came into cars . Give me five, if you remember how interesting it was to delve into the glove compartment of my father’s car in search of cool collections on cassettes. This is a continuation of our cycle of historical materials, and this time we are talking about the advent of automotive compact audio. It all started with the fact that in the fifties the same Fidelipac was finalized to increase the playing time, so the Stereo-Pak appeared. Later they developed a radio tape recorder for this carrier, but it was not without problems either - the tracks had to be switched manually, rewinding forward threatened to break the tape, and the price bit. But Stereo-Pak played in car dealerships until the early seventies until it entered the market - 8 Stereo - a new compact audio format.

Cassettes with eight soundtracks. This format was invented by the self-taught William Lear, the one that gave the world the first Motorola (only it was a radio). When the Stereo-Pak cassette fell into his talented hands, he immediately saw good potential in it and decided to “tighten up” the medium. Increased the number of tracks to eight, removed some elements to simplify production - and the world saw a new format, which already housed 90 minutes of recording. Tracks no longer needed to be switched manually, but such a cassette had its own drawbacks - the recording quality suffered due to narrow tracks. But the cunning Lear made a knight's move - he convinced the Ford Motor Company to sew a player for Stereo 8 into the basic equipment of their cars, which helped him capture a considerable share of the American market. But already in the mid-seventies this format began to concede to those tapes that we know now,although some indie bands like RTB2 and Melvins are still trying to revive Stereo 8 by releasing their small albums on it.

Miniature bobbins in a cassette form factor . Attempts to enclose the film in a compact plastic case were made even before the advent of Fidelipac. In the mid-fifties, RCA craftsmen made a cartridge with two small reels inside, although there was no urgent need for a new product. The company already felt very good (for example, only they could sell Elvis records). But the complexity of handling reel players cut off a significant part of the audience, so RCA decided to simplify the format a bit. It was possible to record 30 minutes of sound on four tracks of such a cassette; it was called Sound Tape. But the format never became widespread - the quality was inferior to bobbins, and such a cassette cost a dollar more. By the mid-sixties, Sound Tape had become history, having existed on the market for only six years.

German answer to cassettes . All previous developments were from the USA. But what about compact audio in Europe? In this article, we understand the history of DC International. In the 63rd year, Grundig developed a cassette, inside which two reels with film were placed (hence the abbreviation DC - Double Cassette). It was possible to record 45 minutes of content on such a medium. But it was not possible to conquer the DC market - Philips compact cassettes have been around for two years, and the Germans did not manage to lure the mass audience, and production stopped in the 67th. Now such cassettes are of interest only to collectors and enthusiasts.


Photo by Peter Stumpf / Unsplash

History of Speech Synthesizers: First Mechanical Installations. The ancestors of Siri existed as far back as the 18th century in the Russian Empire. Christian Gottlieb Kratzenstein developed a system of acoustic resonators - the mechanism could reproduce the vowels a, e, and, o, y (a replica of the device is in the Dresden Technical University). But in a similar historical period, other gadgets were introduced. So, the Austrian Wolfgang von Kempelen invented a machine that simulated the voice path and pronounced short phrases in several languages, although it was very difficult to drive. Then came the talking head - a hybrid of piano and Kempelen device. The keys controlled the jaw, mouth and tongue, and the blacksmith bellows performed the role of the lungs. The car was called "Euphonia," but because of its frightening appearance, it never found a place in the houses of the gentlemen.

The Age of Electrical Solutions. In this article, we consider approaches to the invention of twentieth-century speech synthesizers. As early as the end of the 19th century, the German scientist Hermann von Helmholtz showed the world a diagram of a device that could synthesize some simple vowels and diphthongs. But only in 1930 did the first real speech synthesizer appear, VODER (in the material we talk about the principles of its work - we even found an archive record with a demonstration). Next came the synthesizers based on spectrograms. The first, developed by the American L. Schott, used linear light sources that shone through spectrographic patterns, and the second, Franklin Cooper, read images on a transparent tape and modulated the harmonics of the fundamental tone. Such devices became the progenitors of computer speech synthesis.

Computer era. This article comes close to the present. In the mid-fifties, the Japanese entered the speech synthesis market, who showed the first transistor computer with built-in memory. A few years later, a synthesizer appeared on its base, which was able to pronounce about 1,500 words - not bad for the last century. Later, fully digital synthesis based on IBM appeared. But a real breakthrough occurred when the principle of linear prediction was introduced into the technology (there is a video in the material where, using the speech synthesis system, on the basis of this principle, they tried to order pizza by phone). We conclude the story with examples of inventions from the 70-80s: a machine that could read text from paper, and a children's toy for training pronunciation of words.

Additional reading:



PS On April 18, the Pro-Ject Vinyl Academy and Audiomania will host the “ Online Apartment Building ”.

More articles:


All Articles