How we made the compact disc

Abstract

An international collaboration between Philips and the Sony Corporation lead to the creation of the compact disc. Kees A. Schouhamer Immink explains how it came about.

Full text

260
reverse engineering
© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. © 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
How we made the compact disc
An international collaboration between Philips and the Sony Corporation lead to the creation of the compact disc.
Kees A. Schouhamer Immink explains how it came about.
Kees A. Schouhamer Immink
In 1973, I joined the optics group of Philips
Research in Eindhoven to work on the
videodisc, a 30-cm-diameter optical disk
that can store up to 60 minutes of analogue
video and sound. The launch of the
videodisc in 1978 was a technical success,
but since consumers showed little to no
interest it was deemed a failure for Philips.
In the early 1970’s, Lou Ottens, technical
director of Philips’s audio division, requested
a prototype of an audio-only disk based on
the videodisc technology. Believing that
such a sound-only disk was a trivial matter
since the videodisc that had been previously
demonstrated, my research director at the
time, Piet Kramer, asked two engineers from
Philips’s audio division, Loek Boonstra and
Toon van Alem, to develop the audio-only
disk. They fared well and, in 1976, they
demonstrated the first prototype of a digital
audio-only disk using videodisc technology.
In October 1979, Philips and Sony
decided to join forces in the development of
a digital audio disk standard. I had nothing
to do with the hard work on prototyping
the audio-only disk. I was first involved
when the small joint Sony/Philips task force
of experts was formed. Since there was no
one else available within the optics research
group, I was named the ‘expert’ who dealt
with servos, electronics, and coding (apart
from reading a textbook, I had no expertise
in this particular field).
Between 1979 and 1980, five expert
meetings, alternating between Tokyo and
Eindhoven, were held. There were many
important parameters to be discussed and
decided: playing time, disk diameter, error-
correcting code, and sound quality. Playing
time and disk diameter are probably the
most visible parameters for the general
public, and these two are, of course,
related: a 5% increase in disk diameter
yields 10% more disk area, and thus a 10%
increase in playing time. Philips director
Lou Ottens, who developed the compact
audio cassette in the 1960s, proposed a
115-mm disk diameter. The cross diameter
of the compact audio cassette, which was
very popular at that time, is 115 mm. He
reasoned that since it was a major success,
the compact disc should be the same size.
Sony, no doubt with portable players in
mind, initially advocated a 100-mm disk.
In December 1979, during the meeting
in Tokyo, the partners agreed that the sound
quality would be 16-bit (44.1 kHz sampling
frequency), the error-correcting code would
be Sony’s CIRC (cross-interleaved Reed–
Solomon code), the playing time would be
74 minutes, and the disk diameter would be
120 mm. The disk diameter is a remarkable
choice as both partners initially preferred
a smaller disk. Philips’s website with the
‘official’ history states that “the playing time
of the CD was determined posthumously
by Ludwig van Beethoven”. Sony’s vice-
president at the time, Norio Ohga, insisted
that he wanted the composer’s ninth
symphony (Fig. 1) to fit on a single disk.
He had identified that the longest known
performance lasted 74 minutes and 33
seconds. This was a noisy mono recording
made during the Bayreuther Festspiele in
1951 conducted by Wilhelm Furtwängler.
For the state-of-the-art at the time we
calculated that a diameter of 120 mm was
required for Beethoven’s ninth. The change
of diameter from 115 to 120 mm had a
dramatic effect on Philips’s planning as their
player design and disk manufacturing were
all based on 115 mm.
It is plausible that Norio Ohga was a
great fan of Beethoven’s ninth, but he must
also have been well aware of the significant
consequences for Philips’s planning.
There are also other reasons to suggest
that the Beethoven story for determining
the playing time might be a myth. In the
first place, the channel code was not yet
decided in December 1979. Within two
months, I was able to increase the playing
time by 30% by designing a new code, ‘EFM’
(eight-to-fourteen modulation). A little
arithmetic shows that with the new channel
code a disk diameter of 100 mm would
accommodate the requested 74 minutes
playing time. The second reason is that in
October 2017 I had a lunch conversation
in the Takanawa Prince Hotel in Tokyo
with my brother in arms, Sony’s top-
engineer, Toshitada Doi, who, when I told
him about my doubts about Beethoven,
said, “Of course you are right, but it was a
good story, wasn’t it?” ❐
Kees A. Schouhamer Immink
Turing Machines Inc., Rotterdam, e Netherlands.
e-mail: immink@turing-machines.com
Published online: 17 April 2018
https://doi.org/10.1038/s41928-018-0063-7
Fig. 1 | Sketches for Ludwig van Beethoven’s
Symphony No. 9 in D minor, Op. 125.
Credit: Granger Historical Picture Archive / Alamy
Stock Photo.
Nature electroNics | VOL 1 | APRIL 2018 | 260 | www.nature.com/natureelectronics