ArticlePDF Available

The future of digital audio recording

Article

The future of digital audio recording

Abstract

The following keynote address was presented to the 104th Convention of the Audio Engineering Society in Amsterdam during the society's golden anniversary celebration on May 17, 1998.
KeesA.SchouhamerImmink*
Philips Research Laboratories, Eindhoven, The Netherlands
-?
j5
The following keynote address was presented to the 104th Convention ?
-, of the Audio EngineeringSocietyin Amsterdamduring the society's ':_
goldenanniversarycelebrationon May17, 1998. !
__ redicting the future is lowed Moore's prediction. In the early seven, and the industry came up--very
I ! 1 )\, very difficult, as you 1980s, HDDs with a capacity of a surprisingly--with an increase in reso-
Il _k'/J/probably know. On the megabyte were quite common. Now lution and sampling frequency by a
] _- .... other hand, predicting the small laptops equipped with HDDs factor of two. The industry also came
d-_ -,: future of audio recording having a capacity of a few gigabytes up with the "great" idea of using more
technology may not be that difficult, are commodity products. Again, an in- channels, usually five, instead of the
crease by a factor of a thousand in 18 usual two stereo channels. This is a
One person has been very successful in
predicting the future of electronics tech- years time. nice beginning to spending the bit bud-
The capacity of optical recording get, but if we have a factor of a thou-
nology during the last 30 years or so.
His name is Gordon Moore, the founder products, say the Compact Disc, has sand to spend, we must try to act as a
not followed Moore's rule that the ca- really big spender. For example, let us
of Intel, which manufactures the elec-
tronics used in many computers. As a pacity would double every 18 months, think of playing time: we could store a
The capacity of the Compact Disc, 1000 hours of music on the new disc.
young electrical engineer, Dr. Moore
which was introduced 15 years ago, is That is 40 days of around-the-clock
was one of a team that made the first in-
tegrated circuit. This chip contained five approximately 600 megabytes. The ca- CD-quality music. With eight discs
to 10 transistors, pacity of optical discs changed only one could play music for a full year.
Some 30 years ago, Dr. Moore pre- once when the DVD was introduced a Note that the average Western house-
dicted that the number of transistors few years ago. The capacity of the hold owns less than 200 Compact
DVD is 4.7 gigabytes, which is a fac- Discs. So with just one "super disc"
per square millimeter (or square inch)
tor of seven more than that of the clas- one could store Mr. Average's reper-
would double every 18 months. This
prediction, now called Moore's law, sic CD introduced 15 years ago. Ac- toire. Unfortunately, the price of such a
has been very accurate. For example, cording to Moore's law, one might disc would be, say, $20,000, which is
in 1980 at the time of the introduction have expected a growth in capacity of well above the budget of an average
a factor of a thousand instead of seven, household.
of the Compact Disc, a standard mem-
ory could store 16 thousand bits, while Evidently, optical recording is a poor There is a natural limit to the in-
performer in terms of capacity growth, crease in the sampling frequency and
current memory units can store ap- This is not the right setting to discuss resolution. Some people even question
proximately 16 million bits. This is a
growth factor of a thousand in just 18 the various factors that have played a whether the average consumer is very
role in this underperformance, but in much interested in the improved per-
years time. Not only has the storage
capacity increased by a huge factor, the coming years the capacity of opti- formance that DVD audio is providing
but computer speed has also increased, cal discs will grow much faster than in this respect. So we have to think of
hitherto, somethingmoreattractive.Anincrease
prices went down, and so forth. The
Let us assume, for the sake of dca- of spatial resolution may give a radical
audio engineering industry has benefit-
demic interest, that from now on the answer to the technology push of the
ed a lot from this electronics revolu- storage capacity of optical recording factor thousand we have to spend. My
tion. In fact it was the driving force be- will follow Moore's law. What does ultimate audio format would have the
hind the digital audio regolution. All
kinds of operations can be done in the this mean for the recording industry? following main parameters:
digital domain much more accurately What kind of data-hungry formats can
we think of that will consume a factor Resolution: 32-bit PCM
than was ever thought possible, of a thousand more bits than the Corn- Sampling frequency: 196 kHz
Thecapacityofstoragedevicessuch Numberof channels: 256.
as hard disk drives (HDDs) has not fol- pact Disc? The change from CD to
DVD gave a slight flavor of what kind This format would be, for the time
*Now with the Institute of 'Experimental of options the audio industry has. This being, future proof.
Mathematics,Essen, Germany. change involved only a small factor of The small technical problem is e:>
j. Audio Eng. Soc., Vol. 47,No. 3,1999March 171
*_-'Future of AudioRecording
·.}
,. that, apart from a few audio buffs, Technical Committee on Loudspeakers
;probably no one would install 256 and Headphones to take a closer look at
loudspeakers in the living room. invisible loudspeakers.
Clearly, state-of-the-art loudspeakers Let me conclude with the following
are ugly, require cables, and so forth. In observation: In the future there will be 1999 April 13-16: 13th European
order to solve this technicality we must a proliferation of audio formats,proba- Frequency and Time Forum and
(re-)invent the concept of invisible bly mutually incompatible. Since the 1999 IEEE International Fre-
quency Control Symposium,
sound or invisible loudspeakers. You 1970s we have witnessed a rapidly Micropolis, the Parc des Exposi-
may hear them, but you cannot see growing number of recording formats tions et des Congres de Basan-
them. A name for this could be "the that have been introduced into the
con, France. Contact Isabelle
emperor's new loudspeaker system." In- market. The average time interval be- Bourgon, tel: 33 3 81 40 28 21
visible sound is not science fiction. It is, tween the introduction of new formats or fax: 33 3 81 88 57 14.
at least in principle, physically possible is getting smaller and smaller. After
to make invisible loudspeakers by using analyzing ample statistical data, I con- 1999 April 17-22: NAB 99, Las
modulated laser light in air using the clude the time is ripe to-postulate a Vegas, NV, USA. For informa-
photo-acoustical effect. There are a few new law that simply states that the tion, tel: 800-342-2460 or e-mail:
technical obstacles with the cables, in time interval between two consecutive register@nab.erg, fax: +1 732-
particular the power cables, as evident- introductions of new audio formats 544-2888.
ly batteries cannot provide the required will halve every 18 months. On the ba-
power. So we have to think of altema- sis of this law, in 2023, at the Society's 1999 April 28: The American
tive power supplies, such as solar pow- 75th anniversary, the introduction of Loudspeakers Manufacturers
Association (ALMA) Spring
er or the beaming of infrared energy, an entirely new audio format will cc-
Symposium, Marriott Nashville
Why not design a micro nuclear teac- cur every hour. Then, fortunately, we Airport Hotel, Nashville, TN,
tor, the size of a battery, which could will have just enough time to play the USA. For information contact
supply the energy. Anyway, it seems new record before both the player and Carol Bousquet, ALMA execu-
like a good idea to ask the Society's the disc have become obsolete, tive director,tel: 978-448-5658.
1999 May 8-11: 106th AES Con-
_F_F__Ull'_O_ vention, Munich, Germany,
MOC Center, Munich. For infor-
m Kees A. Schouhamer Immink, a mationsee page 200.
native of the Netherlands, worked
at Philips Research Laboratories 1999 June 23-25: 9th AES Re-
from 1967 to 1998, and since gional Conference, Tokyo,
1996, has been an adjunct profes- Japan, Sunshine City Conven-
sor at the Institute for Experimen- tion Center, Ikebukuro. For
tal Mathematics, Essen University, information contact: Takeo Ya-
Essen,Germany. mamoto,conventionchairmanat
He has contributed to the tech- tel: +81 3 3489 4760 or e-mail:.
nologies of a variety of digital au- yamatake@t3.rim.or.jp.
die and video recorders such as the Compact Disc, Compact Disc Video,
DAT, DCC, and, recently,the DVD. Dr. Immink holds 36 U.S. patents,has 1999 September 2-5: 17th Inter-
authored or coauthoredthreebooks, and has written numerouspapers in nationalConference,Florence,
the field of digital audio recording technology. He received many awards for Italy, "High Quality Audio Cod-
his part in the digitalaudio revolution. He is a fellow of the AES, lEE, lng,"Villa Castelletti,Signa, Italy.
SMPTE, and IEEE, and was honored with the IEEE Edison Medal, the lEE See page 200.
Sir J. J. Thomson Medal, the SMPTE PoniatoffGold Medal, and the IEEE
Masaru Ibuka consumer electronicsaward. He was elected a member of 1999 September 10-14: IBC99,
the Royal NetherlandsAcademyof Sciencesfor hiscontributionsto science Amsterdam,the Netherlands.
in 1996. He serves as a govemor of the IEEE Information Theory Society For information contact Gina
andthe IEEEConsumerElectronicsSociety. Christison, IBC Office, Savoy
Dr. Imminkjoined the AES in 1980, has been a member of the Review Place,London WC2R OBL, UK.
Board of the Journalsince 1984. He was a committee member of the tel: +44 (0) 171240 3839 or fax:
Netherlands Section from 1986 to 1992 and its chairman from 1988 to +44 (0) 171 240 3724.
1990. He was electeda governorfor the 1995-1997 term, and vice-presi-
dent, Northern Region,Europe,for 1997-1999. He was made an AES fei- 1999 September 24-27: 107th
Iowin 1985 and was honoredwiththe Society'sSilverMedalin 1992.Email AES Convention,New York, NY,
address:immink@exp-math.uni-essen.de. USA. For information see page
200.
172 J.AudioEng.Soc.,Vol.47,Nb.3,1999March
Thesis
Full-text available
In this work we propose the development of a new method of digital audio mixing that aims to automate two main usual processes for the task of mixing musical instrument sounds. The practical purpose of this computational automation is to facilitate the sound engineer's work for the balance of individual musical instruments, improving the sound of the recorded music approaching the phonographic industry standards. Firstly, we discuss the perceptual aspects of audio, the sound manipulation equipment usually found in recording studios, as well as some aesthetic-musical aspects. As an original contribution to the field, we describe a new method of automatic mixing that aims the optimization of the musical dynamics and spectral distribution of digital audio. Our model is based on the approximation of the mixing spectrum to a reference spectrum via sound volume matrices and individual equalization for each instrument. Some proposed types of reference spectra are based on Fletcher & Munson's equal-loudness contours. The model, with all its modules was computationally implemented in MATLAB and eventually can be exported as a stand-alone program for use by musicians, technicians and sound engineers. This software achieves our propose of sound balance through the automation of three processes: a) optimization of the instrumental dynamics in the time domain, b) tonal / spectral correction of each recorded instrument, c) improvement of tonal / spectral balance of the final sum resulting from the mixing.
Thomson Medal, the SMPTE PoniatoffGold Medal, and the IEEE
  • J J Sir
Sir J. J. Thomson Medal, the SMPTE PoniatoffGold Medal, and the IEEE