Dbx 1981 3

Page 1

THE SOUND ENGINEERING MAGAZINE

SEPTEMBER 1981 VOLUME 15. NO. 9

Broadcast Audio-
The Increased Need for
Audio Processing

ROBERT L. DEITSCH and GEORGE BERGER

Here, authors Deitsch and Berger provide us with a behind-the-
scenes look at the problems {and the solutions) faced by

engineers in the broadcast field.

HIS ARTICLE will deal with the basic concepts which

underlie the design of facilities for the ABC radio

broadcast studios in New York. Since I975, when a

slow but major reconstruction was begun, we have
argued with our friends about what we believe are the funda»
mental and philosophically different aspects of broadcasting
which distinguish it from the rest of the audio field. There are
to this day several misconceptions about what broadcast audio
is and ought to be. which still disturb audio professionals
and amateurs alike.

Is there any reason why that which is broadcast should be
identical to the original source? Well. there are two parts
necessary to the answer. The first. and we grant some nit-
picking. is that none of the records and tapes which are our
primary sources are unprocessed. and secondly. shouldnt the
broadcaster be granted latitude in the same sense in which
recording engineers use processing to compensate for the
limitations of their media or to create a uniqueness ofsound for
a group? We suggest that to answer no is intellectually dis-
honest. and thus we have ceased to feel any necessity to defend

Robert L. Deirsch is (hie-[engineer WPLJ/NY & pros
ject manager AM stereo ABC radio division, George
Berger is the senior design engineer ABC radio, NY

Reprinted by petmission at at: Magaztne. .7 1931 Sagamote Publishing Company. Inc

our position. We must make it understood. but the proofcomes
from that understanding a priuri.

Our somewhat dogmatic philosophy about our studios is that
an on-air control room should be as specifically tailored to
the format as possible. and that a production facility must be so
general and versatile that it never limits the creativity ofeither
the producer or technician. Technical equipment should be
of the highest possible quality. All of our installations since
I975 have been centered around custom-designed mixingdesks
from Rupert Nevet all reel~to~reel tape decks have been Studer
A-80 mastering recorders,

It would be unfair for the reader familiar with the two
ABC»owned radio stations in New York to be saying. What a
waste of good facilities: by the time I hear the product it sounds
very different." The AM station is WABC which. over the years.
has been the premier clear channel station in the US. ABCs
FM station in New York is WPLJ. which is the top-rated album
rocker in New York. The two stations provide some excellent
examples of the use of audio processing within our medium.

EASE VERSUS QUALITY

One of the problems broadcasters have faced since the early
sixties has been the conflict between the desirability of using
tape cartridges for ease of production and the degradation of
quality. especially in stereo. which seemed to be inherent in
using them. The two major problems were phasing and de-
gradation of signaHo-noise ratio. As early as 1973 we were
experimenting with a matrix encoding system to minimize

Page 2

(dbx ms»
('nxnpremm Limitcrs

Studto Ashly
switcher - 5077
crossover

tum
tum

Figure 1. WPLJs better mousetrap consists of an
Ashly SC-77 Crossover and six dbx model 165 "Over Easy
Compressor/Limiters. (A) Block diagram; (8) rack-
mounled equipment.

phase instability. In 1975 we decided to use dbx® noise reduc-
tion as well.

There has been much debate over the relative merits of dbx
and Dolby noise reduction. Because so much of the source
material we use comes to us already compressed, and also
because dbx offers a far greater amount of noise reduction to
precede the additional compression which we use. dbx was to
us a far preferable choice. Both of our stations now use this
combined processing system. This type of processing should
be transparent; the listener, either in the studio or at home,
should not be able to detect its presence.

The type of processing which causes controversy is com»
pression, limiting, and clipping, as well as any equalization
or reverb which may be added. It should be realized that the
sound of a radio station is an important programming
consideration. Psychoacoustic analysis is one of the elements
which enters into such decisions, but the broadcaster must also
consider competitive loudness and dynamic range.

Many listeners are in environments with high ambient noise
levels, such as automobiles; a station can ignore this part of its
potential audience only at its peril. On the other hand, if
listeners find the signal unenjoyable because too much
processing is added to fix this type of problem, audience
may also be lost, Consequently, the final decisions, within the
limits imposed on all broadcasters by the FCC rules and
regulations, must be made among what are still a scientifically
ill-defined set ofpossibilities, Having made those tough choices,
the broadcaster must enter into the area of receiver technology
in order to determine how to best accomplish those aims.

It is reasonable to look at the FM broadcast chain, including
the home receiver, as offering the potential of true repro-
duction. But it is by no means true that all FM receivers can
reproduce what is in the broadcast signal with true high fidelity.
Thus, even with FM, the broadcaster must attempt the clearest
possible definition of the audience. With AM, the waters
become much more murky. AM transmission is for all practical

purposes capable of equal fidelity! It is. of course. subject to
atmospheric and electrical interference, but the broadcast
signal is high fidelity. THD and lMD figures of less than one
percent are achievable. Signal to noise ratio approaches 60 dB.
Today. most AM receivers begin to roll off frequency response
at l kHz and are down more than 6 dB by 5 kHz.

In defense of receiver manufacturers (but certainly not to
agree with them), it is, of course, more expensive to build wide
response high fidelity AM receivers, Most firms claim that there
is no mass market for such equipment. We as broadcasters
counter that this is a matter of consumer education; there was
no immediate large-scale rush to FM equipment either.

So, we the broadcasters are left attempting to compensate
for the varying quality of receivers. both AM and FM, that are
in the hands of the public. Both of our New York stations play
popular" music. The problem faced by the technical staff to
meet the desires of the WABC programming depanment is to
provides a sound which makes WABC-AM sound as good on
the radio as an FM station. This means that we must compensate
for the poorer" AM circuitry in AM, FM combination radios
and receivers.

Through experimentation, we have found that no single
equalization curve can accomplish our purpose. The single
most obvious objection to a simple equalization attempt comes
from the announcers and diskjockeys. They feel that they don't
sound like themselves. They are correct, We have found that a
multiband compressor/ limiter system has been the answer.

Many of the ideas for use of equipment start at our visits to
Audio Engineering Society conventions. For example, we were
extremely impressed with the Over Easy® compression
curve used in some dbx equipment. Many years of experience
have taught us that the use of any single band compression
device does not yield the desired audio density without destroy-
ing the clarity of tone at the lower and higher ends of the
frequency spectrum. Such single»banded attempts typically
yield muddy drums and cymbals.