That Radio Network Sound

by Fred Krock

Back in the days before satellites, radio network broadcasts had a certain characteristic sound. Every beginning radio announcer's dream was to work for a network some day. Most despaired of ever developing that network sound in their voice.

Network Announcer What they didn't realize was that the network announcers didn't have that characteristic mellow sound in their voices either. That sound came from the telephone company transmission, not from the network announcer's throat.

Listeners sometimes were amazed when visiting Los Angeles or New York how different some of their favorite network personalities sounded when the program originated locally. Even today a quick listen to one of the golden age of radio recordings reveals whether that recording was made where the show originated or whether it was recorded on the end of a network line.

Frequency response was not a major problem in those days of AM broadcasting. Radio network lines had frequency response up to 8 kHz. Telephone company customers paid according to the amount of bandwidth used. After World War II the networks cut back to 5 kHz lines to save money.

Frequency response was essentially flat to 5 kHz. At 5,100 Hz it was 30-50 dB down.

A few stations in extremely small markets used 3.5 kHz circuits. Networks paid for delivering programs to most affiliates. If the market were too small to be worth the expense, the station had to pay for the circuit from the nearest network access point. The cheapest circuit was 3.5 kHz. In a few other cases that was all the telephone company could provide into remote areas.

Considering all the things that were happening to the sound during network transmission, what is amazing is that it sounded as good as it did.

The amount of degradation was a function of distance.On the west coast, programs from Chicago sounded better than those originating in New York.

Even a relatively short transmission distance would impart noticeable network sound. I was surprised to hear it on a network newscast I read in San Francisco rebroadcast from Chico, California, a distance of 183 miles by road. It was audible even on a car radio.

At one time the telephone company played a major role in radio broadcasting. Virtually all studio-to-transmitter circuits and most remote lines were provided by the telephone company. The FCC would not license radio links for broadcast use unless the station could demonstrate that the telephone company could not provide service. Those few radio studio-transmitter links authorized usually were to FM transmitters on remote mountain tops where the telephone company could not provide an equalized 15 kHz circuit.

Only in larger cities did the telephone company provide a facility dedicated to broadcast circuits. In smaller markets you were lucky if you could find a test board man who even knew where the broadcast circuits were located. Often station engineering personel had to show the telephone company installer how to equalize a broadcast line in small markets.

The operative word is man. In the 1950's and 60's, I never heard a female voice while talking to any telephone company technicians.

Broadcasters referred to the telephone company broadcast circuit test board as toll. When television broadcasting began, the same telephone company crew handled pictures as well as sound. Later the television duties were split off to what was known as TOC for television operations center. Audio circuits were handled by what was renamed Audio Operating Center (AOC). We still called it toll.

Most telephone company employees belong to the CWA union. In San Francisco broadcast toll employees belonged to IBEW.

Stations usually bought one full-time circuit from toll to the station for incoming remote broadcasts. Then a circuit would be bought from the remote site to toll. Circuits between telephone company central offices could be bought by the quarter-hour as needed. The crew at toll would patch the various circuits as scheduled which saved stations a lot of money. The telephone company did not charge extra for this service at that time.

A lot of remote lines were routed via toll even though a shorter path might have existed. This allowed quick access by trained personnel in case of trouble.

Circuits were bought either as transmit or receive. Since passive equalizers were used, if no amplifiers were in the circuit, audio could be fed in either direction. Equalization was not perfect when audio was fed in the wrong direction, but it was better than no audio at all in case of a line failure. After the telephone company switched to active equalizers, this emergency backup capability was lost.

In 1958 the station where I worked became the Mutual affiliate in San Francisco. In addition to the network audio circuit, the telephone company installed a ringdown telephone to toll. Pick up that telephone twenty-four hours a day and someone answered at toll.

Ringdown telephones were supplied free to all major market network affiliates. About ten years later the ringdown was disconnected after a telephone company budget cut.

In 1960 the station became the west coast hub for the Mutual network. Our job was to time shift commercials in network newscasts, insert regional commercials in newscasts, and to supply all service to the west coast until 11 PM Pacific Time after the eastern network went goodnight at 9 PM.

Laxative spots always were shifted. Laxative spots at meal time brought listener complaints. A spot fed at 9 PM Eastern Time, a prime time for a laxative account, arrived on the west coast at 6 PM dinner time. Other accounts paid a premium for spots to run in drive time. They were delayed three hours.

Some spots were tape delayed from their earlier network broadcast. Most were played from transcription discs supplied by advertising agencies. For some reason Preparation H commercials always ran short and never fit into their holes properly.

Radio network circuits between New York and Chicago were called the round robin. They made a big loop from New York to Chicago and then back to New York. Any station within the round robin could feed the net.

Switching from one point on the round robin to another was instantaneous. The loop must be opened at the station which begins feeding. Occasionally an operator would forget to open the loop when starting a feed. The result sounded like a tape echo as the sound went around and around the loop until the operator woke up.

From Chicago to the west coast the network was one way westbound. The circuit could be reversed by the telephone company during a silent period so a west coast station could feed the nation.

Networks allowed thirty seconds for the telephone company to reverse the circuit.

Reversing the network was a major operation. All amplifiers in the circuit had to have their input and output connections reversed.

Starting in 1936 the telephone company would supply at extra cost customer controlled reversing equipment. Reversing the line between the west coast and Chicago caused about three seconds of dead air. Literally thousands of relays would throw. On air reversals usually were done only during newscasts. The east coast newscaster would say something like, "Now with a pause for switching we go to Los Angeles for a report from (name of newscaster)." Three seconds later the Los Angeles announcer would begin talking.

Mutual had discontinued customer control reversing between Chicago and the west coast long before we became the west coast hub. Mutual did use customer controlled reversing between Los Angeles and San Francisco. Some newscasts were fed to the west coast from KHJ in Los Angeles.

Network reversing control equipment at the station occupied two rack units. It had a small two- position rotary switch and red, white and green lights. The same type of lights and switch were used on telephone switchboards built by Western Electric.

The switch turned on phantom power on the network line. This control voltage was repeated from each amplifier to the next all the way to the far end of the network line. If neither end were feeding control voltage, a white light was displayed on both ends. This indicated that the network was unlocked and could be switched to feed from either end. The network audio path did not reverse until the receiving end began to send control voltage. If one end had control the transmitting end displayed a green light while the receiving end displayed a red light.

The network could be reversed only when the white light was on. If the receiving end turned on the switch, nothing would happen while the red light was on.

A few seconds before a hot switch the transmitting end would turn off the control voltage. Ideally the white light would come on at the receiving end at same instant the switching cue ended. Half of the switching time was required for the white light to come on at the receiving end. You didn't want to drop the control voltage too soon because a lightning strike or other disturbance along the line could cause a premature reversal. When the receiving end heard the cue and saw the white light, the operator would turn on the control switch and cue the announcer after waiting for the network to finish reversing.

We would experience line trouble on the incoming feed from the east between once and twice a week on average. Sometimes the network would operate for a few weeks with no problems and then be followed by a dozen outages in a single week.

Much of the circuit was underground cable. It was subject to backhoe fade. A backhoe has been described as the perfect tool to find a buried cable.

The telephone company maintained spare circuits for use in case of trouble on the regular network circuits. These spares also were available for occasional use customers.

Our friends at toll took pride in restoring service very rapidly in case of trouble. Sometimes they had to re-route circuits half way across the country to make good service. Once after a major line failure somewhere in Nebraska, our network service was routed from Chicago to Dallas to Los Angeles to San Francisco. From San Francisco it was routed east to Denver to serve the Mountain Time Zone stations.

If the line failure were west of Denver, San Francisco was responsible for restoring service. If the problem were east of Denver, the problem was given to the AT&T office in Chicago.

The first place San Francisco toll would call when the incoming network line failed was Denver. San Francisco was always happy to let someone else solve the problem. The ringdown telephone at the station would ring and the voice at the other end would say gleefully, "The problem's east of Denver."

This led to a lot of friendly teasing between the station and toll. If we had problems with one of our local remote lines, say from Oakland across the bay to San Francisco, the problem always kept getting described as east of Denver.

The telephone company employees kept a log of all telephone calls involving trouble. At the end of the call they would ask, "How do you sign?" Your signature was your initials. Everyone used phonetics for their initials. I would reply "Fox King" for my initials FK.

Imagination ran rampant. One telephone company employee with initials SJ would sign Stump Jumper. Calls between telephone company employees were logged in the same way. If any question ever arose about who said what and when, the log would tell.

Today we get our network programs from satellites. The sound quality is nearly identical with a local origination. Toll as we knew it is long gone. Too bad when Galaxy 4 failed and the whole NPR network went dead that we couldn't pick up the ringdown telephone and let our friends at toll take care of the problem.


What did Mother Bell do to make radio networks sound that way? Network sound was degraded in five major ways:

  1. Harmonic and intermodulation distortion
  2. Group delay
  3. Ringing
  4. Noise modulation.
  5. Single sideband carrier transmission problems
The distortion was not surprising since the sound may have passed through hundreds of amplifiers on its way to an affiliate station. Since the frequency response was limited to 5 kHz, no second harmonics were heard from frequencies over 2500 Hz or third harmonics from any frequencies over 1667 Hz. Total harmonic distortion on a transcontinental broadcast line probably was in the 10% range. Limited frequency response kept it from sounding as bad as it was.

The circuit equivalent of a twisted pair, such as used by the telephone company, is a very large number of extremely small value inductors wired in series shunted by a very large number of extremely small value capacitors. The result is a low pass filter. The Telephone company would compensate for the high frequency loss by connecting a passive equalizer consisting of series capacitance shunted by inductance. The result is relative phase shift. When a large number of these circuits are connected in series, group delay will reach a very high value.

You might also recognize these equivalent circuits as similar to those used in delay lines. As a result network radio signals traveled across country well below the speed of light. Attempts to use existing radio network circuits to transmit audio for early network television programs resulted in loss of lip-synch in as short a distance as between New York and Washington, DC.

All that reactance in network circuits would cause a number of resonant frequencies in the circuit. A transient near the frequency of one of these resonances could excite the circuit into producing a damped wave at the resonant frequency. This was called ringing. The effect was audible on program material.

The telephone company frequently used companders. ( Compander = COMpresser-exPANDER) The signal was compressed on the sending end and expanded at the far end. This increased the signal- to- noise ratio of the overall path. It also meant that the noise level went up and down as the signal level went up and down. The noise was not completely masked if the signal were primarily high frequency. A soprano voice or solo violin usually produced audible noise modulation. Even if the noise were masked, it caused the sound to become muddy.

The telephone company often used carrier circuits for long hauls. To allow the maximum number of circuits on a single pair, single sideband suppressed carrier signals were used. Carrier equipment was prone to all sorts of problems.

The most common problem with network radio feeds was what we called carrier whine. A continuous tone would appear between 30 and 40 dB below program level. Sometimes several of these tones would appear at the same time.

Even after the telephone company started using microwave transmission equipment, radio networks remained on the same old land lines they had been using for many years.

Copyright 2000, Fred Krock. All rights reserved.