The unbridled development of radio broadcasting in the 1920s and 1930s demonstrated the need for the standardization of audio equipment, studio-to-transmitter links (STLs), and static and dynamic audio signal measurement methods. Different concepts and solutions were developed on the two sides of the Atlantic, and today we still bear the consequences. Remember NTSC and PAL?
The power-matching concept
In the early days of broadcasting, there were many so-called standard audio reference levels, including 1mW, 6mW, 10mW, 12.5mW and 50mW. Bell Telephone introduced the concept of power matching. It wanted to develop reliable, high-performance STLs. It seemed normal to have an impedance-matched source (studio output), distribution link (cable) and destination (radio transmitter input) system. The matching of impedances to tight tolerances is necessary to avoid echoes on long cable lengths. In 1939, the standard reference level of 1mW into a 600Ω line was proposed. The result was a voltage of .77459V RMS. This reference level conforms to Bell Telephone's standards of limiting the signal level to a value that would produce a minimum of cross talk and provide a satisfactory signal-to-noise ratio (SNR).
After creating the reference level, the development of a new audio level meter was jointly undertaken by Bell Telephone, CBS and NBC. The result was the volume unit meter (VU meter), as well as the standardization of the reference level of 1mW, a unit that was adopted by the electronics industry. A standard operating level (SOL), also known as an alignment level of +8dBm into 600Ω, was initially chosen in North America. Some authorities, including sound recording studios, opted for a +4dBm SOL inside the plant. The SOL represents the steady-state maximum level or peak program level as measured with a VU meter.
The VU meter was developed primarily for the control and monitoring of audio programs. The specifications of the VU meter reflect the philosophy of the 1930s. Essentially, the VU meter is a moving-coil RMS-type audio signal level measuring instrument. It is fitted with two scales:
- A VU scale marked 0 (reference deflection) at about 71 percent maximum scale reading extending to +3 (maximum) and -20 (minimum).
- A percentage scale with 100 percent corresponding to VU reading.
The VU meter has an input impedance of 7500Ω and has a minimum loading effect on the 600Ω source impedance. Its sensitivity is adjustable such that the VU reference level (0VU) can be made to correspond to the SOL under steady-state sinusoidal audio voltage conditions. Its dynamic characteristics are such that, if a sinusoidal signal of a frequency between 35Hz and 10kHz and such amplitude as to give reference pointer deflection under steady-state conditions is suddenly applied, the pointer will take 0.3s to reach reference deflection. If this signal is suddenly removed from the input of the VU meter, the needle will take 0.3s to return to its initial position.
This characteristic was chosen in order to approximate the assumed response of the human ear. The 0.3s rise time and decay characteristic of the VU meter introduce a masking effect. Essentially, the instrument is unable to give accurate audio signal level indications under complex wave, fast rise time, input signal conditions. The instantaneous speech or music signal level may, in reality, be 10VU or more above the average readings of the VU meter.
Typically, if a device is designed to handle an SOL of +8dBm, it will be capable of supplying an output level in excess of +18dBm at a total harmonic distortion not exceeding 1 percent. Such undistorted audio peaks, unnoticed by the operator, are likely to reach the audio tape recorder or transmitter and overload it.
The situation is further complicated by FM audio transmitters, which use high frequency pre-emphasis with a time constant of 75µsec, resulting in a 14dB boost at 10kHz. Various types of limiter/compressor combinations are used in an effort to avoid transmitter over-modulation and achieve an acceptable SNR.
The voltage-matching concept
This concept is typical of modern studio installations. The signal source has 0Ω impedance. This impedance is raised to 50Ω by inserting a 25Ω resistance in series with each of the balanced cable conductors to avoid possible output transistor breakdown in case of a short circuit. The load is of the order of 10kΩ or higher. The signal level is expressed in dBu; 1dBu is equal to .775V RMS or the voltage resulting in dissipating a power of 1mW across a 600Ω load.
The voltage matching considerably reduces the power requirements of the signal source since it is required to dissipate only a minute amount of power across the bridging load. An added advantage is the improved frequency and transient response of the system. This is due to the fact that the capacitive loading of the shielded-balanced audio cable has a lesser effect across a source impedance of 50Ω than it has across a source of 600Ω. The interface with common carriers retains the power-matching philosophy to avoid return-loss problems with long cables resulting in echoes. The SOL in North America is +4dBu or +8dBu.
Figure 1. The masking effect of a typical VU meter and peak program meter (PPM) for a given tone duration. Click here to see an enlarged diagram.
The voltage-matching concept normally uses a peak program meter (PPM) for audio signal-level monitoring. Since the late 1930s, the European broadcasters — with the exception of France, where VU meters are still used — have been using some type of PPM. The PPM is a peak reading instrument capable of accurately displaying audio signal transients. The input impedance is bridging, meaning it is greater than 6000Ω. Some current designs feature a 10ms attack time (rise time) and a 2.65s fall back time. This characteristic amounts to a sample-and-hold approach to audio-signal-level monitoring. It allows the user to accurately monitor audio signal levels under steady state as well as program conditions and reduces the need for large amounts of headroom in amplifiers. Neither the scale nor the display is universally standardized. Some type of compression is required to reduce the dynamic range of the audio signal, which otherwise might exceed the transmitter and receiver capabilities.
Figure 1 shows that the PPM is capable of more accurately displaying audio signal peaks than the VU meter. There are, unfortunately, two entrenched camps steadfastly preferring the PPM or the VU meter. In an effort to satisfy all users, some contemporary equipment manufacturers offer equipment with selectable VU or PPM rise/fall times.
Figure 2. Upper-scale details of some audio level meters used throughout the world and the corresponding steady-state signal level in dBu. Click here to see an enlarged diagram.
Figure 2 shows details of the upper part of the display scale of some audio level meters used in various countries. It shows clearly that in addition to transit response differences, various organizations have different reference levels (SOL) and meter display scales. This situation creates problems in international television program exchanges and is not likely to change in the near future. The problem is complicated by the digital equipment that normally references all audio levels to the maximum signal level before clipping, which is identified as 0dBFS. All audio levels have, therefore, a negative value with the SOL set normally to -20dBFS, indicating that the equipment has 20dB headroom. Interestingly, the EBU suggests headroom of 18dB.
This new approach creates confusion with audio operators having an analog background and a strong attachment to the VU meter. Some efforts to change the reference signal from 0dBu = 0.775V RMS to 0dBV = 1V RMS have been poorly received by the audio community. So, until further notice, the reference audio level is 0.775V RMS, strongly related to 600Ω and 1mW. Old habits die hard!
Michael Robin, a fellow of the SMPTE and former engineer with the Canadian Broadcasting Corp.'s engineering headquarters, is an independent broadcast consultant located in Montreal. He is co-author of “Digital Television Fundamentals,” published by McGraw-Hill and translated into Chinese and Japanese.
Send questions and comments to:firstname.lastname@example.org