Television programme sound can have a significant emotional effect on the audience and yet remain quite unnoticeable. Sound does not simply complement the image, it can actively shape how we perceive and interpret the image. Most TV crafts such as lighting, camerawork, editing use a form of invisible technique which achieves its effects without the audience being aware that any artifice has been employed. If the audience does becomes aware of the methods employed, they often become less involved in the production and may even suspect that the programme maker is manipulative. The production contribution of sound is usually the most unobtrusive and difficult for the audience to evaluate – until it is badly done. Visual awareness appears to take precedence over audible awareness and yet intelligibility, space and atmosphere are often created by sound. The selection and treatment of audio shapes our perception and can be used to focus our attention just as effectively as the selection of images.
What is Sound?
When there is a variation of air pressure at frequencies between approximately 16 to 16,000 Hz, the human ear (depending on age and health), can detect sound. The change in air pressure can be caused by a variety of sources such as the human voice, musical instruments, etc. Some of the terms used to describe the characteristics of the sounds are:
Sound waves are produced by increased air pressure and rarefaction along the line of travel.
Frequency of sound is the number of regular excursions made by an air particle in one second (see figure opposite).
Wavelength of a pure tone (i.e. a sine wave, see figure opposite) is the distance between successive peaks.
Harmonics are part of the sound from a musical instrument which are a combination of frequencies that are multiples of the lowest frequency present (the fundamental).
Dynamic range is the range of sound intensities from quietest to loudest occurring from sound sources. This may exceed the dynamic range a recording or transmission system are able to process without distortion.
The ear’s response: to hear an equal change in intensity, the sound level must double at each increase rather than changing in equal steps.
Decibels are a ratio of change and are scaled to imitate the ear’s response to changing sound intensity.
Loudness is a subjective effect. An irritating sound may appear to be a great deal louder than a sound we are sympathetic to (e.g. the sound of a neighbour’s cat at night compared to our own practice session on a violin!).
Phase of a signal becomes important when signals are combined. Signals in phase reinforce each other. Signals out of phase subtract from or cancel out each other (see figure opposite).
Pitch is the highness or lowness of the frequency of a note.
The time taken for a complete cycle of pure tone (A) to begin to repeat itself (B) is the frequency of the signal and is measured in cycles per second (Hz), e.g. 50 Hz = 50 cycles per second. Frequency is inversely proportional to wavelength. For example, a high frequency sound source of 10,000 Hz produces sound with a short wavelength of 3.4 cm. A low frequency sound source of 100 Hz produces sound with a longer wavelength of 3.4 m. Frequency multiplied by wavelength equals the speed of sound (335 m/sec) in cold air. It is faster in warm air.
Phase
Acoustics
Reverberation relates to the time delay before sounds reflected from the wall and other surfaces reach the microphone.
Standing waves effect is due to the room having resonances where the parallel walls enhance certain frequencies.
Choosing which microphone to use in a specific production environment will require consideration to be given to some or all of the following factors affecting a microphone’s performance:
nature of the sound source (e.g. speech, pop group drums, bird song, etc.)
matching the technical characteristics of the microphone to the sound source (e.g. frequency response, transient response, ability to handle high/low levels of sound (sensitivity), and directional properties)
mechanical characteristics such as size, appearance, robustness, wind shields, affected by humidity, stability, reliability, etc.
compatibility – cable plugs, connectors and adaptors, matching electrical impedance, cable run required, interface with other audio equipment
powering arrangements (see condenser microphone below)
programme budget, microphone cost/hire and availability.
Frequency Response of a Microphone
Microphones convert acoustical energy (sound waves) into electrical power either by exposing one side of a diaphragm (pressure-operated) to air pressure variations or by exposing both sides of the diaphragm (pressure-gradient). Directional response of the microphone will depend upon which method is chosen or a combination of both and the physical design of the microphone. Response of the microphone is also related to frequency of the audio signal. Directional response can be plotted on a polar diagram which places the microphone in the centre of a circular graph indicating the sensitivity at each angle with respect to the front axis of the microphone (see polar diagram opposite).
There are three basic types of microphone: moving coil, ribbon, and condenser.
The moving coil: The polar diagram of this microphone can be omnidirectional or cardioid, i.e. having a dead side to the rear of the microphone. This type of microphone can be used as a reporter’s ‘hand held’ but care must be taken in its handling.
The ribbon: This microphone’s polar response is ‘figure of eight’ – open front and rear, but closed to the sides. This microphone has been used mainly in radio studios with the interviewer and interviewee sitting across a table facing one another.
The condenser: The condenser microphone achieves the best quality of the three and requires a power supply to make it operate. Initially the power was supplied by a separate mains driven unit but this was superseded by an ‘in line’ battery supply. Today most condenser microphones will be powered directly from the audio mixer unit, the supply being known as the 48 volt phantom power. There are other forms of condenser microphone power supply known as 12 volt A/B and ‘T’ power. Always check that the mic power supply matches the microphone to be used before connection.
Sound intensity range: The intensity of sound is a function of the pressure variation set up in the atmosphere. Intensity is proportional to pressure squared. A microphone is used to convert sound energy into electrical energy and the voltage produced is proportional to the sound pressure. The range of intensities from quietest to loudest is the dynamic range of that sound situation. For example, the ratio of the loudest sound produced by an orchestra to the quietest sound can be as much as 60–70 db in an orchestral performance. This dynamic range is greater than can be transmitted and therefore sound control and possibly sound compression is required when recording large variations in sound levels.
Decibels: Our ears do not respond to changes in sound intensity in even, linear increments. To hear an equal change in intensity, the sound level must double at each increase rather than changing in equal steps. To match the ear’s response to changes in sound intensity, it is convenient to measure the changes in the amplitude of the audio signal by using a logarithmic ratio – decibels (dB). Decibels are a ratio of change and are scaled to imitate the ear’s response to changing sound intensity. If a sound intensity doubles in volume then there would be a 3 dB increase in audio level. If it was quadrupled, there would be a 6 dB increase.
Zero level voltage: Just as light is converted into a TV signal with a standard peak amplitude of 0.7 V so sound energy when it is converted into electrical energy requires a standard signal, a baseline voltage to which all changes can be referred. This is known as zero level and the standard voltage selected is 0.775 V – the voltage across 600 Ω (a common input and output impedance of audio equipment) when 1 mW is developed. 1000 Hz is the frequency of the standard zero level signal. Increasing or decreasing sound intensity will alter the level of this voltage.
Sound Monitoring
The first and last evaluation of sound quality is by ear, but meters are a valuable aid in monitoring signal strength to prevent overloading the audio system’s maximum permissible transmitted or recording level. Beyond this point the signal will be progressively distorted without increasing the perceived loudness. The relative levels between each sound source must be balanced by careful monitoring. The mix heard on loudspeakers, or headphones should be used to establish the correct balance of the various sound sources, using meters as a guide to ensure that the technical requirements are correctly fulfilled. Two main types of meter are used to measure and monitor changes in audio level. A peak programme meter (PPM) measures peaks of sound intensity. A programme volume meter (VU) measures average level of sound and gives a better indication of loudness but at the expense of missing brief high-intensity peaks that could cause distortion.
VUs (Volume Units)
VU meters give an average reading of signal volume and are calibrated from –20 to +3 dB (some may have extended scales) and are good for reading continuous signals such as tone, and the zero position may be used for line-up.
Care is needed to interpret programme material and it is difficult to read low level signals.
Some high peak, low energy material may cause distortion before any real level is shown on the meter.
VU meters do not respond well to short transients signals and may indicate as much as 10 dB below the actual level. Many transient sounds will barely move the meter. Be careful when judging the recording level with VU meters as their slow response time may prevent them from showing short duration peaks present in voice recordings (especially female voices). It is good practice to peak 3 to 6 dB below zero VU if limiting of the signal is to be avoided.
Decay and rise time are identical and may be too fast to read the signal.
Being relatively cheap to manufacture, many sound sources can be individually metered avoiding switching signals to a single meter.
As a guide to help prevent over-modulation, speech may be held back to –6 dB or lower (50% mod.), some transient musical instruments to –10 dB, loud crowds or compressed material can reach 0 dB.
Ensure signals do not peak beyond 0 dB (100%) to avoid distortion.
PPM (Peak Programme Meter)
The PPM is designed to have a very fast rise time, allowing short duration peaks likely to cause overloads to be clearly seen but give little indication of the perceived loudness of the material.
PPMs are designed with different scales (e.g.
BBC, EBU, IRT/DIN, etc.), but all have the same mechanical characteristics. The BBC version, for example, is calibrated from 1 to 7 with 4 dB steps between graduations on a linear scale.
PPM 4 is used as a reference line-up, representing a signal of 1 mW (0.775 V into 600 ohms). PPM 6 represents the maximum allowable signal.
PPMs are very expensive as the mechanical construction has to be very accurate and requires associated electronics to enable the parameters to be adjusted.
The audio may be allowed to peak up to 6 but needs to be controlled according to the programme material and its loudness, monitored on loudspeakers.
Speech should be normally peaking 5 with only occasional peaks to 6, some acoustic instruments might be allowed to peak 5 as well but more complex music mixes will need to be lower, compressed music (e.g. rock bands) will need to be held to a maximum of 4, small crowds may peak 5 but very energetic ones only 4.
Background effects behind a voice should be no greater than PPM 2 to 3.
Compression can be seen on the meter as the range of levels becomes reduced and loudspeaker monitoring will require close attention.
A television production usually involves combining a number of visual and audio sources. The vision mixing panel (see Introduction to vision mixing, page 170) switches the visual sources whilst the sound mixing console is designed to handle all programme audio sources. To process a variety of different audio inputs during a live transmission or recording into a single controlled output signal, the sound mixer will require some or all of the following facilities:
Audio assignment assigns an audio input to its associated channel by means of a jack field, matrix or routeing switcher.
Input select and gain: Input signals to the desk may originate at line level (e.g. tape machines or incoming lines) or low level output signals from microphones with the availability to adjust the gain depending on the type of microphone in use.
Routeing: Each channel output can be assigned to a group fader or direct to the main output fader.
Channel fader allows individual level control and audio processing of each sound source. A mono channel will be fitted with a pan control to allow the signal to be positioned at the required image position (see Stereo, page 94). A stereo channel has two identical paths to allow processing of left and right signals and a balance control to adjust the relative levels of the left and right signals.
Group fader allows all assigned sources (usually grouped for a common production purpose) to be combined in whatever proportion required.
Master fader controls the combined output of the sound desk and will normally be set at a calibrated ‘zero’ position or at 10 dB below maximum.
Auxiliary outputs are additional outputs to the main desk output and allow for foldback feeds, etc.
Equalization is the process of adjusting the frequency response of the signal usually banded into control of high, medium and low frequencies. Switched filters are used to remove low frequency rumble, etc.
Monitoring of the output of the audio signal is achieved by means of high quality loudspeakers installed in a good, acoustically adequate, listening area (i.e. the room should have no significant effect on the sound). As well as the final production mix, provision is often required to allow monitoring of network audio and/or other linking audio sources to the production. The correct output of signal level is checked by an audio level meter (see figures on page 83). Meters may also be fitted to group faders, channel faders or switchable to a selected combination.
Pre-fade listen (PFL) is the facility on an individual channel to pre-hear the designated audio source without affecting the mixer’s main output.
Limiters prevent the signal from exceeding a predetermined level.
Compression: The dynamic range of some sound sources (e.g. an orchestral concert) can exceed that which can be transmitted. If manually riding the level is not always feasible, sound compression can be used. The aim is to preserve as much of the original dynamic range by judicious use of ‘threshold’ (the level at which compression starts); ‘slope or ratio’, which controls the amount of adjustment; and ‘attack and release’, which determines the speed at which the compression equipment will react to changes.
Reverberation uses digital delay to the audio signal to simulate the effect of audio echo. Audio delay is also required when a visual source has been delayed by a frame store or satellite link. The sound is delayed to match the timing of the image.
Audio record and playback facilities available to a sound desk include, gram decks, / inch tape, mini disc, audio cartridge machines and telephone hybrid. This allows an interface with the sound desk, compensating for line resistance and received levels, of a two wire telephone system.
Other facilities available from a sound console include phantom power supply for condenser microphones, communications and talkback to audio source originators for production purposes.
A digital console converts analogue inputs into a digital signal to be combined with the input from digital devices. Signal routeing between audio source and channel can be displayed on liquid crystal displays (LCD) or computer screens. Menus display information about all operational aspects of the desk and allow adjustment of any audio parameter. Sometimes this can be less flexible than manual adjustment if the required control can only be accessed by way of a number of menu page selections.
Loudspeaker Monitoring
Assessing audio quality requires:
quality loudspeakers that produce a flat frequency response and can reproduce fast transients or changes in level.
a quiet listening environment with suitable acoustics that have no significant effect on the sound.
a clear understanding of how to interpret the chosen metering method (see page 83).
A television studio requires isolation from external sound, a reverberation time suitable for the production content, and sets which are designed to be compatible with good audio pick-up (e.g. the avoidance of curved hard surface backings). Ventilation noise, talkback leakage, equipment and staff movement during production can all add to unwanted background noise.
Audio Rigging
Wall boxes around the studio are the means of connecting audio feeds and associated power to and from the sound control room. A number of dedicated microphone inputs, feeds for foldback speakers, production and sound control talkback are available on these panels. Care should be taken to avoid cross talk between low level signal output from microphones and the higher level foldback or talkback feeds. Use the nominated input/output on the box with the correctly screened cable and tie off the cables at the wall box to avoid disconnection if the cable is accidentally pulled.
Cabling
Start at the microphone position and run the cable back to its wall box or ‘break-out’ or ‘stage box’, a unit which can handle many microphones and take their feed on one cable back to the studio wall. Keep sound cables tidy and attempt to find cable runs across sets that are visually unobtrusive, are not endangered by other studio ‘traffic’, and do not run parallel or adjacent to lighting cables connected to dimmer circuits. If a microphone stand or boom is required to be repositioned during the production, make certain that surplus cable is neatly coiled and immediately available to be used as required. Any cable crossing fire lanes or entrances/exits should be ramped or flown (i.e. tied securely above head height).
Foldback
Programme sound not originating in the studio is often required. Pre-recorded inserts or reports, feeds from outside sources, etc., need to be heard by presenters and guests if they form part of a discussion. Music tracks or sound effects may form part of speech or action cue. The location of the speaker and the level of foldback are critical to avoid pick-up by live microphones. Musicians require foldback of amplified instruments and voice often at a level that may cause coloration and/or prevent other production staff hearing talkback. This type of foldback is often controlled on the floor from a separate mixer.
Audience Foldback
Many productions are staged in front of an audience who require to hear the content of the programme. In some discussion or game shows, members of the audience may also participate in the programme and appropriate techniques need to be employed to avoid howl-round.
The boom arm’s length (3–6 metres) is adjusted by a wheel (1). The microphone (2) can be tilted and rotated by a control lever (3), (squeeze for microphone tilt, turn for microphone turn).
The overall height of the boom arm and platform can be hydraulically adjusted (4) between 1 and 1.5 m. A pump handle (5) can be used to recharge the hydraulic cylinder. The boom pram is steerable with wheel (6) and can be held stationary by a brake (7). The back wheel drive can be disengaged (8) from the two front wheels to allow the boom to be manoeuvred. (9) Talk-back microphone, script board, programme junction box, mini-monitor. (10) Seat. (11) Platform allows operator 360° movement.
Clean Feeds
If a discussion or interview is conducted with individuals who are located away from the main studio, all participants need to hear each other without hearing their own voice delayed. This requires an elimination in the recipients’ fold-back of their own contribution. Clean feed is the output of the audio desk minus one or more of the sources going into it. In the USA, clean feed is more accurately called ‘mix minus’ and can be derived from a purpose designed clean feed output from the desk group outputs, or it may be provided for each channel on the sound control desk. This becomes more complex where there are many remote studios (e.g. an election coverage), requiring a number of clean feeds, allowing each location/studio to hear all the other sound sources.
XLR Connectors
XLR (eXternal, Live and Return) type connectors are used for all balanced microphone connections and balanced high level signals. The connector has three pins – earth, live and return – and follows the convention that cable connectors with holes are inputs to the mixing/monitoring point. Connectors with pins point in the direction of the signal source.
There are two broad categories of audio recording:
analogue – where the magnetization of a tape varies with the amplitude of the audio signal
digital – where the audio signal is converted into a digital code (see pages 38–39) and then recorded.
Analogue Recording
When an electrical current flows through a wire conductor it produces a magnetic field around the conductor proportional to the original current. This magnetic field can be intensified by wrapping the wire around magnetizable material to form an electromagnet. If the electric current flowing through the electromagnet has been created by an audio signal and a specially coated tape capable of being magnetized is driven past this varying magnetic field at constant speed, the fluctuating magnetic field will be imprinted onto the tape in proportion to the original signal’s amplitude and frequency. This method of recording sound has a number of shortcomings and solutions:
Tape has poor magnetic retention and does not record in a linear fashion. A fixed level bias frequency (100–200 kHz) is added to the audio to ‘bias’ the tape to a more linear part of its operating range.
Tape hiss due to the finite size of the individual tape particles can be corrected by noise reduction circuits (see opposite).
Pre-emphasis and replay equalization ensure a flat frequency response output to compensate for losses in the magnetic recording/replay process.
Head alignment is important to ensure the head gap is at right angles to the tape, otherwise an azimuth error may cause a loss in the recording or replay of high frequencies.
Magnetic saturation is reached when all tape particles are fully magnetized and any further increase in signal level will fail to produce a change in recorded level. Maximum recording levels must be controlled to avoid trying to produce a magnetic field that the system cannot achieve.
The speed and tension of the tape must be constant otherwise wow and flutter will be detected in the replayed signal. In broadcasting, tape machines are often required on cue, to have instant start-up to correct operational speed.
Tape heads and guides need to be clean to avoid a build up of tape coating deposited on the heads affecting the machine’s performance.
Tape drop out is caused if there are irregularities in the tape coating.
Off-tape monitoring or a replay check on a recording and the use of high quality tape are the only methods of avoidance.
Print through is caused by a modulated tape inducing its magnetism into an adjacent level of tape resulting in a pre-or post-echo. Storing master tapes ‘tail out’ (i.e. not rewinding after recording or replay) is a standard industry practice to reduce this degradation.
Noise such as hiss, hum and other interference may exist alongside the required analogue audio signal. If the recorded tape is dubbed onto another tape the problem is increased on each recording. Methods of reducing this unwanted portion of the recording include:
Correct levels of recording and replay are always applied.
Pre-emphasis amplifies the high frequency portion of the wanted signals before recording to allow a reduction of high frequencies on replay, reducing high frequency noise whilst leaving the original audio unaffected.
Dolby A divides the signal into four frequency bands having its maximum effect when the level is 40 dB below reference. The system provides between 10 and 15 dB of noise reduction.
Dolby B is primarily a system of equalization with some compression applied to the signal. This can increase the signal-to-noise ratio of domestic cassette to about 60 dB.
Dolby C uses a larger compression and expansion ratio as well as frequency adjustments and can produce an improvement of around 10 dB over Dolby B.
Dolby SR (Spectral Recording) system uses ten separate bands of frequency processing operating signal compression that varies with its level. The result can be a reduction in noise of approximately 25 dB.
DBX uses compression across the whole frequency range. There is frequency pre-emphasis and corresponding de-emphasis on replay, the whole system providing up to 30 dB of improvement in the system signal-to-noise. There are two versions of the system, type 1 installed in professional equipment and type 2 used in the domestic market.
Noise gates define a set level below which the signal gain is reduced or even cut altogether. This process can be very unpleasant if not used with care. One example might be the feed to an audience PA system. Here, setting the gate threshold to around 20 dB (ref. 0.775 V) would ensure the speakers had no output unless the signal exceeded this level, thus reducing unwanted hum from the speakers.
Digital audio recording
Video recording resolved the problem of recording a much higher frequency spectrum than analogue audio on tape by means of a rotating record/replay head as well as moving the tape. Digital stereo audio needs to record at 1.4 Mhz (see Digital audio, p. 92) and the DAT (digital audio tape) system was developed to meet this criteria. For editing purposes, an identical analogue audio track is added (sometimes called cue track) as rocking a slow moving digital signal past a head produces no sound, unlike an analogue signal.
The frequency response and signal-to-noise ratio of digital recorders are superior to analogue machines and repeated copying results in far less degradation. Drop out can be corrected and print through has less significance or is eliminated. Wow and flutter can be corrected by having a memory reservoir system (see Frame store, p. 194) from which data is extracted at a constant rate eliminating any variation in the speed at which it was memorized. In general, when a signal is converted to digital, there is more opportunity to accurately rebuild and rectify any imperfections and to allow signal manipulation.
Audio and Programme Production
The selection and treatment of sound in a production shapes the perception of the audience and involves a range of audio techniques. These include:
Adjustment of loudness in order to emphasize production content; to indicate priority (e.g. dialogue louder than background traffic); to exploit the dynamic range of sound (e.g. quiet contrasted with loud passages in orchestral music).
Pitch is a subjective quality of the frequency of sound vibrations and each frequency band, similar to colour, has associated feelings.
Timbre is the tonal quality of a sound as judged by the ear. It describes the texture or feel of a sound, and can be manufactured for specific production purposes (e.g. the use of distort for a telephone voice).
Reverberation is the reflections from surfaces and can be artificially created to suggest environment.
Fades and mixes mark the transition between sound sources, production sections or complete programmes. The rate at which a sound dies away can be highly evocative.
Sound effects heighten realism, add space and atmosphere, reinforce the action, and can guide the audience to a new situation or scene. Combining sound effects can create an aural impression that is greater than its parts.
Smooth transitions between contrasting images can be created by effects or music.
Anticipation or preparing the audience for change can be achieved by sound leading picture (e.g. the sound of the incoming scene beginning at the end of the outgoing scene).
Sound perspective focuses the attention of the audience on the visual space depicted (e.g. close-up sound for foreground action, distant sound for long shot). Sometimes dialogue clarity is more important than the realism of sound perspective. For example, the sound level of orchestral instruments in close-up are not normally boosted to match picture content. The overall orchestral balance is an integral part of the performance and therefore takes precedence.
Off-stage sound suggests space and a continuing world beyond the frame of the image. It can also be used to introduce or alert the audience to new events. Sound from an unseen source creates space and mystery.
Narration voice-over acts in a similar way to the thoughts of a reader.
Careful balance between voice, effects, and music combine to give a powerful unity and authority to the production.
Music is a sound source that has immense influence on the audience’s response. It can create pace, rhythm, emotion, humour, tension, and requires a very subtle balance to weave the music in and out of effects and speech in order to create unity in the final production.
Silence is an attribute of sound that has no visual equivalent. It emphasizes or enhances a production point but only in a sound film can silence be used for dramatic effect.
Radio microphone equipment
1. Transmitter.
2. Receiver.
3. Transmitter top panel.
4. Receiver front panel.
5. Hand held radio microphone.
It is important to have a basic understanding of how audio is converted into digital information so that errors are not made, particularly when recording on digital camera/recorders.
Sampling
The audio signal is sampled at a constant rate like a series of pulses. For each of these pulses, the level of the sound is checked and this value is given a number. This number is transferred into binary code and this code is recorded as the digital information. The actual recording is therefore a string of codes, each representing the level of the audio at a particular moment in time. In order to achieve a good frequency response, the sampling is carried out at a very high frequency; 48,000 samples per second (48 kHz) is used for most professional systems. The level of the sound to be converted into binary numbers must also have sufficient steps to ensure accuracy, this is the quantization. With the 16 bit system employed by most camera/recorders, 65,536 finite levels can be represented. Put simply, this means that in order to encode and decode the signals, all the systems must operate at the same sampling rate and the signal level must never exceed the maximum quantization step or no value will be recorded.
Digital Audio Meters
The audio meter on a digital camera is very different to that on an analogue device as it shows values up to the maximum signal that can be encoded. This is calibrated from infinity (no signal) to zero (maximum level). It will also be noted that a line-up level (–20 dB) may be marked with a solid diamond shape or square on the scale (see figure opposite). This –20 dB point is used as the reference when aligning an external device such as a separate mixer. When taking a level prior to recording, it is safe to allow peaks to be well below 0 (maximum level). –12 dB to –8 dB is fine for normal peaks, ensuring the signal is properly encoded and the safety margin allows for any unexpected peak. Limiters are built into the system to prevent loss of signal should levels become too high, but as signal noise is not a problem (unlike analogue systems) holding the levels down is acceptable.
Digital Camera/Recorders
Most professional systems have four digital audio tracks. These all have the same specifications and can all be recorded with the picture or used for insert editing. The camera/recorder will have the ability to select any of these as mic or line inputs and will provide phantom voltages for the mics if required. On many digital camera/recorders, audio input/channel selections are not made with switches on the camera but are enabled via a menu displayed on an LCD screen positioned on the side of the camera. Many allow for other adjustments like audio limiters or the introduction of low frequency filters to help reduce unwanted noise at the time of the recording.
On Camera Digital Audio Meters
Viewfinder displays are different to analogue camera/recorder viewfinders. In the DVCPRO system, for example, there are indicators for –40 dB, –30 dB, –25 dB, –20 dB, –15 dB, –8 dB and 0 dB.
The Sony DVW-700 series uses the following display:
Note that unlike analogue systems, where signals above the maximum permitted level become progressively more distorted, digital signals do not exist above the maximum quantized level. The digital meters are calibrated from infinity (°) to 0 (maximum level). Many have a calibration mark at –18 dB which is intended to be aligned with tone at zero VU. In fact some signals may produce digital clipping when aligned in this way and in the UK many users align 0 VU to –20 dB which places the zero level of 0.775 volts at –24 dB.
Cue Track
As well as the four digital audio tracks, the digital formats also allow for a longitudinal audio cue track. This can be used for time code or have a guide version of the sync., or feature audio to assist in editing. This track will be easier to listen to when the tape is being played at a non-standard speed whereas digital tracks will be decoded in bursts, omitting material when played faster than standard and repeating material when slower.
The cue track can be recorded at any time, either with the picture or inserted separately. The information to be recorded on the cue track is normally set via the set-up menu in the viewfinder or on the camera/recorder output.
On the higher specification camera/recorders, a return output of audio from the camera/recorder to separate audio mixer no longer needs to come from the tiny headphone socket. Instead it is fitted with a male XLR socket on the rear of the camera. The track or mix of tracks and the level of the signal to be fed from this socket may also be adjusted via the menu-driven software on the camera/recorder.
Normal hearing provides depth and space, perspective and separation, effects and atmosphere by 360 degree spatial hearing. Television mono sound originates from a single speaker but television stereo sound attempts to provide an approximation of this sound experience.
For effective stereo reception, it is recommended that the listener and the two stereo speakers are positioned at the three points of an equilateral triangle. Many stereo television sets fix their speakers either side of the screen (a distance of approximately three feet) with the viewer positioned six feet or more away. This limits the full potential of television stereo sound even though an effective stereo signal is transmitted. The stereo ‘sound stage’ that is transmitted is much wider than the screen size but it enhances the illusion of a ‘third dimension’ missing from a two-dimensional image.
Many television sets in use are not equipped with stereo sound and therefore stereo transmissions must be compatible with mono reception.
Locating Stereo Sound
Effective stereo sound requires the ‘on screen’ sound to be augmented with ‘off screen’ sound. Stereo adds to the screen image sound, background sound located left or right of the screen whilst still being compatible with mono listening. If approximately 15° is the central wedge of sound image on screen, everything else that does not have an obvious geographical relationship to the picture is ‘elsewhere’.
Microphone Placement
A stereo effect will occur if the time of arrival of an audio signal at two coincident microphones is different and this difference is faithfully reproduced on two loudspeakers. This inter-channel difference can also be artificially replicated from a mono source by the use of a pan-pot which determines which proportion of the mono signal is directed to each stereo channel.
The angle of acceptance of the two microphones, which are usually positioned 90 degrees to each other, will depend on their polar diagrams and their position relative to the sound source. Similar to the angle of view of a lens (see The zoom lens, page 50), the closer the pair of microphones are to the sound source, the greater the sound image will fill the region between the monitoring loudspeakers. The sound can originate from a point source such as a voice (excluding reverberations, etc.) or occupy a wide physical area such as a symphony orchestra. As television stereo sound pick-up is normally required to match the field of view of the displayed image, the width of the sound image and the polar diagram of the microphones are important. For example, a figure-of-eight ribbon microphone close to a camera will pick-up sound in front as well as behind the camera.
The M and S System – Using Cardioid and a Figure-Of-Eight Microphone
The middle and side (M and S) pair of microphones above consists of a cardioid facing forward, providing the M (middle/main) signal or centre image and a figure-of-eight microphone providing the S (side) signal.
Advantages of the M and S technique in news and documentary production are:
The proportion of side component can be decided and adjusted away from pressure of location work.
A good mono signal is available if stereo problems occur in post production.
The middle forward-facing microphone gives a clear indication of the direction the microphone is facing.
The main disadvantage for location work is that the S microphone (a figure-of-eight response) is prone to wind noise and rumble.
Stereo on Location
Points to consider when recording stereo sound on location exteriors when there is likely to be less control of placement and choice of microphone are:
Are the left–right position images correct relative to the picture?
Are there any time delays between picture and sound?
Are there any time delays between the left and right legs of a stereo signal producing a hollow sound, lacking clear imagery?
Do the sounds appear at the correct distance from the camera (perspective)?
Is the relationship between the primary sound sources and the background effects satisfactory, and is the effect of the acoustic suitable?
Does the sound match the picture?