Audio accompaniment for TV programs

By Lavrenti Lishin, Andrey Kuripjatnik

Review contents:

Audio accompaniment for TV programs The audio measuring equipment from LES Company Headphones and sets from Beyerdynamic Systems from D.A.S. Audio Monitor systems from Digilab Electronics Sound monitors ESI Acoustic systems for audio control in TV production. Monitors from Event Electronics Active sound monitors of the Fostex Company Devices for the audio control from Genelec Company Sound monitors from JBL Pro Equipment of Klark Teknik Company Loudspeakers KRK Audio equipment M-Audio Audio equipment Mackie Marshall Electronics monitors Monitors and headphones from Samson Snell and Wilcox equipment for the embedded audio control Tektronix instruments for audio signal measurements Production of Wharfedale Pro Yamaha monitor loudspeakers

Now the majority of radio stations in Russia prepare the radio programs on the digital equipment that cannot be affirmed about preparation of audio track for TV programs. Apparently, it is connected by that until recently all TV channels have transferred the audio in the “mono” mode, therefore the requirements to preparation of multichannel audio track have appeared only with opening of broadcasting in “stereo” mode on the “The first channel” and on “Muz-TV” channel. At the same time the translation of numerous foreign motion-picture films are supplied, as a rule, by stereo- and/or multichannel audio, under preparation of TV programs results in necessity of multichannel digital equipment application and appropriate normative documents development.

Let’s consider the existing subjective and objective specifications and methods of evaluation of quality of TV audio.

There are two evaluation of audio quality: objective and subjective.

Subjective evaluation of the audio quality

Subjective evaluation of technical quality of TV production is based on principle of the information perception (information method). Here as word “information” it is necessary to understand the planned content of the TV audio. For evaluation the concepts of technical waste (under broadcasting and recording) and technical stop (under broadcasting) are used.

The technical wastes are the following:

• Inadmissible reduction of the TV audio quality by duration 5 sec and more;
• Incorrect level adjustment of audio channels of TV program. Under evaluation of correctness of the level maintenance it is necessary to be guided by the ratio of music and speech level established for TV under video recording and film post scoring (table 1);
• Delayed transition from the faulty equipment on the reserve exceeding established values;

The technical stops are the following:

• Full absence of audio on air during 5 sec and more;
• Off program audio during 5 sec and more;
• Unplanned pause between announcer’s notice and the beginning of transmission during 15 sec and more.

Technical quality of audio is evaluated according with five-point scale (see table 2). For the TV program audio quality evaluation the acoustical control of transmissions will be carried out. For this purpose in control room the two control units (for stereo broadcasting) or one control unit (for mono broadcasting) are installed. The TV program audio quality evaluation is determined according with the parameters given in table 3.

Units of the audio signal control and measurement shall provide the measurement of technical parameters both of signals in internal paths of equipment rooms, and signals of the broadcast standard established for the concrete TV organization (region).

Before the beginning of recording or translation of staged program the assignment of path rehearsal which purpose, except for improvement of actions of participants, is the reception of optimum audio quality, is recommended.

Objective evaluation of the audio quality

The objective evaluation of the TV audio quality is based on measurement of parameters of the TV audio signal under recording and transmission.

The parameters of the mark “badly” (waste) are the following:

• Background and noise by level -40 dB and more;
• PPM level overestimation from rating value more than on 1 dB (10%);
• PPM level underestimation on 3 dB (70%) and more: during voice information transmission by over 1 min duration; during voice art and publicist transmissions over 3 min duration; during music programs over 5 min duration.

Note. During musical and voice art transmissions the minimal PPM level should be not lower than -25 dB. The level reduction below given value by duration no more than 5 sec is supposed. At presence of the specified defects in transmission the mark “badly” (waste) is established.

The measuring devices used for objective audio evaluation

The general requirements are the following:

• Frequency range of measurement instrumentation — 20…20000 Hz;
• Module of full input resistance of measurement instrumentation should exceed the value of module of electric resistance of measuring circuit not less than in 20 times.

Further the requirements to the various equipment used for the control and measurement of the audio signal in TV production and broadcasting are given.

Voltmeters for measurement of signals of the sine wave form:

• Range of the measuring voltage — 0.1…40 V;
• Measurement error — no more than ±2.5%;
• Relative error of the voltmeter during measurement of the frequency response irregularity should not be more than one third of determined deviation of the frequency response.

The voltmeter for measurement of noise signals:

• Range of the measuring voltage — 0.1…40 V;
• Measurement error — no more than 4%.

The meter of harmonics:

• Range of measuring values — no more than 0.05…15%;
• Absolute value of the basic error — no more than ±0.05 HF (Harmonic Factor) + 0.05%, where HF is the value of final (top) limit of a scale on which sample is made, in percents.

The device for measurement of security from the weighted noise (psophometer):

• Should have the broadcast psophometric filter. Under application of devices with the detector of norm mean-square values should be reduced on absolute value on 5 dB;
• Range of the measuring voltage — 0.1…5 V;
• Basic error — no more than 6%.

Electronic accounting frequency meter:

• Range of output voltage — 0.1…5 V;
• Basic measurement error of frequency — no more ±0.1%.

Technique of measurement

Under measurement of the noise level there are some fine points. If to connect the voltmeter to output of the studied device, to measure noise, unfortunately, it will be not possible, as noise are rather small and is a portion of mV. By virtue of their small value even the most insignificant presence of other signals may affect in the radical mode to result on measuring value. Therefore it is necessary to connect an oscilloscope for the visual control of a signal to an output of the voltmeter. The modern equipment is completed with a set of transistors and microcircuits which noise spectrum leaves far beyond the audio range. However the voltmeter can measure everything, and its indications will be obviously increased. Therefore for reception of more objective data it is required to establish the filter limiting a frequency band by signals of audio range 20 Hz…20 kHz in a circuit in front of the voltmeter.

Table 1. The ratio of levels of music and voice under video recording and post scoring of TV production
Character of transmission	Voice (*)	Music (*)
Information program with music: without compression of speech (maximum level) with compression of speech on 6…8 dB	0…-2 dB (100…80%) 0 dB (100%)	-6…-8 dB (50…40%) -4…-6 dB (60…50%)
Art voice programs with music	0…-14 dB (100…20%)	no more than -4 dB (60%)
Music programs with announcements and comments	-2…-4 dB (80…60%)	0…-25 dB (100…5%)
Musical headbands, endings for all programs		-3…-6 dB (70…50%)**
Advertising without compression of speech (maximum level) with compression of speech on 6…8 dB	-2…-4 dB (80…60%) -2…-4 dB (80…60%)	-20…-10 dB (10…30%) -15…-12 dB (18…25%)
Simultaneous sounding speech-music:
Information programs with music without compression of speech (maximum level) with compression of speech on 6…8 dB	0 dB (100%) 0 dB (100%)	-20…-10 dB (10…30%) -5…-12 dB (18…25%)
Art speech programs	0…-14 dB (100…20%)	-14…-25 dB (20…5%)***
Alternation speech-music (on air):
Announcer’s notices before information programs	0 dB (100%)	no more than -6 dB (50%)
Announcer’s notices on all other programs	-2…0 dB (80…100%)	no more than -6 dB (50%)
Reserve music, headbands and endings: Maximal value (depending on saturation and sense) Minimal value		-3…-7 dB (70…45%) -25 dB (5%)
Notes: * Measurements will carry out with the meters of PPM level of audio signals according to a GOST 21185; ** On the top limit depending on saturation; *** In special cases, during a long speech pause, it is possible from -3 dB (70%).

Table 2. Subjective evaluation of the audio quality
Evaluation	Influence of defects on perception of the information	Defects of the audio
Very bad (technical stop)	Information completely does not perceived. The main contents of the program does not perceived.	Defect in full or in part does not suppose the information perception.
Poorly (waste)	Under careful listening the information is perceived complicatedly.	Defect interferes with audio perception or breaks art integrity of the program.
Satisfactorily	Information is perceived freely, but presence of defects is conclusive.	Defect does not distract from free perception of the information (for example, insignificant distortions).
Good	Information is perceived freely, definition of defects is difficult.	Defect is not highlighted, is monotonous and low-pressured and does not influence on the information perception.
Fine	Information is perceived completely and without distortions.	Defect is absent.
Notes: The acoustic noise describing real conditions on object of TV transmission (recording) if they do not leave for the quality frameworks mentioned above, do not concern to a category of technical defects. Rejection attributes are the defects which the following are broken: legibility — for speech information programs; legibility and timbre — for speech art programs; legibility, timbre and transparency — for music programs

Table 3. Fundamental parameters for subjective evaluation of the audio signal quality
Evaluated parameter	Notes
Legibility of the text (recognition of the words which were told or song).	At an evaluation of this feature its dependence on the partita and on language of the text it is necessary to take into account. For example, at simultaneous singing with various texts it is possible to expect beforehand, that the legibility of words will be less.
Error of editing (audibility of junctions between separate phonogram portions during playback).	The editing errors are the following: unplanned distortion of tempo, intonation, interpretation in junctions of portion of phonogram.
Volume (audio force subjectively felt on some fragment of phonogram, correlated with expected).	Objective parameters — a measure of sound volume, a level of sound volume.
The timbre (sensation of typicalness of sounding of separate audio sources, their groups).	Timbre is determined by frequency spectrum, transients and level of a source of audio. The timbre may change on the acoustic or electro-acoustic characteristic of transmission, from amplitude-frequency and phase-frequency characteristics, and also from acoustic irregularities of the room.
The dynamic range (subjectively felt ratio of maximal and minimal soundings of sound phenomenon under its playback in a room on the side of the reception, correlated with expected sounding and with level of acoustic noises).	Evaluation of a dynamic range of playback is carried out on the audio channel output. Therefore this subjectively estimated parameter includes the objective parameters also — a dynamic range of the program channel.
Nonlinear distortions (signal waveform changes at passage through nonlinear four-port networks with the result when the spectrum of an output signal differs from spectrum of an input signal by its components).	Distortions which are audible as hoarseness, chatter, “dirty” sound because of nonlinearity of the amplitude characteristic of the channel of audio signal transmission. Objective parameters: factor of harmonics, factor of difference tone.
Detonation (unwanted fluctuation of playback signal frequency, caused by fluctuations of media speed).	Detonation results in unwanted sounding distortion under playback of record from magnetic tape or gramophone record. Depending on media speed oscillation frequency the detonation is perceived orally differently. Slow speed oscillation are perceived as change of its pitch, as “wow” sound. At the increased oscillation frequency of phonogram speed the record vibration is audible; faster oscillations of media speed are perceived orally as nonlinear distortions and noise. Objective parameter is wow-and-flutter.
Tonality deflection (tonality deflection during playback from tonality of audio source).	Change of tonality of an initial sound may take place in the audio channel of magnetic recording-playback under tape speed change, and also in communications channels with a frequency-division multiplex because of frequency shifting. Objective parameters — deflection of media average speed from rating value (drift).
Electric disturbances (periodic, noise or pulse processes arising in channels of record of transmission and playback of sound).	Specific phenomena, accompanying the operation of the electro-acoustic equipment, for example: noise, bang, background, clicks, tone signals, transitive noises, echo effect etc. Objective parameters — relative level of noise, relative level of transitive disturbances.
Acoustic disturbances.	Unwanted audio phenomena on the place of performance or listening caused by operation of ventilation, lighting devices, creak of furniture, bang of doors, movements of executors and public, acoustic defects of the room, specific processes in systems of audio amplification, and also the sounds penetrating from the outside (industrial, transport etc.).
Linear distortions.	Signal distortion in transmission channel (audio), caused by dependence of a complex transmission index from frequency, perceived as timbre change. Objective parameter — variation in frequency response.
Duration of reverberation.	Audible duration of an echo.
Transparency of sounding.	Opportunity of time and qualitative distinction of separate sources in the general audio picture. Objective parameters — a measure of the audio transparency, a measure of the audio legibility, the echo factor, sound structure (including processes of switching-on and switching-off).
Spatial impression.	Perception of joint action of audio sources and their spatial environment which basically is characterized by partial parameters: impression about the room size; duration of reverberation; acoustic balance; acoustic environment; timbre of echo; distribution of directions of spatial sound.
Localization.	Opportunity of recognition of the audio source position by direction and distance to the listener. Special cases are the following: localization of middle, localization in the foreground, localization in a head, shift of the audio image, impression of presence.
The width of a audio picture (subjectively evaluated during listening of the audio picture width).	During stereophonic playback the audio picture width may be more or less than the width of base. Objective parameters — width of base, angle of base, extent of the audio source, and also influencing transitive attenuation and shift of phases.

It is much more difficult to organize a competent evaluation of a multichannel audio track. First of all, it is necessary in TV studio center to create the multichannel sound equipment room (see fig. 1), in which it would be possible to control by level of reproduced audio signals, frequency correction of signals, alignment of the signal level, dynamic processing etc. Also it is necessary to take into account, that all manipulations will be based finally on that audio image which will be created in the equipment room. The area of multichannel control room should be in the range of 30…80 m². Three speakers should be located ahead and two back speakers — under an angle of 110°. Frontal speakers are mounted at height of 1…2 m, back speakers settle down at the same height. Frequency dependence of audio pressure should be uniform in the frequency range of 200 Hz…2kHz with the tolerance equal to ±3 dB. Average reverberation time in the band 200 Hz…4 kHz should be equal to 0.2…0.4 sec. The weighed noise level should not exceed NR 10.

Speaker placement in multichannel audio production studio

The background noise should not be pulse, cyclic or audio-frequency. The multichannel audio control room will allow to make quality check of the multichannel audio track, receiving TV programs and TV films and to make the decision on their translation and post scoring.

Misbalance of TV program volume is the major problem. Thus two following factors are distinguished: the volume misbalance inside one program and the volume misbalance between channels. The decision of this problem is bound with uniformity of measurements, maintenance of regulations of the modulation level, installation of broadcast processors. It is well-known, that the evaluation of the music and speech volume level can be frequently resulted in different results. But if the analog equipment is tolerant to the signals exceeding an allowable level then there are unpleasant distortions in the digital equipment under its overloading. Always it is necessary to mean, that level regulation in TV should be carried out during any process stage: initial record, editing, post scoring and program release. It is necessary to recognize as the most suitable meter of digital audio signals is the peak program meter with time integration equal to 10 ms, analog scale and indicator of overloading. It is normally, that for regular perception of the audio program the ratio of level average value to peak (for example, for sound of the household TV set with decoder NICAM — 12 dB) should be maintained. thus as volume average value the volume level under zero indication of VU-meter (from time constant of 280…300 ms) is accepted, then the peak value accepts the absolute value or the indication of the digital peak level meter with time of integration equal to one sample. Under digital recording the ratio of average value to the peak level, equal to 18 dB (because of the great S/N reserve) is easily achieved. At the same time during recording process the ratio of average value to peak value reduce up to 14 dB normally and maintain it up to the end of the program preparation processing. As a rule before on-the-air the following ratio are used: for stereo NICAM — 12 dB, for DVB-T — 14 dB. For reception of the mono signal compatible with stereophonic, the addition of stereo signal components with 90° shift and level reduction on -3 dB is used. The even multichannel signals of environmental audio (LCRS or 4:2:4) are exposed to the same processing under their conversion to a mono signal. Always it is necessary to mean, that because of tape saturation under analog magnetic audio tape recording the level may achieve the value +6 dB only. Under digital recording the tape saturation is absent. Digital signal distortion arise at 0 dB level at once. At the same time because of the wide dynamic range at digital recording the compression in microphone channels do not use by operators frequently. Under further processing of the recorded audio signals it may result in decrease of the recorded voice level for televiewer perception.

Also it is necessary to take into account, that it is impossible to mix the sound at 0 dB level in jingles and advertising. In this case it is necessary to entrust more to your ears than to the devices.