1、Rec. 644-1 1RECOMMENDATION 644-1*AUDIO QUALITY PARAMETERS FOR THE PERFORMANCE OF AHIGH-QUALITY SOUND-PROGRAMME TRANSMISSION CHAIN(1986-1990)Rec. 644-1The CCIR,CONSIDERING(a) that the performance characteristics of international sound-programme circuits used for programme exchangeare based on a hypot
2、hetical reference circuit (HRC) defined in Recommendation 502;(b) that the performance of a realistic international sound-programme circuit shorter or longer than the HRC canbe derived by calculation according to rules given in Recommendation 605;(c) that the performance of the sound-programme trans
3、mission chain from the broadcasting house to the output ofthe receiver should therefore be based on a reference chain;(d) that sound-programme signals intended for international exchange are conveyed from the broadcasting houseto the international sound-programme centre (ISPC) on national circuits w
4、hich are either the same as, or similar to, thenational circuits which form part of the sound-programme transmission chain;(e) that the performance of the sound-programme transmission chain should be described by a number ofparameters;(f) that for reasons of comparison, clearly defined parameters me
5、asured using test signals, should be used;(g) that these parameters and their target values should be based on the perceptibility of the impairments by thehuman ear,UNANIMOUSLY RECOMMENDS1. that the performance characteristics of a high-quality sound-programme transmission chain should be based onth
6、e reference chain described in Annex I;2. that the performance of the reference transmission chain from the output of the broadcasting house to theoutput of the receiver should be described by the parameters listed in Annex II;3. that the limits given in Annex III be used as realistic target values
7、for these parameters;4. that for the purposes of planning the entire broadcasting chain from the microphone output to the receiveroutput, the subjective threshold values for quality parameters given in Annex IV should be taken for guidance.ANNEX IREFERENCE SOUND-PROGRAMME TRANSMISSION CHAINThe refer
8、ence sound-programme transmission chain is represented, for the purpose of this Recommendation,by Fig. 1 in which A, B, and C are 0 dBrs audio signal interface points, permitting performance comparison:A : interface between the programme source (broadcasting house) and the sound-programme circuitsys
9、tem;A-B : sound-programme circuit system comprising a single cable, radio, communication satellite oroptical-fibre section;B : interface between the sound-programme circuit and emission-reception systems;B-C : emission-reception system, comprising a single broadcasting emitter or a single cable netw
10、ork and ahigh-quality monitoring receiver under optimum conditions;C : audio signal interface at the output of the receiver._*This Recommendation, derived from Report 293 which has been cancelled, should be brought to the attention of the CMTT.2 Rec. 644-1D01-scFIGURE 1 D01 = 3 cmThe EBU CCIR, 1986-
11、90 has published Recommendation R.50 to define the polarity of acoustic signals,electrical audio signals, audio signals at connectors, magnetic signals on audio and video tapes, digital audio signals andacoustic pressure signals for loudspeakers.REFERENCESCCIR Documents1986-90: 10/313 (EBU).ANNEX II
12、LIST OF AUDIO QUALITY PARAMETERSThe audio quality parameters listed below are those which are considered to be the most important in theanalogue environment. Some parameters are also appropriate for connections incorporating digital systems, but somefurther parameters need to be specified for testin
13、g these digital systems.*All test signals should be at the measurement level (see Recommendation 645) unless otherwise stated. Forstereo, measurements should be made in both A and B channels.1. Nominal bandwidthThe effective transmitted frequency band (for both A and B channels, in the case of stere
14、o).2. Amplitude/frequency responseThe response should be expressed relative to the level at the reference frequency of 1 kHz.3. Group-delay variationThe results of measurements throughout the nominal bandwidth should be expressed relative to the minimumgroup delay.4. Non-linear distortion*Total harm
15、onic distortion (THD) may be measured using a single-tone test signal at +9 dBu0s, at frequenciesin the range from 40 Hz to 1 kHz, and the result expressed as a “separation” value (i.e. the difference in level betweenthe test signal and the harmonics, expressed in dB)._*Administrations are requested
16、 to study quality parameters and testing methods for digital systems, and contribute to the work ofthe CCIR under Study Programme 51B/10. Report 1070 is alo relevant.*Administrations are requested to consider these and other suitable distortion measurement methods, and contribute theirconclusions to
17、 the CCIR under Study Programme 50E/10.Rec. 644-1 3Distortion may, alternatively, be measured using the total difference frequency (TDFD) method Thiele, 1975.This is an intermodulation test in which the test signal comprises two tones at frequencies chosen so that second-orderand third-order intermo
18、dulation products (i.e. one of each) occur close in frequency, and may be measured together by ameter with a selective filter. Tests in the UK CCIR, 1986-90a have shown that the TDFD method is less sensitive tonoise than THD, and is not restricted to the use of test signals at frequencies below 5 kH
19、z. This could be important whentesting pre-emphasised connections, although care must be taken to avoid the possibility of overloading resulting fromthe pre-emphasis.The TDFD method is described in Annex V.The duration for which the test signals are transmitted should be restricted in accordance wit
20、h the appropriateCCITT Recommendations of the N series.5. Error in reconstituted frequencyThe error in reconstituted frequency should not exceed the values in Annex III, for individual channels.For stereo, any differential frequency shift between A and B channels is unacceptable.A sensitive measurin
21、g method is described in Thiele and Bonner, 1985.6. Error in amplitude/amplitude responseA 1 kHz tone is applied to the input of the transmission chain at +9 dBu0s and at 31 dBu0s alternatively. Theoutput levels should be measured selectively, and the difference should be within the limits specified
22、 in Annex III.The duration for which the test signal is transmitted should be restricted in accordance with the appropriateCCITT Recommendations of N series.7. Level stabilityThe difference in the level of a constant reference signal, at the measurement level, applied to the input of thetransmission
23、 chain, from its nominal level should be determined for a 24 h period. This difference may vary over the24 h period, and include a fixed adjustment error.For stereo, the level stability should be determined concurrently for both A and B channels.8. Noise (and single-tone interference)Weighted noise
24、voltage level (and, where appropriate, single-tone interference voltage level) is measured inaccordance with Recommendation 468. The permitted maximum signal-to-(weighted) noise ratio (PMS/N, expressedin dB) is determined as the difference between the permitted maximum signal (PMS) and the weighted
25、noise levels. Idle-channel noise. Programme-modulated noise, measured with a 60 Hz sinusoidal test signal applied to the input of the transmissionchain at +9 dBu0s and 31 dBu0s. The fundamental and low-order harmonics are suppressed at the input to themeasuring instrument by a high-pass filter with
26、a cut-off frequency between 200 Hz and 400 Hz. The insertion lossof the filter should be at least 56 dB at 60 Hz, taking into account the attenuation of the weighting filter (24 dB atthis frequency).The duration for which the test signal is transmitted should be restricted in accordance with the CCI
27、TTRecommendations of the N series. If necessary, steady single-tone interference should be measured selectively at the frequencies at which tones aredetected or may be expected. The measurement should be made using a bandpass filter (having effectively 0 dBinsertion loss in the passband) in conjunct
28、ion with the noise measuring apparatus described in Recommenda-tion 468, or using a spectrum analyzer and correcting the measured level by the corresponding weighting factor.9. Disturbing modulation by power supplyThis is the ratio of a 1 kHz sinusoidal test signal to the highest level unwanted side
29、-component resulting frommodulation of that test signal caused by interference from conventional a.c. (50/60 Hz) line power supply sources.4 Rec. 644-110. Stereo: level difference between A and B channelsThe difference in level between the A and B outputs of the transmission chain should be measured
30、 with thesame sinusoidal test signal applied to both channel inputs simultaneously.11. Stereo: phase difference between A and B channelsThe difference in phase between the A and B outputs of the transmission chain should be measured with thesame sinusoidal test signal applied to both channel inputs
31、simultaneously.Conservation of the polarity CCIR, 1986-90b of audio signals is recommended.12. Stereo: crosstalk between A and B channels Linear cross-talkA sinusoidal test signal is applied to each channel in turn, and the level of the signal in the other channelis measured selectively. The cross-t
32、alk attenuation (expressed in dB) is the difference between the levels of thesignal in the two channels. Non-linear cross-talkThe conventional test signal for simulating sound-programme signals specified in Recommendation 571is applied to each channel in turn. If the other channel is influenced by n
33、on-linear cross-talk, an increase in theweighted noise level may be observed.The increased noise voltage Nsformed by the sum of two noise contributions: the idle-channel noise N0and the non-linear cross-talk noise NcT. The latter voltage can be calculated using the following formula:NcT= (Ns)2 (N0)2
34、The non-linear cross-talk is expressed as the ratio of the permitted maximum signal to this level of non-linear cross-talk noise (i.e. PMS/NcT).REFERENCESTHIELE, A. N. September, 1975 Measurement of non-linear distortion in a band-limited system. IREECON InternationalConvention Digest, 480-482, Sydn
35、ey, Australia.THIELE, A. N. and BONNER, D. J. September-October, 1985 Measuring reconstituted frequency error in f.d.m. sound-programmecircuits. IREECON International Convention Digest, 684-686, Melbourne, Australia.CCIR Documents1982-86: a. 10/213 (Sweden); b. 10/224 (France).1986-90: a. 10/236 (Un
36、ited Kingdom); b. 10/313 (EBU).Rec. 644-1 5ANNEX IIITABLE I Realistic target values for the referencesound-programme transmission chainParameter(for details seeTest signalSound-programmecircuit systemEmission-reception Whole referenceAnnex II)Frequency Level(A-B) (1) system (B-C) (1) chain (A-C) (1)
37、Nominal bandwidth(see 1) 40 Hz-15 kHz 40 Hz-15 kHz 40 Hz-15 kHzAmplitude/frequency response(see 2)40 Hz -125 Hz125 Hz - 10 kHz10 kHz- 14 kHz14 kHz- 15 kHz12 dBu0s+0.2/ 1.0 dB+0.2/ 0.2 dB+0.2/ 1.0 dB+0.2/ 1.4 dB+0.4/ 1.5 dB+0.4/ 0.4 dB+0.4/ 1.5 dB+0.4/ 2.3 dB+0.5/ 2.0 dB+0.5/ 0.5 dB+0.5/ 2.0 dB+0.5/
38、3.0 dBGroup-delay/variation(see 3)40 Hz15 kHz12 dBu0s18 ms4 ms37 ms8 ms55 ms12 msTotal harmonic distortion(see 4)40 Hz -125 Hz125 Hz - 1 kHz+9 dBu0s46 dB52 dB42 dB48 dB40 dB46 dBError in reconstitutedfrequency (see 5)Any 12 dBu0s 0.2 Hz 0 0.2 HzError in amplitude/amplituderesponse(see 6)1 kHz+9 dBu0
39、s31 dBu0s0.4 dB 0.8 dB 1.0 dBLevel stability (over 24-hperiod)(see 7)1 kHz 12 dBu0s 0.6 dB 0.8 dB 1.0 dBNoise and single-toneinterference: idle-channel noise programme-modulatednoise programme-modulatednoise single-tone interference(see 8)60 Hz60 Hz31 dBu0s+9 dBu0s53 dB53 dB43 dB73 dB56 dB56 dB46 dB
40、76 dB51 dB51 dB41 dB71 dBDisturbing modulation bypower supply(see 9)1 kHz 0 dBu0s 51 dB 47 dB 45 dBLevel difference between Aand B channels(see 10)40 Hz -125 Hz125 Hz - 10 kHz10 kHz- 14 kHz14 kHz- 15 kHz12 dBu0s1.0 dB0.5 dB1.0 dB2.0 dB1.0 dB0.5 dB0.5 dB1.0 dB1.5 dB1.0 dB1.5 dB2.5 dBPhase difference
41、between Aand B channels(see 11)40 Hz40 Hz -125 Hz125 Hz - 10 kHz10 kHz- 15 kHz15 kHz12 dBu0s15(2)10(2)1510(2)8(2)1020(2)15(2)25Cross-talk between Aand B channels: linear cross-talk 40 Hz40 Hz -300 Hz300 Hz - 4 kHz4 kHz- 15 kHz15 kHz0 dBu0s46 dB(2)56 dB(2)46 dB22 dB(2)(2)26 dB21 dB(2)35 dB(2)25 dB no
42、n-linear cross-talk(see 12)Sound programmesimulating signal(see Recommendation 571)60 dB (3)(3)(1) The values given for sections A-B and B-C, and for the whole transmission chain (A-C) are desirable, and are to be taken intoaccount when designing new national broadcasting networks.(2) Values within
43、this range are obtained by linear interpolation between the values for adjacent ranges on a graph with a logarithmicfrequency scale.(3) Administrations are requested to suggest target values.6 Rec. 644-1ANNEX IVSubjective threshold values are thresholds of perception, found by subjective statistical
44、 research under ideallistening conditions. They have been ascertained from test results obtained by CCIR members and from internationalliterature.TABLE II Subjective threshold valuesParameterFrequency oftest signalSubjectivethreshold valueAmplitude/frequency response(see 2 of Annex II)40 Hz -125 Hz1
45、25 Hz - 10 kHz10 kHz- 14 kHz14 kHz- 15 kHz 1.0 dB 0.5 dB 1.0 dB 2.0 dBGroup delay variation (see 3 of Annex II) 40 Hz-15 kHz (see Note 1)Non-linear distortion THD (see 4 of Annex II) 40 Hz-1 kHz 52 dB(see Note 2)Error in reconstituted frequency (see 5 of Annex II) Any 0.25 HzError in amplitude/ampli
46、tude response (see 6 of Annex II) 1 kHz (see Note 1)Level stability (over a 24-h period)(see 7 of Annex II)1 dB(see Note 3)Noise and single-tone interference: idle-channel conditions test signal level: +9 dBu0s test signal level: 31 dBu0s single-tone interference(see 8 of Annex II)60 Hz60 Hz70 dB(se
47、e Note 1)(see Note 1)80 dBDisturbing modulation by power supply(see 9 of Annex II)(see Note 1)Level difference between A and B channels(see 10 of Annex II)40 Hz -125 Hz125 Hz - 10 kHz10 kHz- 14 kHz14 kHz- 15 kHz2.0 dB0.5 dB1.5 dB2.0 dBPhase difference between A and B channels(see Note 4) (see also 1
48、1 of Annex II)40 Hz40 Hz -125 Hz125 Hz - 10 kHz10 kHz- 15 kHz15 kHz45(see Note 5)30(see Note 5)90Cross-talk between A and B channels: linear cross-talk non-linear cross-talk(see 12 of Annex II)40 Hz40 Hz -300 Hz300 Hz - 4 kHz4 kHz- 15 kHz15 kHzSound-programmesimulating signal(Recommendation 571)15 d
49、B(see Note 5)20 dB(see Note 5)15 dB(see Note 1)Note 1. Administrations are invited to contribute figures for this value.Note 2. This figure assumes that the distortion is predominantly second and third harmonic.Note 3. A change of 1 dB is perceptible only if it is a sudden change.Note 4. The tolerances given for phase differences between A and B channels are for stereo listening, and they provokeunacceptab
