ITU-T J 16-1988 Measurement of Weighted Noise in Sound - Programme Circuits《声音节目电路中加权噪声的测量》.pdf

1、INTERNATIONAL TELECOMMUNICATION UNION)45G134 * TELECOMMUNICATIONSTANDARDIZATION SECTOROF ITU4%,%6)3)/.G0G0!.$G0G03/5.$G0G042!.3-)33)/.-%!352%-%.4G0G0/ amended at Geneva, 1976 and 1980)The noise objectives for sound-programme circuits are defined in terms of psophometrically weighted noisepower level

2、s at a zero relative level point. Psophometric weighting is used to ensure that the objectives and the resultsof measurements are directly related to the disturbing effect of the noise on the human ear. The psophometric weightingfor sound-programme circuits consists of two operations: a frequency-de

3、pendent weighting of the noise signal, and a weighting of the time function of the noise signal to take account of the disturbing effect of noise peaks.To achieve results which are comparable, it is recommended that for the measurement of noise in sound-programme circuits, a measuring set be used wh

4、ich conforms to the characteristics laid down in CCIRRecommendation 468 which is reproduced at the end of this Recommendation.Annex A gives symbols and definitions used in noise measurements.ANNEX A(to Recommendation J.16)Symbols and definitions used in noise measurementsA clear distinction should b

5、e made between measurements performed with equipment conforming to theRecommendation cited in 1 and those with equipment conforming to CCIR Recommendation 468.It is recommended that the definitions and symbols in Table A-1/J.16 be used.TABLE A-1/J.16Definitions and symbols for the specification of n

6、oise measured on sound-programme circuitsDefinitions SymbolsUnweighted noise level, measured with a quasi-peak measuring instrument complying withCCIR Recommendation 468 and referred to a point of zero relative sound-programme leveldBq0sWeighted noise level, measured with a quasi-peak measuring inst

7、rument complying with CCIRRecommendation 468 and referred to a point of zero relative sound-programme leveldBq0psReference1 CCITT Recommendation Psophometers (apparatus for the objective measurement of circuit noise), GreenBook, Vol. V, Rec. P.53, Part B, ITU, Geneva, 1973.2 Fascicle III.6 - Rec. J.

8、16CCIR RECOMMENDATION 468-4MEASUREMENT OF AUDIO-FREQUENCY NOISE VOLTAGELEVEL IN SOUND BROADCASTING(Question 50/10)(1970 1974 1978 1982 1986)The CCIR,CONSIDERING(a) that it is desirable to standardize the methods of measurement of audio-frequency noise in broadcasting, insound-recording systems and o

9、n sound-programme circuits;(b) that such measurements of noise should provide satisfactory agreement with subjective assessments,UNANIMOUSLY RECOMMENDSthat the noise voltage level be measured in a quasi-peak and weighted manner, using the measurement systemdefined below:1. Weighting networkThe nomin

10、al response curve of the weighting network is given in Fig. 1 b which is the theoretical response ofthe passive network shown in Fig. 1 a. Table I gives the values of this response at various frequencies.The permissible differences between this nominal curve and the response curve of the measuring e

11、quipment,comprising the amplifier and the network, are shown in the last column of Table I and in Fig. 2.Note 1. When a weighting filter conforming to 1 is used to measure audio-frequency noise, the measuring deviceshould be a quasi-peak meter conforming to 2. Indeed, the use of any other meter (e.g

12、 an r.m.s. meter) for such ameasurement would lead to figures for the signal-to-noise ratio that are not directly comparable with those obtained byusing the characteristics that are described in the present Recommendation.Note 2. The whole instrument is calibrated at 1 kHz (see 2.6).(A constant-res

13、istance realization is described in Annex I)(A tolerance of at most 1% on component values and a Q-factor of at least200 at 10 000 Hz are sufficient to meet the tolerances given in Table I.)(The difference between the responses at 1000 Hz and 6300 Hz may beadjusted more precisely by a small adjustme

14、nt of the 33.06 nF capacitor, orby a different approach using an active filter CCIR, 1982-86aFascicle III.6 - Rec. J.16 3 4 Fascicle III.6 - Rec. J.16TABLE IFrequency(Hz)Response(dB)Proposed tolerance(dB)31.5631002004008001 0002 0003 1504 0005 0006 3007 1008 0009 00010 00012 50014 00016 00020 00031

15、500 29.9 23.9 19.8 13.8 7.8 1.90+ 5.6+ 9.0+ 10.5+ 11.7+ 12.2+ 12.0+ 11.4+ 10.1+ 8.10 5.3 11.7 22.2 42.7 2.0 1.4 (1) 1.00.85(1) 0.7 (1)0.55(1) 0.5 0.5 0.5 (1) 0.5 (1) 0.50 0.2 (1) 0.4 (1) 0.6 (1) 0.8 (1) 1.2 (1) 1.4 (1) 1.6 (1) 2.0+ 2.8(1)- 1)This tolerance is obtained by a linear interpolation on a

16、logarithmic graphon the basis of values specified for the frequencies used to define themask, i.e. 31.5, 100, 1000, 5000, 6300 and 20 000 Hz.Fascicle III.6 - Rec. J.16 52. Characteristics of the measuring deviceA quasi-peak value method of measurement shall be used. The required dynamic performance

17、of the measuringset may be realized in a variety of ways (see Note). It is defined in the following sections. Tests of the measuringequipment, except those for 2.4, should be made through the weighting network.Note After full wave rectification of the input signal, a possible arrangement would consi

18、st of two peak rectifiercircuits of different time constants connected in tandem CCIR, 1974-78.2.1 Dynamic characteristic in response to single tone-burstsMethod of measurementSingle bursts of 5 kHz tone are applied to the input at an amplitude such that the steady signal would give areading of 80%

19、of full scale. The burst should start at the zero-crossing of the 5 kHz tone and should consist of anintegral number of full periods. The limits of reading corresponding to each duration of tone burst are given in Table II.The tests should be performed both without adjustment of the attenuators, the

20、 readings being observed directlyfrom the instrument scale, and also with the attenuators adjusted for each burst duration to maintain the reading asnearly constant at 80% of full scale as the attenuator steps will permit.6 Fascicle III.6 - Rec. J.16TABLE IIBurst duration(ms)1(1)2 5 10 20 50 100 200

21、Amplitude referencesteady signalreading (%)(dB)17.0 15.426.6 11.540 8.048 6.452 5.759 4.668 3.380 1.9Limiting values: lower limit(%)(dB) upper limit(%)(dB)13.5 17.421.4 13.422.4 13.031.6 10.034 9.346 6.641 7.755 5.244 7.160 4.450 6.068 3.358 4.778 2.268 3.392 0.7(1)The Administration of the USSR int

22、ends to use burst durations 5 ms.2.2 Dynamic characteristic in response to repetitive tone-burstsMethod of measurementA series of 5 ms bursts of 5 kHz tone starting at zero-crossing is applied to the input at an amplitude such thatthe steady signal would give a reading of 80% of full scale. The limi

23、ts of the reading corresponding to each repetitionfrequency are given in Table III.The tests should be performed without adjustment of the attenuators but the characteristic should be withintolerance on all ranges.TABLE IIINumber of bursts per second 2 10 100Amplitude referencesteady signal reading

24、)(dB)48 6.477 2.397 0.25Limiting values: lower limit (%)(dB) upper limit (%)(dB)43 7.353 5.572 2.982 1.794 0.5100 0.0Fascicle III.6 - Rec. J.16 72.3 Overload characteristicsThe overload capacity of the measuring set should be more than 20 dB with respect to the maximum indicationof the scale at al

25、l settings of the attenuators. The term “overload capacity” refers both to absence of clipping in linearstages and to retention of the law of any logarithmic or similar stage which may be incorporated.Method of measurementIsolated 5 kHz tone-bursts of 0.6 ms duration starting at zero-crossing are ap

26、plied to the input at an amplitudegiving full scale reading using the most sensitive range of the instrument. The amplitude of the tone-bursts is decreasedin steps by a total of 20 dB while the readings are observed to check that they decrease by corresponding steps withinan overall tolerance of 1 d

27、B. The test is repeated for each range.2.4 Reversibility errorThe difference in reading when the polarity of an asymmetrical signal is reversed shall not be greater than 0.5dB.Method of measurement1 ms rectangular d.c. pulses with a pulse repetition rate of 100 pulses per second or less are applied

28、to the inputin the unweighted mode, at an amplitude giving an indication of 80% of full scale. The polarity of the input signal isreversed and the difference in indication is noted.2.5 OverswingThe reading device shall be free from excessive overswing.Method of measurement1 kHz tone is applied to th

29、e input at an amplitude giving a steady reading of 0.775 V or 0 dB (see 2.6). Whenthis signal is suddenly applied there shall be less than 0.3 dB momentary excess reading.2.6 CalibrationThe instrument shall be calibrated such that a steady input signal of 1 kHz sine-wave at 0.775 V r.m.s., havingles

30、s than 1% total harmonic distortion, shall give a reading of 0.775 V, 0 dB. The scale should have a calibrated rangeof at least 20 dB with the indication corresponding to 0.775 V (or 0 dB) between 2 and 10 dB below full scale.2.7 Input impedanceThe instrument should have an input impedance 20 k and

31、if an input termination is provided then thisshould be 600 1%.3. Presentation of resultsNoise voltage levels measured according to this Recommendation are expressed in units of dBqps.Note 1 If, for technical reasons, it is desirable to measure unweighted noise, the method described in Annex II shoul

32、dbe used.Note 2 The influence of the weighting network on readings obtained with different spectra of random noise isdiscussed in Report 496.REFERENCESCCIR Documents1974-78: 10/28 (United Kingdom).1982-86: a. 10/248 (Australia).8 Fascicle III.6 - Rec. J.16BIBLIOGRAPHYBBC 1968 Research Department Rep

33、ort No. EL-17. The assessment of noise in audio-frequency circuits.DEUTSCHE NORMEN DIN 45 405.STEFFEN, E. 1972 Untersuchungen zur Geruschspannungsmessung (Investigations into the measurement of noise voltage).Techn. Mitt. RFZ, Heft 3.WILMS, H. A. O. December, 1970 Subjective or psophometric audio no

34、ise measurement: A review of standards. J. Audio Eng.Soc., Vol. 18, 6.CCIR Documents1978-82: 10/9 (EBU); 10/31 (L M Ericsson); 10/38 (OIRT); 10/225 (German Democratic Republic).ANNEX ICONSTANT RESISTANCE REALIZATION OF WEIGHTING NETWORKBIBLIOGRAPHYAUSTRALIAN BROADCASTING COMMISSION Engineering Devel

35、opment Report No. 106 Constant resistance realization ofCCIR noise weighting network, Recommendation 468.Fascicle III.6 - Rec. J.16 9ANNEX IIUNWEIGHTED MEASUREMENTIt is recognized that unweighted measurements outside the scope of this Recommendation may be required forspecific purposes. A standard r

36、esponse for unweighted measurements is included here for guidance.Frequency responseThe frequency response shall be within the limits given in Fig. 4.This response serves to standardize the measurement and ensure consistent readings of noise distributed acrossthe useful spectrum. When out-of-band signals, e.g. carrier leaks, are present at a sufficient amplitude, they mayproduce readings that are inconsistent between measuring equipments whose responses are different but still fall withinthe tolerance template of Fig. 4.BIBLIOGRAPHYCCIR Documents1978-82: 10/76 (CMTT/14) (Canada).