SMPTE ST 211M-2001 Motion-Picture Film - 16- and 35-mm Variable-Area Photographic Audio Records - Signal-to-Noise Ratio.pdf

1、SMPTE STANDARD SMPTE 211 M-2001 Revision of ANSI/SMPTE 211M-1996 for Motion-Picture Film - 16- and 35-mm Variable-Area Photographic Audio Records - Signal-to-Noise Ratio 1 Scope This standard specifies a method for measuring the signal-to-noise ratio of 16- and 35-mm variable-area photographic audio

2、 records. 2 Definitions 2.1 biased, unmodulated audio record: A variable- area audio record with no input to the photo- graphic audio recorder, but with noise-reduction biasing used in accordance with normal practice for the recorder being used. 2.2 fully-modulated audio record: A var i a b I e - ar

3、ea audio record which has an amplitude equal to the maximum amplitude permitted by the ap- plicable standard defining the dimensions of the photographic audio records (see annex e). 2.3 system noise: The noise output of the repro- ducer itself. 2.4 unbiased, unmodulated audio record: A variable-area

4、 audio record recorded with no input to the photographic audio recorder, and with no noise-reduction biasing. 2.5 weighting network A circuit which alters the frequency response of the measuring apparatus by a prescribed amount to provide agreement between the measured signal-to-noise ratio and the

5、subjective impression of noisiness. Page 1 of 6 pages 3 Test method 3.1 Test track The test trackshall consist of three sections, recorded in sequence at the same lamp-current setting and printed at the same light step. 3.1.1 Section 1 shall consist of an audio record of 1000 Hz recorded at 80% of f

6、ull modulation and shall serve as the reference signal. The length of this section shall provide about 10 seconds of running time. When reproduced, this section shall have less than 5% harmonic distor- tion. 3.1.2 Section 2 shall be recorded with an unbi- ased, unmodulated audio record. The length o

7、f this section shall provide about 15 seconds of running time. 3.1.3 Section 3 shall be recorded with a biased, unmodulated audio record. The length of this section shall provide about 15 seconds of run- ning time. 3.2 Test measurements 3.2.1 The test track described in 3.1 shall be recorded and dev

8、eloped under standard condi- tions for the system being checked. There shall be sufficient unspliced film ahead of the test track to permit stabilization of printer speed. 3.2.2 The test track shall be reproduced, and the output of the reproducer shall be measured with the required test apparatus (s

9、ee 4.1 through 4.4). The signal level of section 1 (the reference signal) shall be measured without the weighting Copyright Q 2001 by THE SOCIETY OF MOTION PICNRE ANDTELNISION ENGINEERS 595 W. HaMale Ave., white Phi NY 108)7 (914) 761-1 1 Approved June 6,2001 SMPTE 211M-2001 - WEIGHTING NMRK REPRODU

10、CER network, and the signal level of sections 2 and 3 shall be measured with the weighting network. The unbiased, unmodulated signal-to-noise ratio (A) in decibels, shall be calculated as follows: AC VLmEER Vsl vs2 A= 20l0g - + 2dB where Vsi is the signal level of section 1 in volts and VS is the si

11、gnal level of section 2 in volts. The biased, unmodulated signal-to-noise ratio (B), in decibels, shall be calculated as follows: vs1 VS3 B= 20l0g - + 2dB where Vsi is the signal level of section 1 in volts and Va is the signal level of section 3 in volts. 3.2.3 Following the above measurements, the

12、 system noise shall be measured with the required test apparatus with the exciter lamp on, all driving and take-up motors on, and a 0.4 neutral-density filter placed at or near the film plane. The signal-to-system-noise ratio (C), in decibels, shall be calculated as follows: Vsl Vn C= 20l0g - + 2dB

13、where Vsi is the signal level of section 1 in volts and Vn is the signal level of system noise in volts. If the signal-to-system-noise ratio is not at least 1 O dB greater than the unbiased, unmodulated signal-to- noise ratio or the biased, unmodulated signal-to-noise ratio, whichever is greater, th

14、e system signal-to-noise ratio measurement shall be reported with the audio track signal-to-noise ratios. 4 Test equipment 4.1 Measuring devices Two types of measuring devices may be used: Type ITWARM consists of a weighting circuit with unity gain at 2000 Hz and an average response voltmeter. The s

15、ystem is described in 4.2. Type ITU consists of a weighting network with unity gain at 1000 Hz and a quasi-peak response volt- meter. The system is described in 4.3. Type ITU measurements should be made when the system to be measured contains significant amounts of impulse noise. The readings made o

16、n the two measuring systems are generally diff erent and cannot be compared. The type of measurement used shall be stated when giving the result. 4.2 ITU/ARM measuring apparatus An acceptable signal-to-noise measuring apparatus is shown in figure 1. 4.2.1 Weighting network 4.2.1.1 The nominal respon

17、se of the weighting network shall vary with frequency in accordance with the numerical values shown in the second column in table 1. Figure 1 - Measuring apparatus for signal-to-noise ratio Page 2 ot 6 pages SMPTE 211M-2001 Table 1 -Weighting curve ITWARM ITU Frequency response response Tolerance (H

18、z) (dB) (dB) (dB) 31.5 63.0 100.0 200.0 400.0 800.0 1000.0 2000.0 31 50.0 4000.0 5000.0 6300.0 71 00.0 8000.0 9000.0 10 000.0 12 500.0 14 000.0 16 000.0 20 000.0 31 500.0 -35.5 -29.5 -25.4 -1 9.4 -1 3.4 - 7.5 - 5.6 0.0 + 3.4 + 4.9 + 6.1 + 6.6 + 6.4 + 5.8 + 4.5 + 2.5 - 5.6 -1 0.9 -1 7.3 -27.8 -48.3 -

19、29.9 -23.9 -1 9.8 -1 3.8 - 7.8 - 1.9 0.0 + 5.6 + 9.0 +10.5 +11.7 +12.2 +12.0 +11.4 +10.1 + 8.1 0.0 - 5.3 -1 1.7 -22.2 -42.7 I 2.00 f 1.40* 1: 1.00 f 0.85 I 0.70* t 0.55* I 0.50 f 0.50 I 0.50* f 0.50* f 0.50 0.0 f 0.20* f 0.40* I 0.60* I 0.80 f 1.20* f 1.40* f 1.65* t 2.00 + 2.80“ -OD This tolerance

20、is obtained by linear interpolation or a logarithmic graph on the basis of values specifiec for the frequencies used to define the mask; .e. 31.5, 100, 1000, 5000, 6300, and 20 O00 Hz. 4.2.1.2 The permissible differences between the response curve of the measuring network and the nominal response of

21、 the weighting network shall be as shown in the last column of table 1. 4.2.1.3 The weighting network shall be provided with a means of bypassing or defeating itself. 4.2.2 Voltmeter 4.2.2.1 The ITWARM voltmeter shall provide a voltage indication proportional to the average value of the rectified si

22、gnal. It shall have suffi- cient sensitivity so that the noise signals will cause a meter deflection of at least one-third of full scale. 4.2.2.2 The voltmeter shall be free from exces- sive overswing, determined as follows: When a 1000-Hz signal is suddenly applied to the input at an amplitude whic

23、h would give a steady read- ing of approximately two-thirds of full scale, there shall be less than 0.3 dB momentary ex- cess reading. 4.3 ITU measuring apparatus An acceptable signal-to-noise measuring apparatus is shown in figure 1. 4.3.1 Weighting network 4.3.1.1 The nominal response of the ITU w

24、eighting network shall vary with frequency in accordance with the numerical values shown in the third column of table 1. J: 4.3.1.2 The permissible differences between the response curve of the measuring networks and the nominal response of the weighting network shall be as shown in the last column

25、of table 1. 4.3.1.3 The weighting network shall be provided with a means of bypassing or defeating itself. 4.3.2 Voltmeter The ITU voltmeter shall provide a voltage indication proportional to the quasi-peak value of the signal, as follows: 4.3.2.1 Meter response The response of the meter to signal t

26、one bursts shall be as shown in table 2. The method of measurement shall be as follows: Single bursts of 5-m tone are applied to the input of an amplitude such that the steady signal would give a reading of 80% of full scale. The limits of reading corresponding to each duration of tone burst are giv

27、en in table 2. The tests shall be performed both without adjustment of the attenuators with the readings being observed directly from the instrument scale, and also with the attenuators adjusted for each burst duration to main- tain the reading as nearly constant at 80% of full scale as the attenuat

28、or steps permit. Page 3 of 6 pages SMPTE 211M-2001 Table 2 - Single tone-burst response (ms)* 1 2 5 10 20 50 1 O0 200 Burst duration Amplitude reference steady signal reading (%) 17.0 26.6 40 48 52 59 68 ao (dB) -15.4 -11.5 -8.0 -6.4 -5.7 -4.6 -3.3 -1.9 Limiting values Lower (%) 13.5 22.4 34 41 44 5

29、0 58 68 Limit (dB) -17.4 -13.0 -9.3 -7.7 -7.1 -6.0 -4.7 -3.3 Upper (%) 21.4 31.6 46 55 60 68 78 92 Limit (dB) -13.4 -10.0 -6.6 -5.2 -4.4 -3.3 -2.2 -0.7 *The rise-and-fall time of the burst envelope shall be less than 5 ps. 4.3.2.2 Response to repetitive tone bursts The meter shall respond to repetit

30、ive tone bursts as shown in table 3. The method of measurement is as follows: A series of 5-ms bursts of a 5-ld-I tone shall be applied to the input at an amplitude such that the steady signal would give a reading of 80% of full scale. The limits of the reading corresponding to each repetition frequ

31、ency are given in table 3. The tests shall be performed without adjustment of the attenu- ators but the characteristic shall be within tolerance on all ranges. 4.3.2.3 Overload characteristics The overload capacity of the measuring set should be more than 20 dB with respect to the maximum indica- ti

32、on of the scale at all settings of the attenuators. The term overload capmty refers both to the absence of clipping in linear stages and to retention of the law of any logarithmic or similar stage which may be incor- porated. Overload capacity shall be measured as follows: Isolated 5-kHz tone bursts

33、 of 0.5-ms duration are applied to the input at an ampiude giving full- scale reading using the most sensitive range of the instrument. The amplitude of the tone bursts is decreased in steps by a total of 20 dB while the readings are observed to check that they decrease by corresponding steps within

34、 an overall tolerance of f 1 dB. The test is repeated for each range. 4.3.2.4 Reversibility error The dflerence in reading when the polarity of an asymmetric signal is reversed shall not be greater Table 3 - Repetitive tone-burst response I Burst repetition frequency iHz) 2 10 1 O0 Amplitude referen

35、ce steady signal reading 48 77 97 (dB) -6.4 -2.3 -0.25 Limiting values Lower limit (%I 43 72 94 (dB) -7.3 -2.9 -0.5 Upper limit (%I 53 a2 1 O0 (dB) -5.5 -1.7 -0.0 Page 4 of 6 pages , SMPTE211M-2001 than 0.5 dB, measured as follows: Isolated 1-ms rectangular pulses are applied to the input in the unw

36、eighting mode, at an amplitude giving an indication of 80% of full scale. The polarity of the input signal is reversed and the difference in indication is noted. 4.3.2.5 Overswing The reading device shall be free from excessive over- swing, measured as follows: When a 1-kHz tone is suddenly applied

37、to the input at an amplitude which would give a steady reading of 0.775 V or O dB, there shall be less than 0.3 dB momentary excess reading. 4.3.2.6 Calibration The instrument shall be calibrated so that a steady input signal of 1 -kHz sine wave at 0.775 V rms, having less than 1% total harmonic dis

38、tortion, shall give a reading of 0.775 V or O dB. The scale should have a Annex A (informative) Additional data Al The effective reference signal level used in this stand- ard is a fully-modulated audio record. However, in order to avoid the production of unwanted harmonic distortion and the possibi

39、lity of uncorrectable overmodulation, 3.1.1 requires that the reference signal be recorded at 80% of full modulation. To bring the effective reference signal level to that of a fully-modulated audio track, 2 dB are then added to the measured signal-to-noise ratio in each of the equa- tions in 3.2.2

40、and 3.2.3. If the reference signal is recorded at other than 80% of full modulation, a correction factor, computed as follows, should be added to the measured signal-to-noise ratio: 0.8 x Wa Wr c = 2010g where C is the correction factor in decibels, Wa is. the width of the fully-modulated audio reco

41、rd, and Wr is the total modulation width of the reference signal. The total modulation width of the reference signal is the sum of the modulation amplitudes, as illustrated in figure A.l, for a bilateral variable-area audio track. The modulation width for a dual-bilateral track is calculated similar

42、ly. A.2 It may be desirable sometimes to make a measurement of the system signal-to-noise ratio of a projector or other reproducer without making it in conjunction with a measure- ment of the signal-to-noise ratio of a particular audio record. In such a case, an appropriate reference signal would be

43、 that contained on the applicable signal level test film. If the reference sianal is not 80% modulated. a correction factor should be applied as described in A.l. A3 The basic measuring method described in this standard is also applicable to 8-mm type S photographic audio tracks. calibrated range of

44、 at least 20 dB with the indication corresponding to 0.775 V (or O dB) between 2 dB and 1 O dB below full scale. 4.4 Test reproducer The area of the film scanned by the test reproducer shall be as described in the applicable standard defining the dimensions of the photographic audio records. The tes

45、t reproducer shall be capable of reproducing all frequencies of a multifrequency test film, as described by the applicable standard, at a uniform level f 2 dB. If the test reproducer does not meet this criterion, the frequency response of the test reproducer shall be reported along with the signal-t

46、o- noise ratios. The meter used for measuring the frequency response of the reproducer shall have no weighting network and shall have either an average response or a true rms meter response. However, no standards now exist describing an 8-mm type S multifrequency or signal level test film; therefore

47、, it is not possible to measure the frequency response as required in 4.4. TOTAL MODULATION WIDTH = MI + M2 Figure A.l - Modulation width Page 5 of 6 pages SMPTE 211M-2001 A.4 The weighting curve for the ITUIARM meter, given in column 2 of table 1, is derived from the weighting charac- teristic spec

48、ified in ITU-R BS.468. It has been modified to have unity gain at 2000 Hz when 5.6 dB are subtracted from the response at each frequency specified in ITU-R 85.468. A5 The ITU measuring apparatus described in 4.3 is in agreement with ITU-R BS.468. The ITU measuring device described in 4.3 is judged t

49、o be the best method for meas- uring signal-to-noise ratios of photographic audio tracks under all conditions, especially when the noise contains impulsive components. A.6 The ITUIARM measuring device described in 4.2 is judged to be effective and useful when the noise signal is uniform and does not contain impulsive components. The ITUIARM method has the advantage of being implemented with more readily available equipment. O A7 One possible network for the realization of the weight- ing characteristic of table 1 is shown in figure A.2. This network is derived from ITU-R BS.468