1、 1 Purpose This guideline provides specifications for recording dialog under conditions normally encountered in motion-picture production. The use of a typical portable analog 1/4-in tape recorder operating at 7.5 in/s is assumed. Variations on typical uses are described in the annex. 2 Scope Many f
2、actors can limit the performance of production audio recordings; among them are acoustical noise and reverberation on the set; placement, directionality, and frequency response of microphones; microphone preamplifier performance; and many others. The object of this guideline is to outline only those
3、 factors which are controlled by the setting of level controls which result in how much flux is recorded on the tape. A background section enumerates some of the problems resulting from under- and over-recording, with the objective being the best possible level-setting practice. A specification sect
4、ion gives recommendations for the levels to be used for the peaks of dialog as measured by vu meters or by quasi-peak-responding meters. The specifications were developed from accumulated experience with varying recording conditions, tape types, etc. The guideline describes two methods for productio
5、n recording of dialog: (a) when a production mixer is fed to a 1/4-in recorder, or (b) when audio from microphones is fed directly into a 1/4-in production recorder. 3 Background 3.1 This guideline was developed with input from the Subcommittee on Audio Production and Post-Production for Motion-Pict
6、ure and Television Entertainment Programming. The subcommittee found that tapes made in production were often over recorded, leading to excessive print-through and distortion. Thus, a guideline outlining good practice was desirable. Several reasons may exist for the tendency to over-record, includin
7、g a low sound pressure level available in headphone monitors which the recordist must use to remain on the set. 3.2 Under-recording causes excessive noise due to tape hiss when the recorded level must subsequently be raised when dubbing to further generations. 3.3 Over-recording may result in: 3.3.1
8、 Audible distortion on peaks of modulation. Such distortion can only become worse in subsequent dubbing because the distortion accumulates from generation to generation. Page 1 of 2 pages EG 15-2005 Revision of EG 15-1987 Copyright 2005 by THE SOCIETY OF MOTION PICTURE AND TELEVISION ENGINEERS 595 W
9、 Hartsdale Ave., White Plains, NY 10607 (914) 761-1100 Approved April 4, 2005 SMPTE ENGINEERING GUIDELINE Level for Analog Recording of Dialog in Motion-Picture Production EG 15-2005 Page 2 of 2 pages 3.3.2 Print-through. “Print“ from layer-to-layer of the stored tape occurs because the most suscep
10、tible magnetic domains in a given layer of oxide can be magnetized by flux from adjacent layers. The result is pre- and post-echoes polluting recordings. Print-through is often the single most limiting factor in the performance of dialog recordings, since dialog recorded on a quiet, non-reverberant
11、set or location consists of syllables of speech interspersed with virtual silence, which affords an opportunity for print-through to become audible. 4 Specifications 4.1 When a production mixer which uses a vu meter is connected to a 1/4-in recorder with a quasi-peak-responding meter, a reference to
12、ne of 1 kHz shall be used to set an indication of “0“ on the vu meter of the mixer, and the line input control of the recorder shall be set to read -10 dB on the quasi-peak-responding meter. 4.2 With this calibration, peaks of dialog should normally be recorded no higher than 0 dB on the vu meter. W
13、ith a 10-dB offset between the vu meter and the quasi-peak-responding meter, the peaks of dialog on the quasi-peak-responding meter should, for most typical dialog recordings, also read no higher than 0 dB. 4.3 When no external mixer is used for recording, peaks of dialog should normally not exceed
14、0 dB on the quasi-peak-responding meter. Annex A (informative) A.1 The majority of production recorders are factory set to record a flux level of 320 nanowebers per meter (nWb/m) with a “0“ reading at 1 kHz on a quasi-peak-responding meter for 7.5 in/s recording. This guideline is for these recorder
15、s; for recorders with other reference levels, the peaks should correspond to 320 nWb/m. A.2 The vu meter is specified in American National Standard Volume Measurements of Electrical Speech and Program Waves, ANSI/IEEE 152. The quasi-peak-responding meter, often called a modulometer, is specified in
16、IEC Publication 268-10A, Programme Level Meters. The faster integration time of the quasi-peak-responding meter, a factor of approximately 30, leads to the higher levels shown by the quasi-peak-responding meter on speech and other program material than on sine-wave tones. A.3 The guideline is writte
17、n for analog recorders without companding noise reduction. If noise reduction is employed, greater freedom from the audibility of print-through and lower noise permits leaving greater headroom for unexpected peaks of dialog, so the recommendation to record peaks of dialog at “0“ is retained, but add
18、itional headroom is available for unexpected peaks. A.4 Magnetic tapes vary by as much as 10 dB in their signal-to-print characteristics. Choice of a “low print“ tape could be as much as 10 dB better than another choice. A.5 Measures which can be used to minimize print-through, besides use of special “low-print“ tape, are to maintain proper storage conditions, especially not exposing the tape to excessively high temperatures, and storing the tape tails out. Reference William A. Manly, “Thinking about Print-Through,“ Audio, 61:54 ff., Sept. 1977.
