JPS5914959B2 - Method for measuring abnormal noise in multi-channel record playback system and record for the measurement - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for measuring abnormal sounds in a multi-channel record playback system and a record for measuring the same. The purpose of the present invention is to provide a method for measuring abnormal sounds that can perform continuous numerical management of noise (noise), and a record for the measurement.

In general, each multi-channel audio signal is a direct wave signal in the audible frequency band and an angle-modulated wave signal in a predetermined frequency band, and a multi-channel record is made by multiplexing these two signals and cutting and recording them into one sound groove. It is known that during playback, abnormal sounds may occur due to various reasons. This abnormal sound suddenly occurs when a certain limit is exceeded; it is discontinuous there; there is no particular abnormality even near the limit, and the abnormal sound does not become particularly loud even if the limit is significantly exceeded. . Therefore, conventionally, continuous numerical management of abnormal sounds has not been possible. On the other hand, when producing multi-channel record playback equipment such as pickup cartridges and demodulators, the extent to which these equipment generates abnormal sounds (that is, the quality with respect to abnormal sounds) must be tested by some method. Therefore, the present applicant previously proposed a method for evaluating and measuring a pickup cartridge for measuring abnormal noise and a record for the measurement in Japanese Patent Application No. 56475/1982.

This proposed method covers the entire audio frequency band (e.g. 20Hz~
A multiplexed signal consisting of a large white noise, particularly at a high frequency level (15 kHz), and an unmodulated carrier of, for example, 30 kHz is reproduced, and the magnitude of the demodulated signal of the unmodulated carrier is measured. Thereby, it is possible to measure abnormal sounds generated when harmonics of white noise in the high range of the audible frequency band interfere with carriers. However, the causes of abnormal noise include not only the harmonics of the high-frequency signal in the audible frequency band that jump into the angle-modulated wave signal band and cause interference, but also carrier breakage due to pick-up needle skipping, record sound grooves, etc. A decrease in the carrier level due to wear, a large amount of noise generated from roughening of the groove due to wear, and when the modulation signal has a large amplitude in the midrange range, the angle-modulated wave signal becomes too large than the predetermined frequency deviation, leaving the detection band. There are various other cases such as overmodulation that protrudes.

Therefore, although the proposed method is effective in measuring abnormal noise caused by the harmonic interference, there is a problem in that it is not sufficiently effective in measuring abnormal noise caused by other causes. Ta.

For this reason, until now, the only comprehensive test for abnormal sounds has been to use a multichannel record on which certain music has been recorded, and to judge the quality of the reproduced sound by hearing, which is extremely subjective and intuitive. This was not universal at all, and furthermore, it was difficult to record as data.

Moreover, when playing other music records that were considered to be very similar, the judgment often changed significantly. The present invention solves the above-mentioned problems, and one embodiment thereof will be described below with reference to the drawings.

FIG. 1 shows the frequency spectrum of one embodiment of a signal used in the method of the invention. By the way, it is impossible to identify the cause of the abnormal sound by auditory sensation because all abnormal sounds are perceived as the same noise regardless of the cause. In other words, it is only when all abnormal sounds from all causes are generated at once and measured that a measurement method that closely matches the sense of hearing (for all types of music) can be obtained. In addition, as a result of investigating the reason why the judgment changes drastically when the music is changed, and why a passing product becomes rejected, or conversely, a rejected product becomes passed, we found that music has a special harmonic relationship,
It has become clear that there is a phase relationship and this is the cause.

Therefore, it can be seen that in order to eliminate the above-mentioned phenomenon that the judgment changes depending on the music, it is sufficient to use independent signals that do not have a specific phase relationship for measurement. This makes it possible to measure any kind of music. To summarize the above, the following conditions are necessary to enable general measurements.

1. Contains high-frequency signals that tend to produce harmonics; 2. Contains mid-range signals with large modulation waves (frequency deviations); 3. Must have a level distribution representative of music in general; 4. Audible frequencies above the midrange. Contains the entire spectrum of the band, 5 has no harmonic relationship, 6 has no specific phase relationship, 7 each channel is an independent signal from each other, 8 is capable of continuous numerical management. .

On the other hand, it was found that the frequency spectrum of the abnormal sound that occurred was almost evenly distributed over the entire band.

Among them, the abnormal sounds that had a particularly large impact on the auditory sense were around 100 Hz to 1 kHz, with the frequency being around several hundred Hz. That is, the present invention focuses on the following points. 1 Mid-range frequency that causes abnormal noise (e.g. approximately 5
00Hz to 2kHz) or higher.

2 Medium and low (middle and low) frequency bands are used for measurements.

3. The middle and low frequencies (approximately 200 to 600 Hz) are almost unrelated to the cause of abnormal sound generation.

4. The middle and low frequencies match well with the sense of hearing.

5. Since there is no input signal in the middle and low frequency bands,
Distortion occurring in this band can be obtained with a simple filter.

An example of a signal that satisfies all of the above conditions is the signal shown in FIG.

That is, this signal has a frequency band of about 2kHz to about 1kHz.
Up to 5kHz, for example 2 to make a signal equivalent to music.
It is random noise (for example, white noise) that is attenuated at 6 dB/0 ct above kHz. Here, in addition to the above, as a weighting to make the signal equivalent to music, for example, to make pick noise, pick noise of 5 kHz or more is weighted by 6 d
Attenuate by B/0ct or 2kfIz to 3d
Methods such as attenuation at around B/0ct can be considered, but
It is best to match the recording condition of the actual record.
To explain the case of measuring abnormal sounds in a 4-channel record playback system using the above signal, for example, FIG. 2 is a block system diagram of one embodiment of the record recording system for measurement of the present invention, and FIG. 3 is a block diagram of the present invention. 1 shows a block system diagram of one embodiment of a reproduction system for an invention measurement record.

In FIG. 2, 1 is a random noise oscillator from which pick noise, for example, is output. This pick noise is attenuated by 6 dB/0ct over at least about 2 kHz in an integrating circuit consisting of a resistor R and a capacitor C, and then supplied to a band filter 2, where it is attenuated by about 2 kHz.
Components from Hz to 15kHz are transmitted. As a result, the pick noise that has passed through the band filter 2 has a frequency spectrum as shown in FIG. 1, and is supplied to the multi-channel record modulation circuit 3, where it is converted into a normal multi-channel record modulation signal. Ru. Here, for example, in the case of 4 channels, three more signals shown in FIG. 1 are used. However, these four signals supplied from the bandpass filter 2 to the modulation circuit 3 are independent signals having no specific phase relationship with each other. In the modulation circuit 3, the first and second signals of the four signals (pick noise) are computed into, for example, a sum signal and a difference signal, and then only the difference signal is angularly modulated.

After that, a direct wave signal in the audible frequency band (here about 2
~15kHz sum signal) and an angle-modulated wave signal in a higher frequency band (here, the above difference signal is, for example, 30kHz).
After the signals are multiplexed (a signal obtained by angle-modulating a kHz carrier wave), they are cut and recorded by a cutting needle 5 on one of the left and right walls of the sound groove of the record 6 through a cutting amplifier 4 as a signal for the left channel, for example. . In addition, the above 4
The third and fourth signals among the signals are multiplexed in the same manner as above, and then cut through the cutting amplifier 4 to the other wall of the left and right walls of the sound groove of the record 6 as signals for the right channel. The needle 5 cuts and records. Here, the above-mentioned multiplexed signal is recorded with the frequency above 2kHz raised by 6dB/0ct according to the RIAA recording characteristics, but as shown in Figure 1, the direct wave signal component to be recorded is boosted by 6dB/0ct above 2kHz. When attenuated by /0ct, the velocity amplitude of the direct wave signal is constant over its entire band.

Therefore, the abnormal sound in the reproduced signal that is generated in relation to the velocity amplitude of the direct wave signal is constant insofar as it is caused by the velocity amplitude. In FIG. 3, the reproduced multiplexed signal obtained by picking up and reproducing the record 6 in the pickup cartridge 7 to be measured is as follows:
The signal is supplied to a demodulation circuit 8, which includes a demodulator for demodulating the angle-modulated wave signal, and an arithmetic circuit to which the output of the demodulator and the direct wave signal are supplied, and is demodulated there.

The abnormal sound level in the reproduced signal varies depending on the multi-channel record reproduction system comprising the pick-up cartridge 7 and the demodulation circuit 8. However, as described above, the distribution of occurrence of this abnormal sound is approximately constant over the entire frequency band. The signal of the channel to be measured among the four channel signals taken out from the demodulation circuit 8 is supplied to a bandpass filter 9 having a passing frequency band width of approximately 100Hz to 1kHz as shown in FIG. While the signal component indicated by I in the figure is removed, only the abnormal sound occurring in the band is subjected to F waves.

The design of this bandpass filter 9 is relatively easy since there is no signal below about 2 kHz as mentioned above. Moreover, since the generation distribution of abnormal sounds is substantially constant, even if the pass frequency band width is narrowed to about 300 to 500 Hz, the indicated value of the measuring device 10 (described later) will only become lower, and the relative value will not change. In this case, designing the bandpass filter 9 becomes extremely easy. Furthermore, as shown in FIG. 4, it is also possible to use only a low-pass filter with a cutoff frequency of 1 kHz or 500 Hz instead of the band filter 9. However, in this case, the number of low-frequency components increases, and hum and the like may also be mixed in, making reading the reading of the measurement device 10 somewhat unstable. The abnormal sound component in the band extracted by the band filter 9 is supplied to a measuring device 10 such as an AC voltmeter or an oscilloscope, and its level is measured there. Incidentally, it is known that the level of random noise recorded and reproduced follows a Gaussian distribution, as shown by the curve in FIG. Therefore, abnormal noise is generated in the shaded area where the limit line of the input level of the pickup cartridge exceeds the limit line indicated by v in FIG. 5, and the amount of abnormal noise is proportional to the area of the shaded area. In the present invention, the level is set so that the normal limit line v is close to the standard deviation of the Gaussian distribution. This makes it possible to carry out continuous numerical measurements within a range of approximately 10 dB above and below the center of the limit line distribution of the pickup cartridge to be measured. This range is practically sufficient. In the above embodiment, in principle, four signals as shown in FIG. 1 are required, but when measuring one channel at a time, two signals may be sufficient, and even one signal can provide some effect. .

Further, two oscillators 1 are required as signal sources when recording one channel, and four oscillators 1 are required when recording both channels simultaneously. However, by recording each channel on a table recorder and reproducing them simultaneously, only one oscillator 1 is required. Furthermore, since the properties of random noise do not change whether the sum or difference of two signals is calculated, during modulation or demodulation, the input and output sides of the arithmetic circuits included in the recording circuit 3 and the demodulation circuit 8, respectively. Since the signal is the same no matter which signal is used, it is very convenient to handle.

Therefore, although the signal measured by the measuring device 10 was explained as a demodulated channel signal in the above embodiment, it is not limited to this, and may be a difference signal (of a calculation circuit) that is demodulated from a reproduced angle-modulated wave difference signal. one of the input signals). Further, it is convenient to provide a reference signal portion modulated at several hundred Hz (for example, 400 Hz) in the record 6 and take a comparison value based on this portion.

Furthermore, in the above embodiment, the frequency band of the recorded and reproduced signals was about 2 kHz to 15 kHz, but it can be arbitrarily selected from the middle range to the treble range (about 500 Hz to 20 kHz). Furthermore, although the direct wave signal has been explained as a sum signal of two channels, and the angle-modulated wave signal is a signal in which a carrier wave is modulated with a difference signal of two channels, the present invention is not limited to this, and various other variations can be considered. It is something that can be done. As described above, the method for measuring abnormal sound in a multi-channel record playback system according to the present invention and the record for measurement thereof include one or two or more mutually independent components whose components are distributed over substantially the entire frequency band from the midrange to the treble range. Random noise or a signal equivalent to this is made into a direct wave signal and an angle-modulated wave signal, and then these two signals are multiplexed, and this multiplexed signal is applied to the multi-channel record playback system that is to be measured on the cut record. The random noise or equivalent noise in the input demodulated signal or the output signal of the arithmetic circuit which is supplied with the demodulated signal of the angle-modulated wave signal and the reproduced direct wave signal and outputs the multi-channel signal. By measuring components that occur in the middle and low frequency bands where signals do not normally exist, abnormal sounds caused by all causes are generated at once, making it possible to perform a comprehensive inspection of abnormal sounds and, therefore, detect similar noises. Playing back such music will not change the auditory evaluation of the multi-channel record playback system, and since we are measuring abnormal sounds in the middle and low ranges, which have the greatest impact on the auditory sense, Auditory evaluation and measurement results agree well, and since there are no signal components in the frequency band to be measured, it is easy to design a band film or low-pass filter that collects a lot of abnormal sound components to be measured from the demodulated signal. Furthermore, since it uses random noise, it has many features such as continuous numerical measurements.

[Brief explanation of the drawing]

Fig. 1 is a diagram showing the frequency spectrum of one embodiment of the signal used in the present invention, Fig. 2 is a block system diagram of one embodiment of the recording system of the record for measurement of the present invention, and Fig. 3 is a diagram showing the frequency spectrum of one embodiment of the signal used in the present invention. A block system diagram of one embodiment of a playback system for a record,
Figure 4 is a diagram showing the Tonami characteristics of one example of the main part of Figure 3;
The figure is a diagram showing the relationship between the voltage and occurrence frequency of random noise used in the present invention. 1...Random noise oscillator, 2...
Bandwidth filter, 3...Multi-channel record modulation circuit, 6...Record, 7...
・Pick-up cartridge, 8... demodulation circuit, 9... band filter (or low-pass filter),
10... Measuring instrument.

Claims (1)

[Claims] 1. One or two or more mutually independent random noises whose components are distributed over substantially the entire frequency band from the middle range to the high range, or equivalent signals as a direct wave signal and an angle-modulated wave signal. After that, the multiplexed signal obtained by multiplexing these two signals is played back by the multi-channel record playback system to be measured.
The nature of the random noise or an equivalent signal in the input demodulated signal or the output signal of the arithmetic circuit that is supplied with the demodulated signal of the angle-modulated wave signal and the reproduced direct wave signal and reproduces and outputs the multi-channel signal. A method for measuring abnormal sound in a multi-channel record playback system, which is characterized by measuring components occurring in non-existent middle and low frequency bands. 2. Multiplexing one or two or more mutually independent random noises having a frequency band from the middle range to the high range, or a signal equivalent thereto, into a direct wave signal and an angle-modulated wave signal, and then multiplexing these two signals. A record for measuring abnormal sound in a multi-channel record playback system used in the measurement method according to claim 1, wherein a signal is cut and recorded on at least one of the left and right walls of a record sound groove.