.

Around the time of the Second World War, audio from performances in studios and halls would be input directly into record cutting machines without modification – a method called direct cutting (also referred to as direct-to-disk recording).

However, as this method made editing or re-recording (cutting another master disk) to correct the performance and adding various other effects difficult, after the end of the war, from the 1950s onward, magnetic recorders came to be introduced.

Magnetic recorders only had enough tracks for monaural recording at first, and two tracks later became popular as they entered the era of stereo, but 3-track recorders (Anmerkung : von Ampex 1952) quickly appeared on the market with the arrival of the era of American pop music to facilitate the adjustment of songs placed in the center.

Later, as “British sound” became came into fashion in the latter half of the 1960s, the number of tracks gradually increased.

The number of tracks for magnetic recorders during that period could be broadly classified into two categories: 2-track/4-track recording and 8-track/16-track/24-track/ 32-track multi-track recording.

The former mainly used a small number of microphones for pickup, and the latter mainly used multi-microphone pickup. An analogue 2-track recorder is shown in Fig. 5.13.

A similar analog 2-track recorder for cutting that could produce advanced signals to control groove pitch and depth is shown in Fig. 5.14.

This reproducer was equipped with bypass and extension rollers and an additional playback head to obtain the advanced signals necessary for controlling groove depth and pitch, and it selected an optimal advance signal depending on the speed of the tape and the rotational rate of the disk, which it then used to regulate the cutting machine.

Figure 5.15 shows a Studer A827 24-track analog recorder from Switzerland.

At the end of the 1960s, disk records reached a plateau in terms of the technology developed up to that point, sales stabilized and there was a desire for records with added value.

From this state of affairs came the desire for a new breakthrough. During this period, record companies attempted to make various improvements in recording and in the record manufacturing process, and these were actually put on the market and reviewed by audiophiles. Some examples of these improvements are shown in Table 5.16.
.

Table 5.16 compares the production and manufacturing process for normal records shown in the top row with the improvements (A) to (G), and assesses what methods were actually accepted.

These improvements captured the interest of audio fans, and as they demanded that they be able to judge the improvements with their own playback systems and their own ears, and as far as possible the various improvements were elucidated when the records were put on sale.
.

.

• (A) These were 30cm, 33 1/3-rpm LP records that had their rotational rate increased to 45rpm. (bei uns Maxi-Singles genannt) These records were released into the market in 1967, and although they had the drawback of being limited to certain kinds of music by their short recording time, they gained support from audiophiles (Anmerkung : und Diskotheken).
• (B) Master Press records were records that were pressed directly from master disks without having made a mother and stampers, and as they omitted the process of making a mother and stampers, they were unsuitable for mass production. Priceless high-value classic editions which would only be available in limited runs of approximately 1,000 copies were sold and were received well.
• (C) Dolby S/N Stretcher was a system developed by Dolby Laboratories in England (who later transferred their headquarters to the U.S.) (Anmerkung : Ray Dolby war ein Engländer, derinUSA studierte und bei Ampex angestelt war) to reduce noise created by magnetic tapes during playback.
  
  This system reduced noise by increasing the low-level signals that contained undesirable noise above the actual levels during recording, and decreasing them again during playback. S/N Stretcher was effective in analog multitrack recordings of pop music in particular, because even if parts with many instruments playing were fine, when only a few instruments were playing, the signal from other instruments was, of course, not added, but the noise from each of the tracks was. Dolby Laboratories released a form of S/N Stretcher for professional use called Type A, and later developed and released S/N Stretcher for cassette tape as Type B and Type C.
• (D) Half-speed cutting is a technique that aims to improve the cutting characteristics and sound quality of a cutter head at high frequencies, and involves sending the signal to the cutting machine at half speed and reducing the cutting speed by half. Cutting systems are generally designed to cover the entire range of human hearing from low to high pitch.
  
  Accordingly, half-speed cutting with music that contains bass increases the risk of decreased levels of bass or distortion. To avoid these risks, the introduction of units that could reproduce and deliver the signal with a stable characteristic even at low frequencies of around 10 Hz (such as PCM/digital recording and replaying machines) was necessary. Even with analog recording, the effectiveness of half-speed cutting for music that does not contain much bass, such as shamisen music, was well known.
• (E) Direct cutting bypassed the magnetic recorder, making it possible to improve the noise, harmonic distortion, modulation distortion, and wow and flutter associated with the use of such units.
• (F) Playback distortion compensated cutting records were made with a technique that carried out correction during cutting to reduce the distortion caused by differences in the shapes of the cutting styluses and playback styluses introduced in 1970 (see Section 5.4, Fig. 5.54 to 5.58).
• (G) These records are manufactured with a process that substituted PCM/digital recorders for former analog magnetic recorders. Records manufactured as a trial were released in 1970 (Fig. 5.21, Fig. 5.22), with sales continuing in the form of LPs and CDs from 1972 up to the present (Fig. 5.24).

.

.

• Anmerkung : Warum hier das von Telefunken / TELDEC perfektionierte "Metal Mastering" nicht erwähnt wird, ist nicht ersichtlich.

.

The method of improvement that was commended as having the greatest improvement among methods (A) to (F) described above in terms of sound quality was direct cutting, which eliminated the noise, harmonic distortion, and modulation distortion created by magnetic recorders.

One of the first experiments in (modern) direct cutting in the era of LP records was at "Nippon Columbia’s Studio No. 2" in Akasaka, Tokyo in 1969, where the output signal from a live performance was sent to three cutting rooms downstairs in the same building.

This differed from ordinary cutting as it was impossible to know the groove pitch and groove depth in advance, so three cutting machines were each set with different cutting levels, groove pitch, and groove depth to provide levels that varied from a wide safety margin to nearly no safety margin.

Recording was then carried out with the expectation of success. Flamenco, jazz, and Argentine tango were selected as they had improvisation and mistakes were not very likely to occur.

The records were recorded at 45 rpm, which took sound quality and the frequency of musical and technical mistakes into consideration, and gave a little over 15 minutes of playback per side. These were released as 45-rpm doubles with a total of just over an hour of playing time.

Some direct cutting disks released at the time are shown in Fig. 5.17, 5.18, and 5.19.

After direct cutting, in second place with good reviews came half-speed cutting, which could cut high pitches cleanly by reducing the cutting speed by half.

The other methods (Master Press, Dolby S/N Stretcher recording, playback distortion compensated cutting, etc.) were introduced, and although they were well received to a certain degree, they did not gain much recognition.

On the basis of the above, it was concluded that simultaneous direct cutting and half-speed cutting would be ideal, although it was thought to be theoretically impossible.

Around the same time in broadcasting, the good sound quality of live broadcasts without intermediate magnetic recorders was recognized.

Between 1967 and 1969, Kenji Hayashi and others at NHK STRL worked to improve the modulation distortion created by magnetic recorders, developing the world’s first PCM/digital recorder (first in mono, later in stereo), using an industrial helical scanning VTR as a recording machine.

Although this recorder did not meet the specifications for being installed at production sites because it could not do editing, nonetheless it was more than adequate as an experimental machine, and took the world by surprise.

Despite that, however, it was not very well received at NHK production sites, and the developers of this PCM recorder sought opportunities to participate in evaluation experiments elsewhere.
.

• Anmerkung : Es fehlt eigentlich eine Erklärung, warum dieser "overwhelming superior" PCM Recorder bei den NHK Kollegen gemieden wurde.

.
This PCM/digital recorder from NHK STRL was used in the assessment experiments conducted by Hitachi Group described in Section 5.1.2 (3), producing valuable evaluation results for participants.

In two channels, the PCM/digital so was overwhelmingly superior to the conventional recorders that all of the people involved in the experiment were led to wonder if it would be better to practically implement PCM/digital recording as a priority over 4-channel sound, and focus on 4-channel sound later.

Four-channel sound did, in fact, die an early death (Anmerkung : etwa ab 1976), and a few years later PCM/digital recording was made a practical reality, and approximately 10 years later, digitally recorded music reached homes in the form of CDs.

Takeaki Anazawa, who facilitated these evaluation experiments, wasted no time in borrowing the PCM/digital recorder shown in Fig. 5.20 from NHK STRL, and began conducting full-scale tests of PCM/digital recording at studios and concert halls with the help of Nippon Columbia’s recording staff and production staff.

During recording tests with NHK’s PCM/digital recorder, the two analog disks shown in Fig. 5.21 and 5.22 were born. The record in Fig. 5.21 was released by Nippon Columbia in January 1971, and the recorder developed by NHK STRL was used to record it. Still, as it was unable to edit the audio, its operational efficiency and the amount of audio yielded were only slightly better than with direct cutting.

Figure 5.22 shows a live recording released in April 1971 made on a recorder (without editing capabilities) developed by NHK STRL.

Required specifications for PCM/digital recorders (1969) and the performance, configuration and format of the Denon DN-023R (1972)

.

• (1) Wide dynamic range.
• (2) Low distortion rate.
• (3) No wow or flutter.
• (4) No crosstalk between channels.
• (5) Flat frequency characteristic across a wide range of frequencies.
• (6) 4 channels of multi-channel recording/reproduction of more.
• (7) Ability to copy data from a single channel.
• (8) Capable of half-speed cutting.
• (9) No loss of sound quality during storage.
• (10) No ghosting, etc.
• (11) Equipped with advanced head for cutting.
• (12) Capable of editing

.

.

• (1) Dynamic range: > 76 dB.
• (2) Distortion rate: < 0.1% (with active levels).
• (3) Wow/flutter: undetectable.
• (4) Crosstalk between channels: -80 dB or lower.
• (5) 0 (DC) -20 kHz (deviation of +0.5dB or lower).
• (6) Eight channels of recording possible.
• (7) Possible to copy all of the data with 2 recorders, individual data copy possible with two systems.
• (8) Possible: 0 (DC) -20 kHz (deviation of +0.5dB or lower). (9) Data errors observed to increase after storage, but without a detectable decrease in sound quality.
• (10) Ghosting level: -80dB or lower.
• (11) Approx. 1 second (1/2 a revolution of a 33 1/3-rpm record) advanced stereo signal output possible.
• (12) Capable of tape cutting/editing.

.

.

• (1) PCM/digital audio converter
• (2) 4-head, low-band VTR (modified)
• (3) VTR activity monitor
  
  
• (4) Format
• Modulation method: Pulse code modulation (PCM)
• PCM signal format: 13-bit natural binary code
• Transmission clock frequency: 7.1825 MHz
• Voice sampling rate: 47.25 kHz
• Transmission waveform: Standard NTSC TV signal (excluding vertical synchronizing signal)
• No. of voice channels: 8/4/2 switchable
• Advance signal recording method: Direct recording of analog signal
• No. of advanced signal channels: 2
• Magnetic tape device: Modified 4-head low-band VTR
• Tape speed: Recording: 38 cm/sec.; playback: 38 cm/sec./19 cm/sec. (switchable)

.

The specifications shown in Table 5.2 were established at the same time as the recording experiments described above were conducted with the help of people such as Kenji Hayashi (former head of the Consumer Products Research Center, Hitachi Ltd.) who oversaw development at NHK.
.

• Anmerkung : Offensichtlich - den Bildern nach zu urteilen - war es 1970 noch eine 2" Quadruplexmaschine mit 4 Köpfen. Die erste 1" Profi-Videomaschine wurde 1974 auf der Messe in Montreux vorgestellt.

.
Later in 1970, with the approval of the new management of Nippon Columbia, a joint development project between NHK STRL and Nippon Columbia was commenced.

Development was carried out by Hayashi, Miyashita, Todoroki, Oshinden and Anazawa, who were Koichi Tsubota’s junior colleagues at the Nippon Columbia factory in Mitaka City where he worked before and after WWII.

In 1972, exactly 10 years before digital recordings reached consumers in the form of CDs in 1982, Nippon Colombia released the Denon DN-023R, the world’s first PCM/digital recorder for record mastering (Fig. 5.23) was completed, and it began to be installed in recording and record production sites from April of that year.
.

Digital recorders such as this require data recorders that can record and reproduce digital data in a stable manner.

NHK’s PCM/digital recorder shown in Fig. 5.20 used a one-inch industrial VTR as a digital data recorder. Recording tests with this recorder at recording sites revealed a problem with its stability during operation as a data recorder.

Other issues such at its lack of editing capabilities emerged, and the importance of drastically improving it was recognized. If eight channels of audio data with a 48 kHz sampling rate and a 16-bit bit depth are to be recorded and reproduced simultaneously, then the machine must be able to record and reproduce digital data at a rate of at least six megabits per second.

Moreover, there were no data recorders at the time that could cut and edit recorded data, produce advanced signals (signals advanced by approximately one second) to control groove pitch and depth during cutting, and replay the tape at half speed for half-speed cutting.

After searching for a data recorder while considering factors such as operational stability and ease of modification, a four-head VTR for broadcasting use was settled upon.

By coincidence, Shiba Electric Co., Ltd., who were manufacturing four-head VTRs for broadcasting stations, became part of Hitachi Denshi Co., Ltd. (now Hitachi Kokusai Electric, Inc.) at nearly the same time as Nippon Columbia became a member of Hitachi Group.

So the four-head VTRs began to be manufactured within the Hitachi Group. The man who took charge of modifying the VTR at Hitachi Denshi was Toshiaki Kawamura, who supervised the writing of History of the Video-Tape Recorder in Japan and the Preservation of (Early) Examples as a systematization report for the National Museum of Nature and Science in March 2001.

The recorder portion of the PCM/digital recording equipment shown in Fig. 5.23 (on the right of the picture) used a 4-head low-band black-and-white VTR for broadcasting use. This VTR had lost its place in broadcasting stations with the transition to color for the 1964 Tokyo Olympics, so it was used to record digital audio in the form of white dots.

As this recorder was developed during the era of 4-channel sound, it had the capacity to record eight channels, and it was configured so that reducing the number of channels would increase the accuracy of the recording.

Editing, which is necessary in record production, could be done in the form of cutting and splicing the tape by hand.

It also had the ability to produce advanced signals to control groove pitch and depth during cutting, and deliver audio data at half speed to enable high-fidelity cutting at high frequencies.

The first recording made on the recorder shown in Fig. 5.23 was made in April 1972 at Aoyama Tower Hall in Tokyo, where the Smetana Quartet, pride of Czechoslovakia, played Mozart’s String Quartet No. 17 “The Hunt” and other pieces.

At the time, string quartets were a kind of music that traditionally demanded the most precise of ensembles (unlike jazz, flamenco, Argentine tango, and percussion ensembles), and getting consent from the performers to release the record without editing was impossible. Figure 5.24 shows the recording made in April 1972 being edited.

During editing, the analog track of the 2-inch tape recording is played and the cutting point is found by listening to the reproduced sound.

By doing this as described above, these analog signals are also used to change the groove pitch and depth of the disk during cutting. Then a method was used where the recording side of the 2-inch magnetic tape near the editing equipment was coated with magnetic powder dissolved in a volatile liquid, the recording tracks were located under a microscope, the tape was cut on both sides of the splice where the recording tracks joined, the magnetic powder was cleaned up, and then the tape was taped with thin aluminum splicing tape.

Tens of thousands of splices were made using this editing method over the course of approximately eight years, continuing until the appearance of the random access editing system shown in Fig. 5.36.

Now, 40 years later, editing can be done on laptop computers that have miniature high-capacity memory media and signal processing capabilities.

The first product to be recorded using the world’s first proper PCM/digital recorder with editing capabilities, and then recorded as an analog record was released in October 1972. This record is shown in Fig. 5.25.

A small and robust recorder was made two years later in 1974, and in the same year, digital recording was commenced in Western Europe.

The first digital recording in Eastern Europe was made in 1975, and the first digital recording in the U.S. was made in 1977.

French president Charles de Gaulle had ridiculed (verspottet) visiting Japanese Prime Minister Hayato Ikeda in 1962 by referring to him as “that transistor radio salesman,” but Japan’s PCM/digital recording work in Europe and America put an end to that and views toward Japan changed, and not only France but also other countries began to present Japan art-related awards and technology-related awards. Some of these awards are listed in Table 5.3.

.

In this way, the dream of the Japanese people to see great works of art perfected was achieved with the help of technology.

Figure 5.26 shows the exterior of the Alte Oper in Frankfurt, where the celebrated Eliahu Inbal conducted the complete Mahler symphonies, and Fig. 5.27 shows a scene from the recording session.

In 1985, after much discussion and practice over the space of ten years, the recording of the complete collection of Beethoven’s string quartets with the Smetana Quartet was completed. Soon after the project was completed, the quartet disbanded.

Recording of Hibari Misora’s album, the complete collection of Beethoven symphonies conducted by Otmar Suitner, and jazz music in the U.S. were also completed, and by the time CD releases began in 1982, there were nearly 400 titles of digital content.

To find the theoretical dynamic range of a given bit depth, the bit depth is multiplied by 6 dB and 1.8 dB is added:

therefore the dynamic range of 12-bit audio is 73.8 dB, and the dynamic range of 16-bit audio is 97.8 dB. Although the bit depth can easily be increased by increasing the number of comparators, achieving precision is a completely different problem.

Sixteen-bit numbers are easy to represent, but actually achieving 16-bits of accuracy over the entire target frequency band with enough absolute accuracy and relative accuracy is extremely difficult.

The bit depth and dynamic range of recorders increased from the end of the 1960s, arriving at the 16 bits per channel specification of the CD in the latter half of the 1970s. This progress is shown in Fig. 5.28.

The recorder with a 15.5-bit equivalent characteristic indicated in Fig. 5.28 is the Denon DN-034R, which is shown in Fig. 5.29. This recorder was transported to New York in the latter half of the 1970s and was used to make 8-track multitrack jazz recordings. The recordings received Billboard’s Trendsetter Award in 1978 for being America’s first commercial digital recording.

The upper part of this particular recorder was a Hitachi Denshi SV-7400, which was famous in the era of 4-head VTRs, and installed underneath was a time base corrector unit in place of the processor shown on the bottom in Fig. 5.29.

This VTR was registered as one of the Essential Historical Materials for Science and Technology by the National Museum of Nature and Science and is on display at the NHK Museum of Broadcasting.

Digital audio processors (Sony PCM-1600, PCM-1630) that used a semi-professional U-matic VTR as a recorder appeared at the end of the 1970s, a few years before the introduction of the CD, and the Denon DN-035RmkII, a 4-channel, 16-bit digital recorder that used the same U-matic tapes and U-matic VTRs appeared at the same time. Figure 5.30 shows a Sony PCM-1630 and a U-matic VTR, while Fig. 5.31 shows a Denon DN-035RmkII.

Sony and Mitsubishi Electric also sold fixed-head digital recorders that did not use VTRs for recording. Figure 5.32 shows a Sony 2-track fixed-head digital recorder. Later, multitrack recorders also appeared.

Figure 5.33 shows a Mitsubishi Electric 2-track fixed-head digital recorder.

Figure 5.34 shows a Sony PCM-3348 48-track digital recorder, and Fig. 5.35 shows a Mitsubishi Electric X-850 32-track digital recorder. Both were fixed-head recorders.

This type or recording system was in use around the time when CDs began to be sold in 1982.