In an analogue recording the frequency and intensity of an original sound is replicated in the form of a varying magnetic field or a mechanical movement. A playback device then converts this back into real sound.
The most common reel-to-reel format is quarter-inch tape. The earliest tape recorders used a tape speed of 120 or 60 inches per second (in/s or ips), although modern equipment works at 30, 15 or 7½ in/s, or sub-multiples of these speeds.
All tape machines work on the same principle. Tape is supplied via a feed spool, passes between tape guides to the tape heads, goes between the shaft of a capstan motor and pinch roller and through further guides to a take-up spool. The capstan ‘evens out’ any speed variations that might otherwise cause unwanted effects, such as wow or flutter.
The tape is first passed in front of an erase head, which is connected to a high-frequency voltage. This randomises the magnetic particles or domains on the tape, effectively removing any existing material.
The tape then passes the record head, which is connected to the amplified recording signal and a high-frequency bias voltage. The latter overcomes the non-linear nature of the tape, improving its frequency response and distortion, but only when adjusted to suit the tape. Having said this, the bias can be increased beyond the optimum level to give over-biasing, where the effects of tape dropouts are significantly reduced, although the high-frequency response is worsened.
The audio signal applied to the record head is also modified by a given recording characteristic. This boosts both the high and low frequencies in the signal, compensating for various losses that occur in the magnetic recording process, as well as minimising background hum or hiss.
Finally, the tape passes the replay head. This is connected to an amplifier which reverses the action of the recording characteristic and boosts the signal to line level.
Although 30 in/s is sometimes used for high-quality music, most recordings are made at 15 or 7½ in/s. The former is invariably used for music whilst the lower speed is perfectly adequate for speech recordings. It’s also reasonably convenient for mechanical editing and splicing, using a chinagraph pencil, razor blade, splicing tape and an editing block.
Most machines use an accepted recording characteristic, such as the American NAB or European IEC/CCIR standard. If you do play a NAB tape on a machine with an IEC characteristic the signal output at 10 kHz is increased or lowered by 3 dB at 7½ in/s and 15 in/s respectively. Any problems with matching an actual tape’s characteristic to a tape machine must be fixed before switching on noise reduction, since this will exaggerate any errors in the frequency response.
A mono professional machine usually employs single-track recording, in which the full width of the tape is used. However, a modern stereo or twin-track machine uses two separate tracks toward the edges of the tape, with an unused guard track running down the middle. Most machines have a wide guard track of 2 mm width, the stereo tracks occupying a similar amount of space. Unfortunately, some recorders employ a narrow guard track, causing discrepancies in levels. A twin-track machine allows separate recordings to be made on each track.
Some older portable recorders use a mono half-track format. This type of machine allows you to record on one side of the tape, turn it over and then use the opposite side. Such tapes can usually be played on a modern stereo machine, although there can sometimes be problems with differences in the guard tracks between half-track and stereo or twin-track recorders.
Although all domestic quarter-inch tape formats are now obsolete you can still encounter such systems. Some of these are variations of professional formats that have been tailored to reduce the consumption of tape. It should be noted, however, that the actual tape used in domestic machines isn’t always suitable for professional models and vice versa. Domestic long play (LP) tape is very thin and can be easily damaged by the powerful transport of a professional machine, whilst the rather abrasive surface of a professional tape can cause damage to the heads of a domestic machine.
Most domestic recorders use lower tape speeds and smaller spool sizes than professional machines. A speed of 7½ in/s is sometimes used, although 3¾ in/s is much more common. For non-critical recordings, usually only containing speech, 17⁄8 in/s or even 15⁄16 in/s can be employed.
Apart from those devices based on professional formats, most consumer machines use the four-track system. This can accommodate up to four separate mono recordings, two on each side, or two stereo recordings, one on each side. If only one recording, in mono or stereo, is made on a single side you can play the result on a professional machine. However, if you create recordings on both sides of the tape the professional machine plays both at once, with one side played backwards.
The four tracks are numbered in sequence across the tape. This means that a stereo recording on the first side uses tracks
3, whilst that on the second side uses
Tapes are made of a durable polyester film, such as Mylar, which is unaffected by humidity. Rolls of this clear film, typically two feet (0.6 m) wide and under 0.001 inch (0.025 mm) thick, are coated continuously and slit into individual tapes.
Most tapes are coated with gamma-ferric oxide. However, high-bias audio tapes, as well as video tapes, contain chromium dioxide or an iron oxide powder treated with cobalt, while a high-performance metal-particle tape contains only iron and other metals. All tapes have needle-shaped particles, typically 5 millionths of an inch thick by 20 or 30 millionths long.
Plastic resins are used as a binder, attaching the powder to the backing, and the particles to each other, but not causing them to clump together. The mixture is applied to a thickness of less than 0.0004 inch (0.02 mm) on the film and a powerful magnet used to align the particles along the length of the tape. The tape is then dried, rolled and cut into the required width.
Professional reel-to-reel multi-track formats include 4-track (½ inch), 8-track (1 inch), 16-track (1 inch), 16-track (2 inch), 24-track (2 inch) and 32-track (2 inch). If your machine has the appropriate tape width you should be able to play a tape created on any machine with a lesser number of tracks. However, you have to work out in advance which tracks to use and you’ll need to realign your machine if you intend to use noise reduction (see below).
Most of these formats run at 15 or 30 in/s, usually with noise reduction in the form of Dolby A, Dolby SR or dbx. Much of the above information about standard reel-to-reel tape also applies to multi-track.
Domestic reel-to-reel multi-track machines are very rare, although attempts have been made with 4-track and 8-track operation on quarter-inch tape. Rather more common is the use of Compact Cassette, often at a speed of 3¾ in/s and in 4-track form, although 8-track machines have also been produced. Recordings made in these formats are usually incompatible with a normal Compact Cassette machine, since both sides of the tape are used at once and the tape runs at twice the normal speed. Some form of noise reduction, such as Dolby B, Dolby C or dbx, is absolutely essential.
This exceptionally popular format was devised by Philips in the 1960s. It began almost as a toy but in its modern form can give a respectable sound quality. However, unlike reel-to-reel tape, the tape guidance system is built into the cassette itself, which causes compatibility problems between some types of cassette and certain machines. For example, cassettes of European origin often have built-in ‘friction’ whilst those from the far-East are designed on the basis that such friction is provided within the machine. So one make of tape will behave perfectly in one machine but not in another.
This split personality of Compact Cassette can also cause problems with variations in head alignment and azimuth. An incorrect setting of the latter can cause a varying loss of high frequencies, especially noticeable when stereo outputs are combined to create a mono signal.
Fortunately, the thin tape required to fit inside the cassette also happens to give a superior high frequency response, although the amount of background noise makes some kind of noise reduction almost essential. The most common systems are Dolby B, Dolby C and Dolby S. In most instance you’ll need to switch the noise reduction on and off manually, although some proprietary systems accommodate data at the start of recording, including noise reduction switching information.
Modern machines can accommodate a wide range of different tape types. However, if your machine isn’t designed for a particular variety of tape you should avoid using it for recording, although playback is often perfectly satisfactory.
The original Type I (ferric) cassette employs a standard recording characteristic of 120 microseconds (µs), which means the tape will sound perfect on any kind of machine. The Type II (chrome and psuedo-chrome) and Type IV (metal) versions use a 70 µs characteristic. This reduces the background noise, although some machines can’t accommodate this characteristic, resulting in harsh sound reproduction. You shouldn’t use these tapes for recording on a machine that isn’t designed for them: the machine won’t provide adequate bias, creating a very edgy sound quality.
The Type III (ferrochrome) version also uses a 70 µs characteristic, but is rarely encountered, since tapes of this kind can give a lumpy frequency response. In addition, automatic switching for this cassette type isn’t provided, meaning you must use manual switching to get the correct bias. Unfortunately, most people forget to operate the switch at a crucial moment.
Provision for automatic tape switching is given by extra notches created in the cassette shell, although some machines ignore these. The Type II notch is located near the record protect notch, whilst the Type IV notch is nearer the centre. Of course, two notches of the appropriate kind are usually provided, one for each side. However, a chrome tape which is pre-recorded on one side will usually have a Type II notch only for the blank side, since the prerecorded material is usually recorded with a 120 µs characteristic to maintain compatibility with older machines.
Pre-recorded material can be produced using fast copying techniques. A master tape is played at around 64 times normal speed, allowing an hour’s recording to be copied in less than a minute. The process normally involved the use of bulk cassette tape, supplied on special pancake reels, which is loaded into empty cassette shells after copying is complete.
Unfortunately, the quality of pre-recorded tapes is variable, especially in older recordings. The problems revolve around high-frequency performance, usually caused by a lack of quality control, either with the tape itself, recording levels, the amount of AC bias or azimuth settings. Unfortunately, Dolby B noise reduction often makes such shortcomings even worse.
The gramophone or phonograph, and successor to the wax cylinder, operates by converting the vibrations of a stylus in the groove of the disk into an electrical signal. Various systems have been used, including hill and dale recording, involving up-and-down movements, and lateral recording, using sideways motion. Modern disks employ a combination of both: the two stereo channels are represented by movements that are 90° apart but both at 45° to the disk’s surface
Many early disks have a nominal rotational speed of 78 revolutions per minute (rev/min or rpm), although the actual speed often varies from 75 to 80 rev/min. Older disks are often made of shellac and can be played with a stylus or needle made of metal or a composite material. Modern ‘78’ styli have a conical or elliptical shape and can be colour-coded as follows:-
|Yellow/Blue||4.0 / 1.2||Elliptical|
|Orange/Blue||3.5 / 1.2||Elliptical|
|Green/Red||2.8 / 0.9||Elliptical|
Modern gramophone records rotate at 45 rev/min for a single and 33 rev/min for a long play (LP) or extended play (EP) disk. These vinyl records must be played with a modern stylus that has a diamond or sapphire tip. The shape of a stylus is crucial to reproduction, although for LPs a standard shape is normally used.
All disks use an equalisation characteristic to minimise hum and hiss. This applies a bass and high-frequency boost during recording, with a complementary cut during playback. A plethora of characteristics are used for ‘78’ disks, as listed below:-
|3180 µs/||Bartok, Capitol, |
|3180 µs/||Angel, |
• US only
* Some recordings use -/450 µs/- or -/636 µs/- characteristic
All ‘78’ recordings produced after 1954 use the 3180 µs/450 µs/50 µs characteristic, whilst all modern disks use the standard RIAA characteristic.
The mechanics of a gramophone player is a bit of a nightmare. The tone arm, carrying the stylus, is usually moved across the disk by the groove itself, although this mustn’t put undue force on the groove, causing distortion or damage to the disk. A servo system can reduce the force, although attempts to do this over the years have met with varying degrees of success.
Ideally, the stylus should always be tangential to the spiral of the groove. This isn’t possible with a conventional arm, although various kinds of linear-tracking arms have been devised.
Finally, there are the limitations of the medium itself. For example, the disk spins at a constant rotational speed, meaning that as the stylus gets closer to the centre the linear speed falls. Hence the noise, distortion and frequency response gets worse towards the middle. Rather a shame really, especially since the best bits are usually at the end of a record!
The master disk for an LP is made using a cutting stylus driven by electromagnetic coils, connected in turn to an audio amplifier. This is then electroplated, this metal creating a metal master, which is a negative version of the original disk. This is plated to produce a positive mother, which in turn is plated to create several stampers. These are then paired up within the top and bottom of a record press, into which a vinyl ‘biscuit’ is inserted and pressed for around 20 seconds. Once the new record has cooled it’s removed, the hole is punched in the centre and it’s labelled.
The inadequate dynamic range of analogue equipment has led to the development of noise reduction (NR) systems. These involve a trade-off of an enhanced range against other complications. They require an encoder or compressor at the input of an analogue device, such as a tape recorder, and a decoder or expander at the final destination.
Professional noise reduction systems often use a line-up level of
+4 dB, corresponding to
PPM 5 or
0 VU. To test such a system, move the
NR switch on the noise reduction unit to the
OUT position and check that the tape machine is getting the correct level. Now move the
NR switch to
IN and adjust the record level control on the unit to give the same level at the machine. A similar process should be used to set up the replay level control on the unit. If the line-up is correct you should be able to record a signal on the machine and operate the
NR switch without any changes in level occurring.
The most common systems, produced by Dolby Laboratories, are as follows:-
This is the original Dolby system, devised for use with a multi-track tape machine. It gives a 10 dB reduction in noise at all frequencies, although care must be taken with line-up to ensure compatibility with other tape machines and studios.
This is the standard noise reduction system for Compact Cassette, giving a 10 dB improvement at high frequencies, but only when the line-up is correct. Unfortunately, many prerecorded cassettes are at low level, giving reduced high frequencies for quiet sounds, whilst the system also exaggerates problems with head alignment. Sadly, during playback, many people switch off Dolby B to get an emphasised high frequency performance.
An enhanced system for Compact Cassette, giving a 20 dB improvement at high frequencies and effectively consisting of two ‘cascaded’ sets of Dolby B electronics. Unfortunately the results aren’t entirely acceptable on a Dolby B machine.
The Dolby Headroom Extension (HX) system isn’t really a form of noise reduction, but is often used with it. Whilst making a recording, the tape recorder’s bias voltage is adjusted to suit the high-frequency content of the signal. This creates a high-level recording with reduced distortion that can be played on any type of machine, with or without noise reduction.
Spectral Recording (SR), recently used for multi-track tape and 35 mm film recording, provides a 30 dB improvement in noise levels at all frequencies. The results are comparable to those obtained using digital technology.
The third-generation system for Compact Cassette, giving a 30 dB improvement at high frequencies and providing good compatibility with Dolby B and non-Dolby machines.
Other types of noise reduction include:-
A reasonably low cost noise reduction system that effectively doubles the existing dynamic range of a system by employing 2:1 compression and expansion. It also uses pre-emphasis to minimise breathing effects, although such artifacts are often audible on certain types of material.
Similar to dbx but employing average rather than root mean squared (RMS) signal detection, reducing cost with little effect on quality. As used in Bel noise reduction and other products.
A playback-only system, also known as a single-ended process, as devised by Philips to compete with Dolby B. It reduces high-frequency noise by about 10 dB, using a sliding-frequency filter whose cut-off frequency is determined by the sound level. When used with signals at the correct level it can be very effective and unobtrusive but can give muffled results if the level is too low. The universal use of Dolby B for pre-recorded cassettes means that DNL has never been popular.
SMPTE timecode, containing information about the time of day of recording, allows a tape machine to be synchronised with other recording machines, video equipment or computer-controlled devices. It’s accurate to one frame, the duration of which is set by your television standard.
In the United States the NTSC standard is used, running at a frame rate of around 30 frames per second (frm/s or fps), whilst Europe uses the PAL system, with a rate of 25 frm/s, and cinema films run at 24 frm/s. Ideally you should use only one standard, otherwise you may need special equipment.
SMPTE timecode can be recorded onto an analogue tape machine as longitudinal timecode (LTC). This is encoded using bi-phase modulation, employing the following frequencies:-
|Frame ||Logic ‘0’ ||Logic ‘1’ |
An example of the signal waveform is shown below:-
In practice, the signal is ‘rounded off’ to make it suitable for transferring directly onto an audio recording device. For a rate of 25 frm/s the average frequency is between 1 and 2 kHz whilst the period length is around 500 milliseconds, with rise and fall times of about 50 microseconds. Overshoot, undershoot and tilt should be less than 5% of the peak-to-peak amplitude.
LTC can be recorded is several ways. If you only need mono sound you can record the audio on the left-hand channel of a professional quarter-inch tape machine, with LTC recorded on the right-hand channel. However, there’s a risk of audible crosstalk from the timecode to the audio track.
However, if stereo sound is required you’ll need to use a machine with a separate timecode track. This employs an extra head, with a track width of 0.38 mm, located centrally in the 2 mm of wide guard track that’s used on a modern machine.
In some machines the timecode track doesn’t use normal pre-emphasis and de-emphasis, making the track unsuitable for other purposes, such as an audio cue track. Some machines use separate heads for timecode and audio, requiring a fixed delay circuit to keep timecode and audio in step. Other models use combination heads, such as a combined head for audio erase and timecode playback with another for timecode erase and audio record. Once again delay circuitry is required.
On an analogue multi-track machine you should record timecode on an outside track, such as track 16, 24 or 32. However, most machines only reads timecode at normal playback speed. To read timecode at speeds between 50 times and 1/10 of playback speed often requires modifications to the machine.
Timecode can be used to synchronise several devices, one of which is designated as master whilst the others work as slaves. A synchroniser consists of at least two timecode readers, connected to the master and slave machines. This compares the timecode from each machine and uses this information to control the function and speed of the slave machines. The synchroniser’s control unit may also allow you to apply a fixed timecode offset between chosen machines.
More advanced studio systems use a controller for each machine. Such controllers are linked together by a control bus such as ES-Bus. Connections for video, audio, cue track, timecode and CTL pulse circuits also be provided for the machines.
1997 Grolier Multimedia Encyclopedia, © 1997, Grolier Inc.
©Ray White 2004.