Contents - Previous - Next

8.3 Guide to technical equipment: audio archives

Dietrich Schüller, Lloyd Stickells and William Storm

1. The Archivists Goal

Before describing re-recording techniques, we must first establish what re-recording means to the audio archivist. Re-recording can mean different things to different people. To a collector it may simply mean the copying or transfer of discs to tape for personal pleasure and use. To a commercial operator it may represent a means to generate new revenues.

To the audio archivist, re-recording serves a very specific function; it is a means to preserve the original sonic content of a recording.

In the first two examples, subjective manipulation of the sound during the transfer process is common. The collector may alter the original sound to suit his own taste or the peculiarities of his play-back system. The commercial operator may choose to sell "enhanced" recordings in the belief that they will be more acceptable to modern audiences. These subjective re-recording processes are a luxury some might call them abuses not permitted the audio archivist.

The archivist's function is to preserve history, not to rewrite it. Given this precept, the archivist must always strive to maintain objectivity in the application of various re-recording techniques. This rule is easy to state but can be extremely difficult to put into practice.

The first obstacle is to determine what constitutes an audio archival recording. Current thinking in the field defines several types of re-recording of which three are legitimate for archival consideration.

- Type 0 Inter-archival "replica"
- Type I Audio history
- Type II Re-creation of the original sound source
- Type III Subjective transfer

Type III contradicts the role of the archivist and is not acceptable for archival purposes. Types 0, I and II, however, all deserve serious consideration by archivists.

1.1 The Three Types of Archival Re-recording

1.1.1 Type 0: The Replica

In Type 0, " . . . the intention must be to provide the researcher with a secondary source which is the equal (or at least equivalent) to the original source in as many respects as possible. Indeed, a replica might be preferred and, at least, imagined. If we take one side of an old disc recording, one might obtain a replica by making a negative impression in a suitable polymer and then make another positive from that. Alternatively, one might do what was certainly performed in 1903: grow a metal negative from the disc recording for use as a stamper in a vinyl compound."

Except for cylinder "replicas" made by pantographic means, today's audio archives do not have the mechanical, chemical or human resources to replicate original disc and cylinder recordings as previously described. This is not to say that the idea should be dismissed: it is simply not current practice. For this reason, Type 0 recording will not be discussed further in this paper. It is, however, a topic that deserves future consider ation.

1.1.2 Type I: Audio History

What did a 1910 recording sound like to the people of that time? That is a question addressed by Type I re-recordings. A Type I re-recording is defined as the perpetuation of the sound of an original recording as it was originally reproduced and heard by the people of the era. A Type I re-recording is meant to give listeners a literal historic perspective on the recording and reproduction of sound. This implies the use of the same type of recording machine and medium as used for the original recording.

The original replay equipment, however, introduces replay distortions typical of the technology of the period while modern equipment will normally only add minor replay distortions. The modern equipment will, therefore, faithfully reproduce the sound captured on the historical carrier with real fidelity. The use of modern equipment to replay historic recordings is, therefore, a sub-group of Type I re-recording and is defined as Type IA. Type IA re-recordings will include all the distortions introduced by the original recording equipment but not those introduced by the replay equipment of the period.

1.1.3 Type II: Recreation of the Original Sound Source

What did Enrico Caruso really sound like in person? This a related but different question to that addressed in Type I re-recording. Historically, all recording and reproducing systems have inherently distorted, to various degrees, the "true" sound (artist) that they have tried to capture.

In a Type II re-recording, the re-recording engineer tries to determine what these distortions are and then, using that in formation, reconstructs as objectively and as accurately as possible a faithful reproduction of the original sound source (artist).

In short, a Type II re-recording strives to recreate the sound of the original sound source (artist).

2. The Re-recording Studio

Given the precept "the function of the archivist is to save, not to rewrite, history" the following guide-lines should be observed in setting up a re-recording laboratory for sound.

2.1 Equipment and Acoustics

2.1.1 Electronic Equipment. All equipment should be calibrated to ensure linearity of the signal path ie. what goes in should come out unaltered. There should be little or no coloration of the original sound because of equipment deficiencies. Properly installed and maintained professional, not consumer, standard equipment has little difficulty in meeting this requirement. All equipment should be calibrated to ensure linearity of the signal path ie. what goes in should come out unaltered. There should be little or no coloration of the original sound because of equipment deficiencies. Properly installed and maintained professional, not consumer, standard equipment has little difficulty in meeting this requirement.

2.1.2 Loudspeakers

The loudspeaker system is worth special consideration. Loud speakers are still the weakest link in the audio chain and yet they are vitally important to our perception of the sound. It is, after all, the loudspeaker system that couples the electronic systems to our ears and minds. Underestimating the loudspeakers importance is a major mistake as it will undoubtedly influence critical recording decisions. Speakers selected to reproduce sound objectively should:

a. have a wide frequency range at least 25 to 16,000 Hertz.
b. Have a flat response +/-3dB or less.
c. Have low intermodulation distortion 0.5% or less.
d. Have minimal differences in arrival time at the listening position of sound from the different drive units within the speaker one millisecond or less.
e. Be properly placed in the room for optimum performance.

2.1.3 Room Acoustics

Item 2.1.2 e. above refers to yet another major acoustic weak link in the audio chain - the listening environment. An acoustically poor room can defeat even the best loudspeaker system. As with the electronic equipment and the loudspeaker, the room itself should not colour the sound. Archivists should be aware of its importance and consult experts in the field prior to setting up a laboratory.

2.1.4 Summary

In summary, the equipment and acoustic environment used in a re-recording laboratory should be of a professional quality and capable of accurately measuring, recording and reproducing the signals fed into the system.

Signals in this case, refers to original recordings or reproducing systems. These are the subjects of study whose characteristics are to be examined. such an examination will be impossible if the re-recording testing, recording and reproducing systems are not neutral to begin with.

2.2 Personnel

Given sufficient funds and the correct technical advice any archive can set up a re-recording laboratory. Success for such a facility will, however, only come with the employment of trained professional personnel to perform the work. The degree that the staff must be technically competent will vary with the types of re-recording they are expected to produce.

Type I electrical re-recording requires significantly less technical equipment and knowledge than is needed for Type II re-recording or for Type I acoustic re-recording. In neither case, however, should re-recording be performed by persons not trained in the idiosyncrasies and requirements of the audio recordings that they wish to re-record. Without such knowledge, irrevocable damage can be done to the original recordings.

In short, despite otherwise good intentions, do not allow anyone to re-record original records unless they are qualified to perform the work.

3. Practical Applications

3.1 Commercial Discs

3.1.1 Overview of Re-recording commercial Discs

Not all archives are going to have the budget and personnel to perform all types of re-recording but, given some training and the right equipment, Type IA electrical re-recording of commercial 78, 45 and 33 1/3rpm discs should be within the capability of most archives. Re-recording commercially issued 78rpm discs is a reasonable starting place.

The re-recording would take the following sequence:

a. The record is selected.
b. The record is thoroughly cleaned.
c. The proper stylus (needle) and tracking force (weight) are determined.
d. The record is played on the turn-table at the correct speed.
e. The signal is sent to a pre-amplifier.
f. The pre-amplifier's output is sent to the recorder and the power amplifier. The power amplifier, in turn, sends a boosted signal to the loudspeaker so that the signal can be heard. Once recorded, the tape signal can also be sent to the power amplifier and loudspeaker to check the quality of the copy and for listening purposes.

The tape recorder selected for producing master preservation copies should be an analogue reel-to-reel machine. Full track recording is recommended for mono recordings and half track stereo for stereo recordings. This system and procedure is similar to most people's home stereo operation. There are, however, a number of very important differences. The stylus is not the same as used on modern stereo recordings. Over the history of 78rpm recordings, stylus size and tracking force varied. So the first important difference is stylus size and tracking force and their proper selection. Please also note that the 78rpm disc was monophonic. This factor must be taken into account in the electrical set-up. In addition to stylus variation, manufacturers also varied the electrical play-back characteristics of 78rpm discs. This factor must be taken into account when playing back the discs. The simplest way to manage this problem is to use a pre-amplifier that permits the selection of different play-back equalization curves. This is not a major problem as pre-amplifiers built for 78rpm discs usually have labeled selector switches or some accompanying literature that indicates the proper curve to use for various makes and labels of disc. In addition, these specialized pre-amplifiers will usually convert a stereo cartridge output signal into the mono signal required for re-recording 78rpm discs. Proper selection of stylus size and tracking force accompanied by correct play-back and equalization does not solve all the problems that might be encountered. Accounting for these variables will, however, normally permit acceptable re-recording of most commercially issued 78rpm discs that are in reasonably good condition. In fact, re-recording personnel who have not accounted for these factors and have resorted to the use of noise reduction devices without first understanding basic play-back parameters, will find that most of the unwanted noise is created by not employing the correct play-back equalization.

3.1.2 Choice of Equipment

Given this general perspective, a number of practical points must be considered in selecting re-recording equipment. Nowadays, record players capable of playing 78rpm discs are becoming increasingly difficult to find. The same will apply to 33 and 45rpm disc players before many years are gone. Indeed, most record companies have already ceased to produce vinyl discs in favour of the CD.

The requirements for equipment to play predominantly shellac based 78rpm discs are basically the same as for the vinyl LP with the following points being emphasized:

a. Variable speed turn-table.
b. More robust cartridge cantilever.
c. Range of stylus sizes.
d. Choice of replay equalization characteristics. Taking these points in turn:

a. Variable Speed Turn-table - Usually the only reason for having a "pitch" control on a 33rpm turn-table is to allow the user to "tune" the record to a "live" instrument. The speed of the record machine is likely to have been close enough to the nominal speed of the replay machine to satisfy all but those blessed (or cursed) with "perfect pitch".

Variable speed is, however, an essential requirement for playing so-called 78s because many of them were in fact recorded at speeds greatly different to that nominal speed. 80rpm was a common speed for many of the earlier recordings and some early recordings on the Victor label were recorded at 76rpm. A range of speeds from 60 to over 86rpm may be necessary to cope success fully with acoustic recordings made prior to about 1920. Pathe records, in particular, usually run at speeds between the high 80s to over 100rpm. A speed range of between 72 and 84rpm will, however' cover over 90% of commercial pressings.

To achieve this variation it may be necessary to use a professional machine although it may be possible to find an example of a good quality "domestic" machine such as those made by Lenco and Thorens. These will serve perfectly well at a cost much less than that of a professional model. Ideally a means of speed read-out is a desirable feature to permit the rapid setting of play-back speed where this is known.

b. Robust Cartridge Cantilever The tracking force requirement of the pick-up for 78rpm discs is usually more than that required for the satisfactory tracking of a vinyl LP, even a warped one. For example, a two or three millimetre ripple on a record revolving at 33rpm is not going to put as much strain on the stylus assembly or try to throw it as far out of the groove as the same degree of warp on a record revolving at 78rpm. Badly warped 78s can require tracking weights of 100 mN (10 grams) or more. A typical 78rpm disc will require a tracking weight of up to 50 mN (5 grams). Very badly warped discs may even have to be played, and re-recorded, at half speed to keep the stylus tip in contact with the groove walls.

c. Styli - A vinyl LP has a well defined groove size and an equally well specified stylus shape to play it. The only real choice is between a round and an elliptical stylus profile. While the same is, to some extent true for 78s, very few manufacturers offer 78 coarse groove styli as standard and a far greater range of stylus sizes is needed to get the best results from a range of records.

The normal size is a 65 µm (0.0025 inch) radius, spherical ended cone but a range of sizes from 50 µm (0.002 inch) to 100 µm(0.004 inch) in 12 µm (0.0005 inch) steps should be aimed for. It may be necessary to go to a specialist supplier for anything but the "normal" 78 tip.

So far the discussion has assumed that the grooves in the disc are modulated from side to side, ie. are laterally cut and that the pick-up used is a standard stereo cartridge. In this case the two channels of the cartridge are connected in series or, preferably, left as separate left and right outputs that are combined electrically in the pre-amp to give a monophonic output. There were, however, a few records that had the grooves modulated vertically - "hill-and-dale" cut; the Pathe label being one noted example. If these discs are played with the same set-up, there will be very little signal produced, but lots of noise.

d. Equalization - To achieve the optimum fidelity of re-production from the records it is necessary to "equalize" for the magnetic and mechanical groove characteristics. These early recordings were made with very different characteristics to those used for the standard microgroove LP RIAA equalization curve.

There is a considerable number of record labels, many of them employing their own equalization curves, current during the era of the 78rpm record. It can also be seen that playing commercial 78s through a pre-amp intended for LPs will give an excess of output in the bass frequencies and a lack of sparkle in the treble. While this configuration may sound quieter, it is incorrect. It will also not permit specialized noise reduction units to work at their best. These rely on the steep rise transients generated by scratches and cracks in the disc surface for their operation andany treble attenuation, whether caused by using the wrong equalization or by: using a tone control, will interfere with this process.

3.1.3 Summary

From a technical stand-point, Type IA re-recording of 78rpm discs is relatively easy. It is a reasonable way to begin teaching potential re-recording personnel that many factors affect the performance of this task. Re-recording modern discs follows the same basic procedure and can also be included in this repertoire.

3.2 Tape Transfer

3.2.1 Overview of Tape Transfer

Tape transfers are the second major activity in the field of audio preservation. Many archives, especially in the academic and cultural domains, produce or receive original tape recordings on formats not generally recommended for archival use, eg. analogue cassettes. In addition, early recordings made on acetate tape may be physically in danger. In all these cases a transfer to a more permanent medium is necessary. At the time of writing this article, the recommendation from all sources - the manufacturers of machines and signal carriers, archivists and the recording industry is that sounds to be stored for archival reasons should be recorded on standard play (50µm/1.5mil) matt backed polyester tape with low print-through characteristics. A brand that has proved its life expectancy and reliability and made by a reputable manufacturer is to be recommended. The choice of tape is not an area to make cost savings in.

Sound recordings have been made on magnetic tape for over fifty years. In this time many formats have been used. The main variations are in tape width, tape speed and in track layout on the tape. These are discussed later in this section. Whatever the format of the recording, there is one fundamental requirement of any replay machine:, it must not alter any of the physical dimensions or damage the tape in any way. For example, the transport must not stretch the tape, the guides must not crease or otherwise damage the edges and the heads must not remove oxide.

For many 0.25 inch/6.3mm recordings, a good quality, modern machine will be available with excellent tape handling characteristics. The amplifiers and heads will be of better quality than the originals. Problems may arise with some of the less common standards,' particularly the very slow speed recordings. With these, the original machines may have to be used with, if suitable ones can be obtained, modern heads and amplifiers.

3.2.2 Re-recording Tapes

Before playing any tape, it is recommended that it be carefully spooled back and forth two or three times before copying. This will help to relieve any packing tensions in the tape and reduce the effect of print-through. A visual inspection of any adhesive tape joints should be made and any splices that show signs of age should be renewed.

The basic system for tape transfers is as follows:

a. The tape to be copied is selected and rewound if necessary.
b. The tape track format and size are determined.
c. The designated play-back machine should have the correct head and speed configuration required by the tape in question.
d. The azimuth of the tape is checked and the replay heads adjusted if necessary.
e. The correct equalization is selected.
f. The signal from the play-back machine (the source) is then sent to the recording machine (tape). A tape/source switch on the recording machine or on the monitoring control desk will permit the replay output of either machine to be sent to the power amplifier and loudspeaker.
g. If the signal from the original tape is under or over-modulated, adjust the replay level controls. This will avoid the introduction of additional unwanted noise or distortion to the new copy.

It must be noted that all of the above procedure assumes the use of a properly calibrated, professional system and trained technical staff.

3.3 Cylinder Records

This section is intended to give some general introductory information on the subject only. Because of the extreme fragility of cylinders and the danger of damaging or destroying them when playing them, it is strongly recommended that no re-recording of cylinders is attempted without thorough training. Furthermore, if only a few cylinders are to be re-recorded, the authors strongly recommend that they are sent to an archive specializing in this type of work.

3.3.1 Machine Requirements

If players for 78 rpm recordings are hard to find, then machines to play cylinder records are extremely rare. In many cases they can only be found in antique shops and museums.

Unless a Type I re-recording illustrating the sound of the original system is being made, such machines are not desirable from an archivist's point of view. The relatively crude play-back mechanism can cause irreparable damage to the soft wax and, being an acoustic rather than electrical system, the use of a microphone to gather the sound from the horn will further complicate the re-recording process.

3.3.2 Storage of Cylinders

If the cylinders have not been stored in a vertical position, then it is likely that they will have deformed from the true, circular cross-section that they started with. They will have become oval or egg shaped. In fact, even if they have been stored in an upright position, it will not guarantee that they remain perfectly circular.

Re-recording of cylinders is a highly specialized field and, it must be repeated, should only be undertaken by fully trained technicians with the right equipment.

4. Test Equipment and Maintenance Routines

The quality of the equipment used to record and replay audiovisual carriers determines the quality of the recording or of the replay. The calibration of the equipment is, therefore, of vital importance to maintaining the quality of the recording. This is especially true for analogue recordings. Moreover, misaligned equipment may actually damage the carrier that is being preserved. The need for competent maintenance technicians, ideally on the permanent staff of the archive, cannot be overemphasised. If the technicians are supplied from another part of the institution housing the archive or by a contractor, the archive staff must insist on checking the work to ensure that the necessary high standards are being maintained.

4.1 Test Equipment

No technician can work efficiently without the correct tools. The minimum required includes all the necessary materials for frequent cleaning of the machines; demagnetizers to de-gauss heads, guides etc; the indispensable calibration tapes and discs for the alignment of the machines; and test equipment such as a low distortion sinewave generator, a harmonic distortion meter, a precision voltage meter that includes internationally standardized weighting filters and a wow and flutter meter to check speed stability. Instruments that produce written documentation are very helpful to the archive by making the keeping of technical status reports of each piece of equipment in the archive easier.

4.2 Frequency of Testing

In setting routine inspection and service intervals, it should be remembered that electronic equipment may fail unexpectedly at any time. Short inspection periods are, therefore, desirable. The following periods have been found by experience to be practical.

4.2.1 Analogue Tape Recorders

Cleaning Daily or, if an abrasive tape is used, more frequently
De-magnetizing Weekly
Short Scale With every new roll of tape
Frequency Test  
Complete Every 50 to 100 hours of operation. More and test frequently if
  Re-alignment and test the machine has been transported or loaned out.

4.2.2 Disc and Cylinder Players

Every 100 to 200 hours of operation. The styli and cantilever whenever the player has, or is suspected of having, been mistreated eg. the tone arm dropped on to a record.

4.2.3 All Other Electronic Equipment

Every 200 to 500 hours of operation.

5. Handling and Storage of Sound Recordings

Although this guide is emphasizing the basic technical equipment required by an audio-visual archive, a short note about the handling and storage of sound records cannot be omitted.

5.1 Physical Damage

In contrast to written information with its comparatively high level of redundancy, both in the formation of the individual letters and the overall text, each fraction of a second of an analogue recording is unique information. This means that the information on an A-V carrier requires, by its very nature, more care and attention in terms of physical and chemical integrity. While a speck of mould on a page of a book rarely prevents the reader from understanding and appreciating the message being conveyed, similar damage to an A-V carrier can cause severe signal degradation if not loss. In fact, the playing of a carrier with a physical defect may actually increase the damage.

Even digital recordings are not immune to physical defects on the carrier. Although a certain degree of auto-repair is possible because of the way the samples of sound are coded, it is not difficult to damage the carrier to such an extent that repair is beyond the capability of the error-correction system. The density of data packing on digital carriers is much greater than on the equivalent analogue recording and this also makes the digital carrier vulnerable to physical damage.

5.2 Storage Conditions

Handling and storage of sound carriers has to be optimised to maintain the physical and chemical integrity of the carrier and its full readability for as long as possible. Recomendations for storage conditions for sound carriers are similar to those for other A-V carriers and may be summarized as follows:

a. Replay and recording equipment has to be kept in conditions that minimize stress and other physical damage. Tape recorders must also be free from magnetic fields that may affect the tape path and, thus, the recorded signal.

b. Handling of carriers must be done in such a way that unnecessary contact with dust and other contaminates is avoided. Under no circumstances should data carrying areas be touched by the fingers. Operational areas should be kept clear of dust; the floors should not be carpeted and the air conditioning should be fitted with dust filters.

c. Special conditions apply to the handling of old and/or fragile recordings such as cylinders, acetate discs and acetate tapes. Personnel unfamiliar with the peculiarities of these types of carrier should be given training by more experienced staff before being allowed to handle them. At no time should tapes be edited or repaired with any material other than proper splicing tape supplied by a reputable manufacturer.

d. Climatic control of the storage vaults for all sound carriers should keep the temperature to 20EC +/-3EC and the relative humidity to 40% +/-5%. These variations are the long term variation. Short term variations those over a period of a week or less must be kept within closer tolerances. This will normally require the continuous operation of the air conditioning plant with, if possible, a back-up system. If a back-up system is not possible, the archive should invest in good thermal insulation for the vault. Caution should also prevail if the recordings are to travel outside the archive.

e. Magnetic tape is susceptible to damage by strong stray magnetic fields. For analogue recordings, a field strength of 5 Oe (400 A/m) in the case of AC fields and 25 Oe (2000 A/m) for DC fields has been determined experimentally as the maximum permissible. In a sound archive, the likeliest sources of such magnetic fields are dynamic microphones, headphones, loudspeakers and meter movements. Tapes should be kept at least 15cm away from such devices. It is wise to keep tapes at a greater distance from heavy power transformers, large electrical motors etc. The vault should also be as far away as possible from lightning conductor rods.

f. The storage position for all recordings other than certain soft, instantaneous disc recordings is generally in a vertical plane. Tape stored on flangeless hubs should be stored in their box on the central hub support; discs and tapes on flanged hubs should be stored in their boxes or sleeves in a vertical position. For all tapes and cassettes of all kinds, it is very important that the wind should be as even as possible to avoid blocking and leafing of the tape (blocking being the raising of a block of tape consisting of a number of turns of tape and leafing the raising of one or two turns of tape above the general level of the spool). This is best achieved by re-winding the entire length of tape in one steady pass.


6.1 The Common Formats

It is reasonable to assume that most audio archives could, with the appropriate equipment and some training of personnel, undertake the re-recording of commercially issued discs. The archive should also be able to re-record the common types of audio tape recordings. These would be examples of Type IA re-recording.

6.2 Less Common Formats

If the archive wishes to undertake Type II re-recordings or wishes to re-record less common originals such as cylinders, vertically discs, direct cut discs with aluminium, steel or glass supports for the lacquer surface, wire recordings etc. then the required investment in time, personnel, training, hardware and facilities is much greater. The average audio archive will not be equipped to perform this work. They may, however, have many of these recording formats in their collection. Rather than risk damaging the recordings by re-recording them in their own laboratories to make access copies, it is advisable for the archive to seek the services of an archive specializing in that type of re-recording.

6.3 Conclusion

This section has outlined the following points:

a. The role of the archivist is to faithfully preserve the content of original sound recordings.
b. Three types of re-recording, with one sub group, are legitimate for archival preservation; a fourth is not.

- Type 0 represents a replica of the original artefact
Type I is the replay of a historic carrier using a machine of the same era.
Type IA is the replay of a historic carrier using a modern machine and technology.

Type II represents the faithful reproduction of the original sound source (artist).

Type III, or subjective, re-recording does not fulfil the requirements for archival re-recording. It puts the aesthetic opinion of the technician between the original sound and the listener and is, therefore, not suitable for archival work.

c. It is essential to fully document the process of re-recording any historic carrier. The make, model and serial number of all machines and audio processing equipment used in the record/replay chain, details of the equipment settings used, the equalizations used, the size of stylus etc. must be noted to allow future researchers to, if necessary, restore the sound to as near its original form as possible. The performance curves of the various equipment should be kept and these should include the results of performance checks carried out by the archive as well as the literature produced by the manufacturers. As with other archive disciplines, the A-V archive technician should, as far as possible, employ restoration techniques and materials that can have their effects reversed. This requires full documentation of the re-recording process to make any required reversal possible.

d. Audio archive laboratories who specialize in audio research and restoration should be used as re-recording agencies by other archives requiring the less common formats to be re-recorded.

e. The attitude that the archivist takes to re-recording is the most important issue. Given historical accuracy as the goal, all of the technical considerations will become apparent with time and experience.

Contents - Previous - Next