Association of Research Libraries (ARL®)

A Dialogue between Anji Cornette, The Cutting Corporation, and Alan Lewis, National Archives and Records Administration

LEWIS: Good morning. I'm Alan Lewis from the National Archives and Records Administration (NARA) in Washington. However, my participation here today is not as an official spokesperson for the Archives. Although I'm officially identified as an "Audiovisual Preservation Expert," I am not in the agency's Preservation Division but rather am an AV person in our machine-based AV curatorial unit. I've become a contract manager for out-of-house audio, film, and video reproduction work. Since much of what is being discussed in this conference deals with preservation policy, my views are my own and based on my experience both at NARA and elsewhere in the AV field.

My co-presenter this morning is Anji Cornette, who will introduce herself in a minute. Her company has been one of my contract laboratories and thus I have had a five-year relationship with her company. Our joint presentation builds not only on that relationship but also on our independent experiences at other times and places, with other laboratories, and with other institutions.

Indeed, government contract administrators strive to an arms-length distance between themselves and their contractors. However, in the real world, once a contract is in place, both parties, if they are intelligent about it, can and should work toward a professional win-win relationship.

CORNETTE: Hi, I am Anji Cornette, Division Director of The Cutting Corporation. The Cutting Corporation has been in audio production for over thirty years and specifically working in sound preservation for over twenty-two years. I have been with the company for fourteen years. In addition to sound preservation, our company is very active in setting standards. I serve as co-chair of the R6 Mobile Electronics, WG11 Spoken Word Committee of the Consumer Electronics Association, which is currently working on standards for the download delivery of digital audio files.

LEWIS: In my experience, a collection manager may have need for two kinds of contract services: laboratory reproduction services and off-site storage services. In our all-too-brief time with you, Anji and I can cover only the first by:

• Defining the needs for laboratory services.

• Talking a bit about laboratory services themselves.

• Discussing conservation versus preservation versus restoration.

• Creating a statement of work by which to solicit and manage a contract.

Turning to the first topic, Defining the Need, I think there are three reasons to do duplication work:

• Failure of the medium. That's the stuff on the shelf that is dying because of its own inherent vice (as the archivists call it) or because of the damaging things we do to it or allow to happen to it.

• Technical obsolescence. That's the machinery needed to playback the media, the replacement parts to keep them going and the knowledgeable people to run and maintain them.

• Researcher need for access. That's all about not allowing your best preservation copy to be used by every researcher who comes through the door because every time a piece of recorded media is put on a machine there's the potential or the fact that damage will occur to it.

Concerning failure of media, during the course of the century plus of recorded sound history, many, many materials have been used as a base material, a binder, or as a surface coating on audio recordings. They include aluminum, Bakelite, cardboard, cellulose (cellulose acetate), celluloid (cellulose nitrate), ferrous metal, glass, paper, plaster of Paris, PVC, rubber ("Vulcanite"), shellac, styrene, wood products, other plastics, wax, etc. Some may have been ideal for the purpose to which they were put, initial recording or distribution for a relatively short time, but most have drawbacks that preservation laboratories must be geared up to deal with. Anji has prepared a series of images documenting some of the challenges her lab has seen recently.

CORNETTE: As a vendor with numerous years of experience in sound preservation, The Cutting Corporation has seen many institutions begin to notice deterioration of formats in their sound collections and therefore have come to us. The problems range from sticky shed syndrome to vinegar syndrome, to broken discs, to discs that have lacquer peeling off. There are discs in collections with issues such as palmitic acid or powder residue. Some reel-to-reel tapes have deposits of dirt and mold. We were even asked to consult with an institution in the Caribbean that had a collection suffering from biological infestation. More on that subject later.

The earliest commercial recording media—wax cylinders—is where we might start. The raw material that typically composed a wax cylinder was vegetable wax. For example, Edison solid wax cylinders were composed of ceresin wax, beeswax, and steric acid. Because of this, it is in the nature of wax cylinders to be fragile and breakable. They are also prone to mold and fungal growth especially in warm, moist, and dark environments. Mold damage is due to high relative humidity (RH) that is created by the individual cylinder packaged in hygroscopic cotton or wool wadding. Cylinders can also be attacked by fungus and the residue is fungal mycelium or animal bacteria that eats the wax. The fungus feeds on the surface of the cylinders and the audio program can be lost. Some waxes had oxides and oils that surfaced as white or blue haze on the cylinders and sometimes were mistaken for mold.

Moving to first generation or instantaneous discs, these so-called acetates were manufactured with an aluminum, glass, or cardboard base. The base was then coated with nitrocellulose lacquer plasticized with castor oil. This was an unstable mixture making these acetates not suitable for long-term storage. Symptoms or problems include continuous shrinking of the lacquer top coating, embrittlement, and irreversible loss of recorded sound because of the loss of the castor oil plasticizer. Since the core does not shrink and the lacquer coating does (or expands under changes in temperature) cracking and peeling of the lacquer coating results.

The production of palmitic acid is caused by the hydrolysis of the castor oil from heat and humidity, which then oozes through the lacquer on a disc. The specks or small mounds on the groove look similar to powder residue but have a more crystallized appearance. Palmitic acid is stubborn to remove and requires extensive cleaning. Powder residue may appear on lacquer discs as dried white specks or pasty mounds on the grooves. The main symptom is caused by glue from the paper label, which has spread over time onto the recording surface of the disc. Sometimes powder residue is mistaken for mold or palmitic acid.

Two years ago, The Cutting Corporation retrieved a collection of 1,300 recorded discs. The archivist thought the records were all in good to fair condition. In actuality, when we retrieved the collection, we found that two thirds of the discs were in poor condition suffering from both powder residue and palmitic acid. Each disc required extensive cleaning.

Before WWII, wire recordings were made with steel wire, which can rust but can easily be cured by wiping them down unless the rust is severe. On magnetic wire recordings, print through is a problem. But the worst problem for wire recordings is what we call the bird's nest syndrome or tangles and snarls that lead to the wire breaking.

Wire recordings existed in parallel to 1/4-inch audiotape and eventually were replaced by tape. Cellulose acetate reel-to-reel tapes and acetate discs are subject to a slow form of chemical deterioration known as vinegar syndrome. The main symptoms of this problem are a vinegar-like odor and buckling, shrinking, and embrittlement of the tape or cellulose disc. Low temperature storage conditions can aid in slowing down this process.

A well-known institution sent us two collections, one with a big surprise. When we opened the box, it was like sticking your head in a bag of salt and vinegar potato chips. The collection to our surprise was afflicted with vinegar syndrome. More on vinegar syndrome can be found in an article, "Vinegar Syndrome: An Experience with the Silent but Stinky Acetate Tape Killer" published at . There is a product by the Image Permanence Institute for measuring the level of acidity. The color on the strip, blue, will change to yellow as it detects high acidity, which is what a mustard yellow color on the strip represents.

Polyester magnetic tape stock that came into wide use in the 1960s can develop a condition known as sticky shed syndrome. This problem occurs when oxidation of the tape sticks to the guides and magnetic heads of the playback machine. The material builds up a residue on the guides and heads as playback continues. This causes distortion to the sound of the recording that is called separation or shed loss. It results in very low-level volume, fuzzy sound or inaudible audio. We have an institution that thought their reel-to-reels were in good condition. The reels did not display any sign of deterioration and had been kept in a climate-controlled environment. When we actually started to make the preservation transfers, we found to our surprise that the reels were affected by sticky-shed syndrome. In some cases, the sticky-shed was so intense that the reels required double or triple baking before a transfer could be made.

Here's example of what a sticky shed tape sounds like. (Played sample of tape with sticky shed syndrome.) Hydrolysis is the process by which moisture is absorbed by a material. With magnetic tape, it is caused by extreme humidity that results in the magnetic tape binder weakening and the binder with its oxide information carrier peeling or dropping off the tape. The results are dropouts, shedding or complete detachment from the base.

Mold is caused by the growth of fungus in elevated temperature or humidity conditions. It can cause serious distortion and physical breakdown in most audio formats, both grooved and magnetic formats alike. The other major agent involved in fungal action is the presence of organic material on the recording medium due to unclean storage areas. We received broadcast tapes from an institution that had been stored for years in a damp and moldy basement. The tapes had tiny specks of mold on them that had to be delicately removed in order to get a clean transfer.

On occasion, due to the environment in which the audio materials are stored, the materials will be affected by unusual conditions such as biological infestation. Due to hot and humid conditions and a basement flood where some reel-to-reel recordings were stored, termites decided that they would make their home in the reels. They ate through the reel boxes and resided on the tape under the plastic reels.

Belt recordings, another technology and one often used for office dictation, suffer from severe creases because it is not unusual to find them being stored flat in file folders along with copies of the letters or reports that were typed from them. Belts were often marked with crayons or wax pencils to denote the beginning and end of letters.

And then we come to the Memovox disc, a cellulose acetate grooved recording medium that is prone to the sombrero effect, in which the edges take on a scalloped shape. This is especially true if they were stored vertically in a box. In order to transfer the disc, it has to be flattened as much as possible.

LEWIS: Having made the case that there are many recording media used over these past 100-plus years and there are problems with them, we also recognize that the technologies have come and gone. The fact that so many have gone gives rise to reformatting in laboratories in order to rescue the content.

First there were the cylinders, the first of the physical forms, both in their one-off original copies and later mass-produced ones. Then came flat discs using mechanical stylus-in-groove vibration-capture technology, in one form or another, that lasted into the 1990s. Now the new disc formats use laser technology. In parallel with discs, ribbon-like media were developed in conjunction with magnetic recording systems. The first used paper tape with a magnetically sensitive coating and then metal bands like bandsaw blades. At about the same time, PVC flexible tape with the magnetic material suspended in the substrate itself as developed.

Recording on wire—another linear medium—was developed, more-or-less successfully. Later, taking a page from motion picture film technology, cellulose acetate-base tape replaced paper-base tape until it too gave way to polyester-base tape. So the question and the challenge to archives and reformatting laboratories becomes, "What do you do when you have media and no machines?"

CORNETTE: As a vendor, one of the exciting aspects of sound preservation is dealing with obsolete recording technology. Playback machines are not necessarily available and each medium has its own challenges. The Cutting Corporation has had to either find and fix an obsolete machine or reverse engineer an obsolete machine usually cannibalizing parts of other machines, to recreate an older technology. We have conducted research to locate existing machines and if those results do not yield anything, then we rebuild or reverse engineer.

Over the past few years, we have had to rebuild a Memovox machine and we had to reverse engineer a Dictalog Magnabelt machine. Fortunately, we had extra parts from other disc and belts technologies so we were able to cannibalize the parts in order to build the replacement machines. There's a lot of trial and error before we get the machines working correctly, and it takes hours and hours of engineering time, patience, and ingenuity. There are many intricate parts that have to work together to get the speed right. For our Dictalog Magnabelt machine we had to deal with a belt of a different length and width than Dictabelts or Magnabelts. Our engineers succeeded by building a machine with a lengthened and wider belt path and a longer lead screw to move the playback head across the wider belt.

Also in the realm of machinery, custom-built devices may be needed to solve the cleaning and handling of items before preservation. This is The Cutting Corporation's proprietary Open Reel Vacuum Cleaning System to clean the front and back surfaces of a tape. Such machines must clean delicately but effectively without hurting the media. To do it, the lab's technical engineers worked closely with the sound preservation engineers to modify a reel-to-reel machine into an audiotape cleaner. The two engineers had lengthy discussions and eventually took the heads off the machine, added 1/2-inch Tape Wipe and a low-pressure vacuum cleaner. The reels are cleaned of dust and dirt and the sound transfer engineers are now able to get clean playback of dirty tapes in order to create a new preservation masters.

LEWIS: The final need for reformatting—or perhaps in this case just duplication—is serving the needs of researchers. Having material on the shelf in good storage conditions is only half the archival task. Access is the other.

It is not sound (pardon the pun) archival practice to allow researchers to use the sole, best quality and likely irreplaceable archival copy of some unique record. Hence, a surrogate copy, a service copy, a reference copy—whatever you might call it in your shop—needs to be made from the original. To my mind, it should reflect the sound of the original recording with only minimal signal processing, if any, done to it so that a researcher hears the program content more or less as the recording exists today or perhaps just as a listener would have heard it at the time the recording was new. Whether you allow minimal signal processing or prefer your access copies to be wholly unprocessed, a flat transfer as some call it, is a matter for your archive to decide.

Also in practical terms, other users may have more than academic needs. They are the repurposers, the people who make new product out of old. They're the ones who should be paying the extra costs of making a Caruso cylinder or disc sound like it was recorded yesterday. (I won't get into the ethics of this whole matter of improving historic recordings.)

CORNETTE: Once preservation masters have been made, institutions often order service copies or reference copies for researcher access. In the past, many institutions asked for analog reel-to-reel tapes or cassettes but over the last few years, there has been a gradual shift to digital formats such as CD-Rs for reference copies. Other formats requested have been digital formats such as MP3 files for download or Real Audio Streaming files. One institution was resistant to having digital service copies because their parent institution was sticking with analog media. After two years of resistance, the institution decided to have us make CD-R reference copies in addition to the analog reference copies and found that the CD-R reference copies sounded very good and were convenient to use. Ever since, they have ordered CD-R reference or service copies instead of analog tape

LEWIS: Having defined the need for lab work, please note that a media preservation laboratory is not a run-of-the-mill audio production house, the one down the street or around the corner run by a rock-and-roller whose hearing is something less than wonderful. It must be an appropriate physical plant, equipped with the right equipment and staffed by the right people who are sensitive to and experienced with old media and who are clear about three services they might provide you. Those services are conservation, preservation, and restoration.

I think of conservation of an original item as providing a number of services. Among them are doing a physical inspection and preparing a condition report, doing proper winding of linear media like tapes, making repairs, and/or cleaning the item, and rehousing the item in preparation for long-term storage.

To my way of thinking, preservation deals with saving the recorded content of the original item: inspection of the original first, cleaning and providing heroic measures if the item is deteriorated, and duplication of it to a current format or technology, and perhaps making duplicating masters and access copies at the same time. If the original item is so physically troubled or its technology is so exotic, this is the time that a Replacement Preservation Copy would be made.

Finally, restoration might mean trying to ferret out what the original recording was supposed to sound like at the time period in which it was made and replicating that. Or maybe, it is making an old recording sound like it was made yesterday. The bottom line is that you and the laboratory need to decide how much restoration is restoration if you are going to do any restoration at all.

To recap a bit, Anji and I have talked about laboratory services, perhaps the major AV archives service that is contracted out. Those services are driven by:

• Failing media—and triage the collection first in order to do the most important failing recordings first.

• Obsolescence of the technology—and triage the collection to do the most important obsolete recordings first.

• Reproduction demands for research or repurposing.

The second contract service I would have talked about is off-site storage services—but because of time limitations and the fact that I covered some of this somewhat obliquely in my session yesterday, we'll save this for the next Sound Savings Symposium.

Now on to the Statement of Work or the SOW as we sometimes call it. It is important for you, the client, to have in mind what the project is all about. This is the point where the vague yearnings that "something needs to be done" must move to the next step: defining what it is that has to be done. The goal statement should be written, in my estimation, in order to help you and prospective vendors focus on it.

CORNETTE: As vendors, it is helpful when the archivist knows the history of the collection. What is the subject matter? When was the collection originally recorded? Under what conditions were the recordings stored before the archivist received them? It was helpful to us that the Smithsonian knew that J.P. Harrington's recordings slowed down because the battery for his recording machine was dying. A U.S. Department of State employee brought back a gift that a Russian friend had given him of a record that was an actual x-ray with embossed grooves on the x-ray. He knew that the record was made during the Stalin reign but did not know what was on the record. It ended up being a recording of Elvis Presley singing a song from one of his upcoming movies. The x-ray record played beautifully on our turntable with a 2.7 mil stylus.

Some recordings we have transferred are rich in American history. For example, we have listened to recordings Jane Fonda made during the Vietnam War. We listened to Sergeant Tom as he defected during the Vietnam War and all the propaganda he created. We have recovered a daughter's memory of her deceased mother's voice. We have listened to recordings of Nobel Prize physicists. We have listened to Duke Ellington and Ella Fitzgerald in jam sessions and rehearsals with their friends and colleagues. We have gotten to know Brownie Wise, the Tupperware Lady, and reminisce with Hills Brother Coffee commercials and Eskimo Pie commercials.

LEWIS: After having gone through the process of deciding what needs to be done, this is the step in which you really get down to nuts and bolts of defining just what's on the shelf that has to be dealt with. A clear inventory is needed for the vendor to understand the amount of work to be done and to bid the job properly.

CORNETTE: As a vendor, it helps when the client or institution knows as much information about their sound collections as possible. In the past, we have had institutions call up at the time that they are ready to go after a grant to find that they really don't know what's in the collection. Often this makes it extremely hard for us to provide a ballpark cost estimate. Collections can have reels with different speeds, more than one track, different thickness, and deterioration issues such as sticky shed syndrome and vinegar syndrome. Some collections may have been badly packed or tight-wound and therefore may have developed issues. Others may have mold and dirt and require extensive cleaning.

The same goes for discs. There are acetate discs and vinyl discs. The acetate discs can be based on glass, cardboard, or aluminum. They also a have a host of issues such as chipping, flaking, powder residue, and palmitic acid. Records also play at varying speeds, although most acetates are at 78 rpm. It helps when an archivist knows these types of information at the time of the bid process or cost proposal.

LEWIS: In this step, the end product of the project is defined. This is the step of getting beyond the stage of vague yearning to the substance of what needs to be done.

CORNETTE: A vendor needs to know what medium(s) the client wants to preserve to. Are they open to digital preservation or do they want to stick with traditional analog preservation and analog formats such as reel-to-reel tape? If they are open to digital preservation, are they willing to experiment with high digital compression formats? Do they have digital storage solutions at their institution? Would they like service or researcher copies for access purposes? Would they like a 44.1 kHz, 16 bit wav file on CD-R, or some other digital format like MP3? Do they have a database they would like the vendor to work with or would they like the vendor to provide a database? What information would they like on the labels of the preservation masters and the service/researcher copies? In some cases, institutions like the vendor to consult with them to provide suggestions for their end solutions.

The laboratory also needs to make sure the customer understands the difference between conservation, preservation, and restoration. There are institutions that have very small budgets and therefore have to conserve their collections by re-housing or tight-winding their materials. Customers can confuse preservation with restoration. In fact, recently a woman had an album of Yale University's exclusive musical group, The Whiffenpoofs, which she wanted us to preserve for her husband. The engineer preserved the album and then made a reference copy on CD with the tracks broken up by the bands in the record. We found out later that what the client actually wanted us to do was some remastering and break the bands into individual tracks of songs. She did not understand the difference between preservation and remastering. We have also had a customer ask for preservation and when they listened to the completed projected, they called to complain because what they really wanted was restoration or cleaned up sound.

LEWIS: In thinking through the project, this is where reasonable, realistic, and achievable landmarks are set to measure the productivity of the project.

CORNETTE: The laboratory should work with the vendor to decide start and due dates. There are also issues of picking up originals and delivering them. Does the client have requirements for picking up originals and returning them? How many pieces of the collection can go over to the vendor's facility at one time? Will it come in batches or in an entire collection? Does the client have specifications for storage and handling while originals are in the laboratory's facilities? The laboratory should follow strict measures for storage and handling of originals while in the vendor's facilities.

LEWIS: If you are a nonprofit or a government agency, you may be able to purchase raw stock for the new products or packaging materials at a lesser price. Will the reproduction vendor allow it?

CORNETTE: The laboratory should coordinate with the client on any outstanding items such as labels. The laboratory also needs to arrange for shipping. Once the lab knows what the approved materials are for the project, the laboratory should arrange with suppliers to get items in bulk for discount prices, which then can be passed on to the customer.

LEWIS: As a purchaser of reproduction services what are you seeking in a vendor for your precious, one-of-a-kind, archival original materials? If you think I'm prejudicing you against the audio recording studio down the block—you may be right! Among the things I'd like to know are the company's length of time in business, its expertise in the specific tasks that will be involved in my project, the qualifications of staff who will have their hands on my materials, lists of equipment that will be used and how—and how often—the equipment will be cleaned, aligned, or otherwise maintained.

CORNETTE: The laboratory should have extensive experience in the area of sound preservation and sound restoration from working on a variety of sound collections for various institutions. The laboratory should have standard audio recording equipment as well as obsolete playback machines. The laboratory should have an impressive selection of styli and a variety of reel-to-reel record and playback machines. The equipment should be cleaned, aligned, and maintained on a regular basis. This requires having proper testing equipment at the laboratory's facilities and well-trained personnel including an outgoing quality control department. Engineers should have a degree or certificate in sound engineering and experience in some aspect of sound preservation both in the analog and digital realm. Some of vendor's employees should have security clearance and the backgrounds of any engineers should be checked, as some of the material could be sensitive or classified.

LEWIS: Here's an interesting dilemma. You've got an original recording that you may not be able to playback because you don't have the equipment or it isn't quite gentle enough for that archival original, so how will you know the vendor has done the best job possible when you can't listen to them side-by-side?

CORNETTE: Quality assurance is conducted by the vendor before the preserved material goes to the client who often has a quality assurance program of its own set in place. The laboratory should adhere to a strict quality assurance program. The engineer should A/B the original to what he/she is preserving as a reference point before completing the entire batch. Once a batch is completed, it goes to the quality assurance department for a quality review. Additionally, before a project is sent on to the client, the engineering staff or manager should conduct an outgoing quality inspection. On occasion, the client and vendor may agree on something that might be redone differently, not necessarily because it was done wrong, but to capture the program in a different way. This is called rework and is conducted immediately to complete a batch.

LEWIS: As to labeling, how will the newly made recording medium be labeled as well as its container? With CDs and DVDs, what is safe? As to packaging, what's safe and protective? Are there institutional or professional standards? Color coding? Corporate logos? May the laboratory's name and address appear on the label?

CORNETTE: The vendor goes over with the client how the labels for the preservation masters and the service copies are to be done. An institution logo can be added on. Questions on color of label, any standard numbering for items, and any other information that needs to be on the label are answered and clarified before the vendor generates labels. The vendor should discuss with the client whether to use direct imprint or paper labels on the CDs. Reels should always be returned to the customer in a tight-wound position. The preservation masters should be housed in archivally approved containers to be housed at the client's facility. Decisions on what size boxes to use should be made, either 7 inch or 10 inch or a combination of both.

LEWIS: Documentation of archival treatment is a standard procedure in the conservation field. We should do no less in providing a paper trail of what was done and why, by whom, and what standards or equipment settings were used.

CORNETTE: Often the client wants a report from the engineer on the condition of a recording. For example, was the condition of the recording in poor, fair, good or excellent condition? If in poor or fair condition, how was the condition treated by the engineer in order to produce a preservation master? What were the technical processes used to make the preservation master such as baking reels to cure sticky shed syndrome? What was the original date of the recording and what was the date of the re-recording?

The laboratory should work with the institution based on their current inventory control system, on how to keep a database and generate labels with the information required for their access purposes. The laboratory could create a database in MS Access for the institution's future use and then also work with the institution to develop the proper metadata for digital files for future access.

LEWIS: The goal of shipping is to insure the safety of the original materials as well as the copies made from them. NARA uses overnight shipping Monday through Thursday or in-person pick-up and delivery services.

CORNETTE: The laboratory should work with the client to determine which is the best way for their sound collection to travel. The laboratory can either hand carry or ship overnight the original masters of the sound collection. The original masters should always be shipped separately from the preservation masters and service copies. The shipments should be on separate days and the deliveries at separate times. The laboratory should carefully pack the sound collections or advise institutions how to pack the sound collections so that they are free from shock or vibration. The laboratory should also be careful with magnetic items, so that they are not accidentally erased in travel. The laboratory should provide temperature controlled jolt-free transportation. Transportation is a large issue with sound preservation because many items are irreplaceable. Just as each collection is unique, so is the transportation to and from the vendor.

LEWIS: A vendor's laboratory may not have the same high security as (I hope) you have the collection in. What compromises are you willing to make and how will you inspect and monitor the facility to insure that the vendor is providing adequate physical and intellectual protection for original materials while in its custody?

CORNETTE: The laboratory should be as secure as possible with several locks and alarms. The alarm systems must detect for fire and for intruders. The building should always be locked at night and if the laboratory has some kind of guard or attendant during business hours for the building, that's always a plus. The laboratory might consider keypad locks that can block any access or reproduction to unauthorized people. Part of the hiring process should be conducting a background check of the laboratory's employees. A limited number of people should have access to the key codes.

Fire extinguishers (Class C or one suitable for electrical fires) must be in the laboratory ready to be used if ever necessary. The laboratory must ensure standard storage conditions for the sound collections while in the laboratory's facilities. A properly set up lab should be in a building that is well-constructed, well-located, and free of environmental hazards. The proper temperature and humidity to be maintained is at a consistent temperature of 60° F to 70° F with humidity values of 45% to 65%. The system should operate 24/7. Extreme changes in temperature and humidity greatly increase physical deterioration and can result in chemical changes and fungal growth in the materials of which the medium is composed. The laboratory should have a chart recorder to take daily measurements of both temperature and humidity.

The sound preservation laboratory should be located in an area free from where harmful vapors might be absorbed. The laboratory also should be free of food and the worktable should be clean of any foreign substances. The laboratory must be in a flood free environment.

The materials from the sound collection should be placed in their proper upright positions on strong shelves when stored in the laboratory. There should also be a fireproof cabinet in the sound preservation laboratory for valuable irreplaceable sound collections housed at the vendor's facility until preservation work is complete. It is good to have a fireproof safe.

LEWIS: Periodic reporting may not be required for short term or small projects but for large or long-term ones, you will need them to monitor progress and finances.

CORNETTE: What kinds of reports does the client want to see? How often? The client will probably require the vendor to send a temperature and humidity report weekly. Is there a receipt of materials report that the vendor needs to cross check when the materials are picked up from the client? Does the client want a quality control check report for every item in the batch or collection?

Modern communication with the client and vendor has made many preservation jobs go smoother. It is important to communicate by e-mail in order to have proper documentation of any changes to the SOW and project. E-mail has allowed the client and vendor to keep in touch during the course of a project. It is a quick and inexpensive way to communicate and it does not require a lot of one's time.

LEWIS: What will be the frequency of invoicing: Weekly? Monthly? As batches are completed? Some funds up front if equipment fabrication is needed? Some other scheme? Will there be penalties for late work? Will there be incentive payments for accelerated performance?

CORNETTE: From a practical, business point of view, it is better to bill in batches of a collection the laboratory is preserving, especially if the collection is large. Often times, this will coincide with a business calendar month for internal P & L reasons. If the laboratory waits until the entire collection is complete, it might be waiting several months or years. That's a long time to go without any revenue. The vendor should work with the client to decide what a reasonable amount of items to be preserved in a batch would be and then bill monthly no matter many batches were completed in the month. Clients often wait to pay the vendor when they have acknowledged receipt of the batch and new preservation masters or have completed quality control of the new preservation masters. If there is rework to complete, rework is completed before payment of the services is issued. This kind of invoicing allows the client and the vendor to keep track of money allotted to the contract and how many items are being done.

LEWIS: To wrap this up, we've talked about:

• Recognizing the three needs for lab services: failing recording media, obsolete technologies, and user demands.

• Some convenient operation definitions of conservation, preservation, and restoration.

• Some laboratory reproduction services.

• Creating a Statement of Work by which to solicit and manage a contract.

© 2004 Anji Cornette