Interview with René Paquet (LAC)

Library and Archives Canada, Gatineau, March 1, 2010


René Paquet has worked at the Library and Archives Canada, situated in Gatineau, Quebec, for more than twenty-five years. For much of this time he was responsible for the electronic archives, and he has recently turned to managing the sound and video archives. Emanuel Lorrain of PACKED met him in order to learn how the material is backed-up, and how the obsolescence of the formats and media on which they are stored is addressed.


Part One: Archiving and Digitisation.


PACKED: Could you tell us about your career and your training, and how you came to managing the archives of the Library and Archives Canada?

René Paquet: First of all, I obtained a diploma in electronics, and then I graduated in Information Technology at the University of Quebec in Hull. Then I worked for Radio Canada Company and at the National Archives of Canada, which later on became the Library and Archives Canada.


PACKED: What type of archives can be found here?

René Paquet: Mainly Canadian archives: satellite imagery, maps, manuscripts, photographs, books, paintings, videos, films, audio documents, but also electronic documents such as websites or databases. During my career, I have worked mainly in the field of video archiving, and it is only recently that I started specialising archiving electronic documents.


PACKED: Are all the archives stored in this building?

René Paquet: No, just part of them. We have several locations and the archives previously occupied as many as seventeen buildings in Ottawa, Gatineau and the region. This new edifice was built twelve years ago to meet particular specifications: the laboratories and the storage of documents have been brought together. The building has five floors and the top floor houses the laboratories, whereas the other floors are storehouses. The building is made of concrete and glass, so that in case of a power failure, the concrete keeps the temperature constant, and the glass windows all around the building are there to control the humidity coming from the outside. It is a sort of building within a building. It was purposely built for optimal preservation of the archives and it is guaranteed by the architect to last for five hundred years. The materials used for its construction are metal plating, cement and glass. Inside, the walls are made of melamine to avoid the accumulation of dust, and the carpets are antistatic. It is what we could call an intelligent building in that if, for example, the temperature in a storeroom varies beyond the parameters required for the conservation of the documents, and then an alarm goes off so that we can react in consequence.


"It is a sort of building within a building."


PACKED: Where do the video archives that you preserve come from?

René Paquet: The AV (for audio-visual) section of the Archives of Canada preserves film, audio documents and video. These can be television archives, as well as archives of private origin. We archive 3 to 4% of Canadian television productions, and this corresponds to what the Archives of Canada consider to be worth archiving. These 3 to 4% are selected by archivists according to pre-established criteria. Information is gathered on each document acquisition, then the document is sent here for preservation. In the 1980s, we were also the archives of Radio Canada and CBC, but in the 1990s they set-up their own archives and they are no longer stored here, except for certain programmes. The private sector involves certain French and English speaking channels: Radio Canada and TVA. Among the private English-speaking channels, there is Global CTV, also for about 3 to 4% of productions. This is equivalent to forty hours of recording per week and mainly involves public affairs, news or special events. The television channels, the provinces and the municipalities keep what we don’t preserve as they also have their own archives. About two years ago, a long-term project to gather together all the archives was launched, which will consist in regrouping the National Archives of Canada, the archives of the provinces as well as the municipal archives.

We also preserve millions of metres of film, for which there is no migration plan as of yet. Any film funded by Canada must have a copy stored here. For the moment we just make a preservation copy and do transfers from film to film, to 35mm1 for example. These consist of negatives, interpositives2, internegatives3 as well as any offcuts, which are also preserved. Regarding formats, there are 8 mm4, 16 mm5, 28 mm6 and 35 mm films. More operators work in film cleaning and transferring than in video preservation, but this is also because we have many more archives on film than on video. I would say that in total, we have nearly 520,000 hours of recordings on all types of formats and media.


PACKED: On what media do you receive the documents?

René Paquet: It can be material picked up directly from satellite in the case of agreements with certain television channels to directly file certain signals that they emit. Then, we may receive all types of videotape: 2” Quadruplex7, 1”8, etc. We also have Kinescope film9 and even sound recordings on steel wire made by wire recorders, which were mainly used by the religious communities.


The main transfer room.


PACKED: It says on your website that you preserve open reel ½” tapes, which is a non-professional format. What type of documents are these?

René Paquet: These tapes come from ministries, government agencies, universities, and also from research institutes. An example on ½” open reel tape10 of the very diverse nature of the material we store here is a video capture of patients being exposed to x-rays.


PACKED: What are the criteria for transfers?

René Paquet: The selection is made by a combination of several criteria: the sampling and research done on the physical condition of the tapes in the collections, the degree of technical obsolescence of the formats and the importance of the material. We are right at the start of our migration programme, which was shelved for several years for both budgetary and political reasons. We started by transferring D211 tapes just over a year ago, amounting to between 3,000 to 5,000 tapes. We chose to start with D2, because this allowed us to set up our IT infrastructure more easily, in terms of the SDI12 communications and operations. To set up the infrastructure, we needed a format that is relatively recent and more or less available. Even though the D2 format is already obsolete, it was relatively simple to keep the machines running.


PACKED: In which climatic conditions are the tapes stored?

René Paquet: The videos are stored at a temperature of 18°C and at a relative humidity of 40%.


PACKED: Do you use machines to clean the tapes?

René Paquet: We use RTI13 machines for the 2”, 1”, ¾”14 and Betacam15 tapes. We have other machines for the ½” open reel tapes.


An Elcon 2000P cleaning machine for 2" tapes.


PACKED: Is cleaning the tapes sufficient when they are problematic?

René Paquet: Not always. During a survey we did of ¾” formats, we detected a collection that included some very damaged tapes from the 1970s. This was very probably linked to poor conditions of storage with uncontrolled humidity and temperature that ended up changing the chemical composition of the tape. The tape oxide peels off and the video heads become dirty very quickly. For these tapes, we will have to set up the means to edit them little by little, using Virtual VTR16.


PACKED: Do you also have problems with sticky shed syndrome17?

René Paquet: Mainly with audiotapes. The only videotapes that are concerned by this are ½” open reel.


PACKED: What treatment have you used on the ½” tapes?

René Paquet: We heat them, but in any case a lot of iron oxide accumulates on the video heads, so we still have to transfer them bit-by-bit before reassembling them. The oven is the last resort, and in that case we use a protocol that was introduced by the Canadian Conservation Institute18, which is a state agency. In any case, it is very unusual for us to be able to transfer a whole tape in one go.


PACKED: Is it a problem of hydrolysis?

René Paquet: Yes, but it is also a problem of equipment because each tape was probably recorded using equipment that was different and not necessarily compatible on the level of interoperability. In those days, each machine had different mechanics and this is a problem as video recordings were first of all made mechanically. Many adjustments have to be made to transfer this type of tape and it is very tedious.


PACKED: What other types of problems do you encounter with the tapes?

René Paquet: We have had problems with glue on the 2” tapes. At one time, 3M used glue on the side of its 2” tape holders, and as the years go by this glue spreads onto the tape itself. We had to come up with procedures to clean and remove this glue so we could then copy them. The most effective technique finally turned out being a manual one, as it was impossible to use a machine to do this task. We used 99% alcohol and removed the glue very gently with a cotton-bud.

We have also had tapes that have suffered damage by flooding, and had to be extensively cleaned before it was possible to copy them to new tapes. Or even broken reels, where we had to replace the tape holder. I worked with Fuji and 3M, and one of the solutions 3M used in their laboratory was to add an acid that worked as a lubricant for the tape. However, the results were not always better and sometimes this made things worse as the video head would no longer make proper contact with the tape during playback.


PACKED: So you worked on preservation with Fuji and 3M?

René Paquet: Yes, mainly with 3M in the 1980s and 1990s. We came to an agreement because we bought a lot of 3M products, and some of their 1” type C videotapes gave us serious trouble. The archives did not directly fund the collaboration with their laboratory; it was an agreement between 3M and us.


PACKED: How do you spot the tapes that are the most at risk?

René Paquet: One of our approaches is to do surveys. This is something we have mainly done for the digital storage tapes, but it can also be used for other types of magnetic tapes. Out of ten magnetic tapes with the same identification code we pick one at random, and we make note of which tapes cannot be read and which tapes have defects but remain recoverable. In the end, we pick one out of 50, then one out of 100. Then, thanks to the identification code, we can assess all the tapes in a given batch.


PACKED: Is there a depot to quarantine mouldy tapes?

René Paquet: Yes, there is a depot where the tapes are handled by specially trained employees wearing masks.


PACKED: Which archiving format do you use for master tapes?

René Paquet: We still have a lot of material on Digital Betacam19, but we are in the middle of transferring everything to a totally file-based system, where the archives are fully stored on computer files. We use the MXF container format and Motion JPEG2000, a lossless compression standard allowing a compression ratio of about 3:1. Having not yet finalised the IT infrastructure, we have not yet completely abandoned transferring to Digital Betacam, and we create archives in parallel on file and on videotape until the system is truly completed. The 1.2 petabyte library of LTO20 tapes is already in place, but the procedures for archiving are still being developed.


PACKED: How does the digitisation process take place?

René Paquet: In the storehouses, we keep the video documents on Digital Betacam, on ¾” tapes, Quadruplex, D2, etc. There is also the satellite reception of television channels, which is encoded directly to MJPEG 200021 in the MXF22 container format. During digitisation, two high-resolution copies in MJPEG 2000 at 50 megabit/second are created. One of these goes to the LTO tape library and will be accessible thanks to the LTO robot/player, and the other copy goes to the physical storehouse at stable temperature and humidity. We use an SDI infrastructure, which will later be replaced by fiber optics. All signals from 2” machines, from AVR-1, AVR-2, helical, etc. are converted to SDI. The machines are directly modified to this effect and the signal is directed to the video switcher. For the time being, the video bit rate is 270 megabit/second and we use LTO-4. Our encoders use the NTSC standard and record at a bit depth of 10 rather than 8, because this does not take up much more space and the next generation of encoders will probably all be 10 bits. This allows us to avoid changing from 8 bits to 10 bits later on, and having a potential loss of quality during this process.


LTO tapes servers.


PACKED: Are there mirrors of the high-resolution files?

René Paquet: Yes, with a different identification code but identical to the original file. Before the MXF file is exported, a checksum is done to ensure that the two copies are identical. The checksum is created from the source, and is part of the metadata stored in the XML file. It is also saved to the MISACS database, and always stays with the file, which allows us to check if anything has changed at any time.


PACKED: What is the MISACS database?

René Paquet: MISACS is a database for preservation developed by the archives of Canada, which incorporates all metadata regarding the collections: films, videos and audio recordings.


PACKED: Where are the mirrors geographically situated?

René Paquet: It is currently situated in the storehouses of the building, which has walls that are about sixty centimetres thick. The building itself was made to resist a nuclear strike, except for the fifth floor. Previously, we had a warehouse 90 kilometres from here, but the storage conditions were poor and the decision was taken to bring everything here. During surveys of the tapes stored there, we had ten times the amount of bit-rate errors than we have for the tapes stored here. Therefore, it was safer to have them here in the same building than in a distant storehouse where the risk of data loss was higher. When the other site has been brought up to standard, the mirrors will be taken back there.


PACKED: Are access copies also made?

René Paquet: Yes, when the master copy is made, a consultation copy is created in H.26423 format and sent to the hard drives to be stored. For this, we called upon a company called Samma, which uses multi-encoders that can encode in several formats such as MJPEG2000, MPEG-2, or even MOV, etc.


PACKED: The hardware and software parts of the system were provided by Samma Systems24?

René Paquet: Yes, it’s a system that is also used by the Library of Congress in Washington. They even have robotic systems, which we should also have soon for our ¾” tapes.


PACKED: Do the files incorporate metatagging?

René Paquet: Yes, and in the future, we will be able to index image to text, so by doing a text-search, we will be able to position the video at the corresponding image. This could also be done by indexing the subtitles for the hearing-impaired. This is still under development, but it should make searches much more easy.


PACKED: Are the video archives also available online?

René Paquet: Yes, but this is not yet relevant, as we have yet to establish the protocols and formats that will be used for the website.


PACKED: How is quality control implemented?

René Paquet: Quality control is done automatically: we are currently developing a list of 48 audio and video parameters that are observed during the analysis of a signal. When the system detects anomalies, they are reported.


PACKED: Are these parameters that used to be checked visually by operators using monitors, oscilloscopes and vectorscopes, etc.?

René Paquet: Yes, and it is much more precise than having an operator who must know, for example, how many drop-outs there are and be able to tell where they are situated. In the current system, quality control checks black level, dropouts, video input, if the RF is too low, sound, the servo-system, etc.


PACKED: Are there other stages of the digitisation of the tapes?

René Paquet: Yes, for certain broadcasts or programmes that last longer than a single tape, and are recorded on two or three tapes just as a film is stored on several reels. In this case, we assemble them once they are digitised, and we create a large file, which is then sent along the chain to be stored on LTO.


PACKED: Is the assembly also automated?

René Paquet: No, we encode the tapes, then we send this to a dedicated temporary storage place, or staging, where an operator will come and do the reassembly before putting it back into the chain. When he sends it, a checksum is done.


PACKED: Is the entire procedure written into the XML file that accompanies the video file?

René Paquet: Yes, and these XML files are very large. It contains a lot of information, including the container format type, the IDC, the file number, the format, but also the checksum, quality control, the transfer history, etc. The XML file is named after the identification code of the digitised programme. The metadata it contains can be very technical, and are mainly used during production, postproduction, etc.


The XML file of a digitised tape.


PACKED: With hindsight, were the D2 tapes a good choice to set up and start the digitisation process?

René Paquet: Yes. What’s more, we were able to make an approximate estimation of the time needed to make a preservation file: of this, 16% is devoted to video processing, but this varies depending on the type of computer that is used, and I took the Samma System we use as a reference; quality control, which a lot of time is devoted to, amounts to 31%; automated monitoring makes up 15% of the total whereas the rest includes the transfer of the data from one server to another, which requires a large number of checksum comparisons. Then comes the tape storage process, that is to say the storage on tape and uploading to the database. Whether it is a D2, U-matic or Quadruplex tape, the structure of the folders is always the same, and the MXF, h.264 and XML files will always be created and archived.


PACKED: How was the MJPEG 2000 format chosen for preservation? Did you look at the formats that other archives have chosen?

René Paquet: We did look at what other institutions have chosen, but we also considered what was best for our archives in the long term and for their storage. The choice of MJPEG 2000 does however have its drawbacks: at present, it is not very popular and when we want to do editing, for example, the choice of available tools is very limited. However, the situation will develop shortly. Before MJPEG 2000, we also considered using MPEG-2, 4:2:2, but this entailed a loss of chrominance. So between a lossy format and a lossless format, we chose the lossless one. Lossless compression is a compromise, and if we could we would have used uncompressed files. But compression is an economical choice: all these hours of recordings require a lot of LTO, a lot of hard drives, of computer systems and high-performance servers in order for us to be able to store them and track them.


PACKED: But to begin with, you started with MPEG-2?

René Paquet: Yes. For video archiving, we used MPEG-2 for about twenty years, decoded and put on Digital Betacam.


PACKED: Did you choose MJPEG 2000 when you switched to LTO?

René Paquet: Yes, although LTO had been used before this, when I worked in the IT department and I used LTO to store all the geomatic data and databases of the federal government. Before LTO there was SDLT25, and before that DLT26, 8 mm backup format27, 4 mm28, and 9 track reel-to-reel tapes29. Incidentally, we still have several thousand of these tapes to transfer. Information technology has evolved much faster on the level of data migration than on the video level, partly because information migration is more critical. If LTO was introduced, it was because it was the most economical, the most stable and the most reliable format, and because it is an open format, or almost.


PACKED: On that subject, what are relationships like between the video and IT departments?

René Paquet: There may be a certain animosity between the two and one of my projects was to connect them in order to bring them closer together. The problem is more a question of culture and of training. On a governmental level, computing is only considered in terms of database systems and office automation. And yet a system adapted to office automation will not be adapted to processing video.

IT likes to control everything, but lacks expertise in the field of video files. It is a relationship that works relatively well however, but there are still conflicts and friction at the level of governance. These are not technical problems; they are problems with mentality, which will probably be solved over the coming years.


Part Two: Managing Equipment and Obsolescence.


PACKED: How much equipment do you own?

René Paquet: We own 1,150 items of equipment used only for preservation. I have divided this equipment into three categories: the mechanical category, mainly including equipment for film, which is made up of 150 serviceable items; the electronic category, consisting essentially of video equipment of the 1960s to the 1990s for a total of 610 items; the last category includes the more recent equipment, the hardware and the software, for a total of 390 items, but this number is constantly increasing.


PACKED: What types of video equipment do you own?

René Paquet: We have all the video systems that have existed in Canada since video appeared, that is to say since 1956. We have 2” AVR-1 machines made by Ampex, which are both low-band and high-band. They are made up of a number of racks full of electronic components, but these are also devices that mainly work by using compressed air: the travel of the tape is entirely managed by pneumatic valves. We also have a dozen AVR-2 2” Quadruplex systems currently in storage. In those days, this was portable equipment used in mobile studios to broadcast baseball games, for example. We own 1” machines from the 1980s, which we will use to transfer about 13,000 three-hour tapes. Then we have U-matic, BVU30, D2, Digital Betacam, etc.


PACKED: Are these machines catalogued?

René Paquet: Yes, thanks to the inventory that was made, they were catalogued and through their asset number, we can find their serial number and model.


PACKED: How frequently is this equipment used?

René Paquet: It depends on the project. We draw up a plan of action and if, for example, it involves the ½” open reel machines, then we will bring out the machines we have from storage, switch them on and do the transfer.


PACKED: How is the maintenance of this equipment managed?

René Paquet: 90% of our equipment is more than 10 years old and the remaining 10% is almost all more than 5 years old. Maintenance is divided into three categories. A green zone that includes 60 items that are still maintained by the manufacturer or the retailer. When the retailer no longer provides the maintenance service, the item will pass on to an intermediate zone represented by the colour yellow that comprises 590 items of equipment and corresponds to the maintenance we do here ourselves. The last zone, which is the red zone, involves ad hoc maintenance, done occasionally on certain equipment that is long since out of guarantee. The red zone currently consists of 500 items of equipment. When an item is in the red zone, it means that it is truly obsolete, if it is in the green zone, it means that the retailer still provides maintenance. The yellow zone is a sort of in-between.


Maintenance is organised in three categories. Courtesy of Library and Archives Canada.


PACKED: What kind of problems do you manage to repair by yourselves?

René Paquet: It depends on the category of the equipment; for equipment from the electronic category, this mainly involves the 2”, 1” and ¾” players. In the mechanical category, it mainly involves film appliances: the film printers and duplicators.


PACKED: Is the maintenance of the equipment in the red zone also done internally?

René Paquet: Yes, for example when we redesign part of a machine, such as an electronic card or a printed circuit board. The red zone is the one in which we try never to be. The situation of the equipment is one of the criteria for migration. If it is in the red zone, then it means that the transfer must be done without delay. The yellow zone is the warning signal for setting up the transfer to a more recent medium; otherwise it will become very complicated and very expensive.


PACKED: What video equipment still remains in the green zone?

René Paquet: The new and very recent equipment, like Digital Betacam is still in this zone. You will also find the high-technology equipment, such as the LTO-4 tape players. Our objective is to complete the migration before this equipment reaches the yellow zone. Eventually, migration will also be automated and, for example, after two new generations of LTO, the robot will transfer to LTO-7.


PACKED: Is a maintenance contract signed with the retailer or the manufacturer when the equipment is bought?

René Paquet:Yes. Ideally, it is better to stay in the green zone, because the yellow zone is relatively expensive, and the red zone is excessively expensive. In the yellow zone, the situation can be a hybrid one, but we will first look for a third party to deal with it. Maintenance costs us about 10 to 15% of the purchase price of the equipment. In the red zone, we will have additional problems finding the components and the very specific knowledge associated with a device or a breakdown.


PACKED: Do you have enough machines to accomplish the transfer of all the tapes in their entirety?

René Paquet: Yes, in theory we should have enough machines to complete the migration.


PACKED: Do you have a stock of spare parts?

René Paquet:Yes. When we acquire spare parts that are available, we buy 10 to 20 % more than the amount we think we will need. If we don’t have this extra percentage of parts, then they are parts that are still readily available for equipment in the green zone, such as the Digital Betacam. For equipment in the red zone, we must get by with what we have, with often 0% extra spare parts, and it is mainly in the yellow zone that we try to have 20% more spare parts.


Spare parts and components for video players.


Spare parts and components for video players.


PACKED: Which formats can be found in the red zone?

René Paquet:We find, for example, the IVC 9000 for a 2” helicoidal format, which was used in the 1970s and early 1980s as the archiving format at the Archives of Canada. After taking this decision, they modified the machine in order to be able to record four hours instead of two hours on the same tape, which allowed storage capacity to be doubled. This entailed a compromise of quality, losses of frequency bandwidth, etc. In addition, it is a machine that is one of a kind and many of its components and circuits are no longer available on the market.


PACKED: What do you do in this case?

René Paquet: We redesign parts internally to make the machine work again. This mainly involves electronic components. Either we redesign the circuits to stabilize the time base correction31, or we stabilize the servomechanism systems that are very unstable, to obtain a production of better quality. In some cases, we have to change the memory circuits and these are sometimes magnetic cores. The idea is to obtain the best possible signal and make a file of it.


PACKED: However, certain machines do not work?

René Paquet:Yes, some of them are simply stored. It depends on the plan of action decided on after surveying the collections and the type of tape to be transferred in priority according to the material and the physical condition of the media.


PACKED: How is your equipment stored?

René Paquet: The main problem with the equipment is finding a place to store it. We are often obliged to do with “whatever is left” in terms of storage places, and the climatic conditions cannot really be controlled. The only thing that we truly have control of is security and physical access to the machines.


PACKED: Ideally, what climatic conditions would you have to store the equipment?

René Paquet: For me, it should be the same conditions as those used for magnetic tapes: 18°C and 40% relative humidity. However, I don’t have any precise data on that.


PACKED: Do you use boxes to store the equipment?

René Paquet: Only when we have to send out equipment, and in this case we have specific made-to-measure boxes. Of course, this does not apply to the 2” or the Quadruplex machines, but rather to appliances such as the Digital Betacam players or D2. They are mainly used for dispatching equipment for maintenance. On very rare occasions, we have lent devices, to the modern art museum of Montreal, for example.


A box used for the handling of video players.


PACKED: Who makes these boxes?

René Paquet: We order them by sending our specifications to a plastics company in Ottawa, which specialises in making this type of box. There are several companies that make this type of box. When we send equipment by air, we put it in this type of box.


PACKED: When the equipment is stored for very long periods, do you plug them in from time to time to switch them on? Is this part of regular maintenance?

René Paquet: No, firstly because the places we use for storage do not have adequate electrical installations for this. Our primary difficulty is to find space for this equipment, especially for the 2” Quadruplex devices that take up a lot of space. The person in charge of managing storage space will tell us that it is too big and ask us to reduce the space we use, and if we can take the machine apart, remove this part, etc. The simple fact we can preserve the equipment whole is already quite something! The space that this represents, per square metre, is very expensive, so we are not in a position to control the humidity or the temperature of the place of storage. Formerly, we kept it all in the main building of the archives, but with very limited space, and with a high cost per square metre, we could not justify this.


PACKED: Where is it stored today?

René Paquet: It is stored in an external warehouse, belonging to the federal government who lent it to us for this equipment. It is a big warehouse, which is also used by certain museums that lack storage space. You might find a bus along side our 2” players, for example.


PACKED: How is the obsolescence of the equipment and migration managed from a budgetary point of view?

René Paquet: If we express migration with a curve, we have curve A for the original standard, and curve B for the destination format. The ideal time for migration will be when format A becomes less popular, whereas format B gains in popularity. The period of time it takes for format A to disappear, which starts at this moment of transition, is approximately the time that must be calculated to accomplish the migration. If we launch the migration too soon, we risk making a wrong choice of destination format, and what’s more, this format and the technologies that accompany it, being of a recent nature, will be very expensive. So we allow the technology to mature, and only then do we adopt it. Even though it is above all a lossless format, it is for this reason also that we chose MJPEG 2000.

If we compare the increase in maintenance costs with the risk that the collection incurs, we can see that whatever the amount of money invested in maintenance, the risk of losing documents will increase when we are in the red zone. If we start the migration too soon, it is very expensive, and if we start too late, then it is inordinately expensive. If we don’t start soon enough, not only does the risk for the collection increase, but also the costs. Demonstrating this visually using curves allows us to show the administrators of a collection what must be done to make the archiving work. It also shows that the funding necessary to invest in new formats will allow money to be saved throughout the entire safeguarding process.


 Courtesy of Library and Archives Canada.


PACKED: So, the risk of a very expensive belated migration process is a means to lobby for more anticipation with regards to safeguarding a collection?

René Paquet:Yes, and if we wait for too long, the risks incurred by the collection are greatly increased.


PACKED: How is the necessary maintenance for the equipment required for the migration evaluated?

René Paquet:For this, just as for the tapes, we did a survey. During an inventory, each device was identified by its format, the manufacturer, and the number of devices of this type that we possess. We also made note of the number of tapes we possess in this format, the number of hours that remain to be transferred, and the number of hours during which the machines are used, by month or by year. This survey also shows who does the maintenance (the manufacturer, someone internally, etc.), how regularly it is done, the problems that are encountered most frequently, and it indicates if we could replace the machine by another using today’s technology.

After making this inventory of equipment and the typical problems we encounter, we follow a maintenance programme to which the data gathered during the survey are added. This programme, currently under development, will be called the electronic maintenance program and will include several levels of maintenance. On one hand, there is preventive maintenance which takes place on a day-to-day basis, and which the operator must do when he gets to his workplace. Then there is corrective maintenance, which involves correcting problems.

The electronic maintenance program will then regroup not just the data from the survey, but also a number of other factors necessary for optimal maintenance, such as the history of repairs (What? When? How? By whom? At what cost? Etc.), the inventory of spare parts (capacitors32, video heads, etc.), professional contacts (technicians, retailers, etc.) as well as other important elements. This system is more or less compatible with the ITIL standard that is used to optimise research.


PACKED: Do you have an inventory of all of your spare parts?

René Paquet:This inventory ceased to exist about five or six years ago when the video maintenance department became part of the IT service. The data on the spare parts was deleted because it was not considered to be useful to the functioning of the IT service, as it concerned video. This was one of the setbacks in setting up the electronic maintenance program, and the inventory of the parts and the recording of this information in a database will have to be redone.


PACKED: What other types of information will be present in this system?

René Paquet: Professional contacts of service providers who are able to do repairs, or of persons who can provide advice about such or such a device, etc. Then all the technical symptoms will be included, or the things that must be looked at first to avoid these breakdowns. For recent equipment, a program that explains what must be done manages all this, but this does not exist for older equipment. One part concerns the management of this knowledge internally so that it is passed on correctly.


PACKED: So this is managing the passing on of expertise from one employee to another?

René Paquet:Yes, it is knowledge management.


PACKED: So all this information will be brought together thanks to the electronic maintenance program?

René Paquet: Yes but it is still just a project, the system does not yet exist. Once it is ready, it will be integrated with the Oracle database system, which manages the collections as a whole.


PACKED: How is maintenance organised?

René Paquet: There are three levels of maintenance: daily, preventive and remedial. In the green zone, external companies will do the preventive maintenance, such as updating their software or the hardware. In the yellow zone, daily and preventive maintenance is done internally, that is to say on an operational level. Technicians trained to perform certain operations do this: then, a more qualified technician will check the responses to frequencies and makes sure that the servomechanisms are still properly aligned. When things go beyond this, it is remedial maintenance, which means changing parts for example, and realigning or redesigning certain parts of the machine.


PACKED: Is documentation intended for the other and future members of the team drawn up by the most experienced technicians?

René Paquet: Yes, and these documents are called guidelines or procedures. These procedures to follow are written, but as I have already mentioned, a lot of this data was lost at one point because of the animosity between the IT and the video departments. It is a job we will have to do again, but in the present state of affairs, the equipment maintenance data and the inventory of spare parts exists only in the heads of the technicians.


PACKED: Will it be possible to know the maintenance and repair history of a device?

René Paquet: At the moment we are developing a system called C2. It is a system that reports all of the problems encountered by the computers and mainframes that we possess. Then, someone is responsible for receiving the anomaly messages and dispatching them to a specialist. The specialist receives a demand and makes the repair, and sends a record of this to a database with the asset number.

At present, this works for IT problems, and we are trying to do the same for video, including LTO, Quad, D2 and AVR-2 tapes, and all the tapes or servers that are useful to the video department. Each device here has an asset number. Thanks to the asset number of each device, we are able to incorporate it into a subsystem of the C2 system. As soon as someone reports a problem, it will be encoded and redirected to an internal or external technician. The problem will be identified then repaired and re-entered into the database with a description of what has been done, of what caused the problem, and how much time it took to fix it. It is still under development, but it should be operational shortly.


PACKED: How many technicians work here?

René Paquet: I am the only one left, all the others are retired. [update: René Paquet should retire from the Library and Archives Canada some time during 2011]


PACKED: Has it been arranged to train someone else?

René Paquet: It is something that should be done, but this is not easy. The problem is that if the person is qualified in new technologies, it will be much more attractive for this person to work on new technologies because this is where progress is made. Now, when you work with old technologies it is, in a certain sense, a step backwards, and this is a serious problem for recruitment. I worked for the engineering office at Radio Canada, and when I offered an interesting job to some of my former colleagues, they replied that they had forgotten all that and had turned to other things. The specialists in the field have evolved with the technology and are not interested in keeping the knowledge connected with these old machines.

Finding the workforce is a really arduous task, and developing the means to attract qualified people to this domain is too. Sometimes we find someone who has already done maintenance and is interested, but hiring him turns out to be a fiasco because he can be qualified to repair a TV but not a broadcast system. Broadcast systems already required specialist skills back then, it was more specific, it wasn’t like VHS33. This is something that can be difficult to explain to administration.

One of the recommendations I made was to hire two electronic engineers to do maintenance, but I still haven’t found them. The problem lies also in the fact that way that electronic circuits are manufactured has evolved over time and the students of today no longer learn this. Today, the field of electronics is much more orientated towards computers and processors, etc. Thirty or forty years ago, mechanics and servomechanisms were much more widely used than today, where miniaturisation has made things much more complicated.


PACKED: So it is rare for you to be able to call on external service providers, except when equipment is in the green zone?

René Paquet: In the yellow zone, we sometimes have shared maintenance: internal and outsourced. The objective is that the knowledge related to a technology be passed on from the company to the employees of the archives before it gets to the red zone, and that the company ends its service completely.


PACKED: What is the equipment acquisition policy?

René Paquet: At the level of acquisition, there is a committee that decides if a collection requires equipment or not, depending on the plan of action. If the need is there, the acquisition is made, and it is through this procedure that almost all the equipment has been acquired. The plan of action takes into account the necessary equipment, the maintenance these will require, and the total cost.


PACKED: Through which channels do you buy equipment and spare parts?

René Paquet: Through eBay of course, but also by developing contacts with companies through the NAB National Association of Broadcasters and the IBC34 International Broadcast Convention. NAB is a North American trade fair, whereas the IBC is more European. These are international trade fairs where you can meet a lot of manufacturers; it is networking among people.


PACKED: And for spare parts?

René Paquet: The first contact is made with the company, who will tell us if they can get the part or not, and what the waiting time will be, except for Ampex, which no longer exists. Then there are specialised firms.


PACKED: Do you have equipment that is used just for its spare parts?

René Paquet: Yes, we keep certain devices just so they can be cannibalised.


The head drum of a 1" tapes Ampex player.


PACKED: Can old players with built-in monitors also be repaired internally?

René Paquet: Yes, as Tektronix35 no longer repairs them. However, it is also possible to replace them by more modern equipment such as LCD screens.


PACKED: What other types of repairs and modifications do you do?

René Paquet: Most of our equipment is not in its original state, because we have done a lot of modifications so that they are more stable. This has become necessary because of temperature changes that change the characteristics of the transistors and make the servomechanisms work outside of their normal limits.

For the 2” helicoidal IVC I already talked about, we are working to install a SDI digital output to convert the signal to a file. Its quality is still relatively good, but it’s a machine that has been modified time and again to allow this.

On this machine, the circuit was modified, because there were a lot of problems with the time base correction. We had to stick in new circuits because there was no more room on the original card, to allow the servomechanism systems of the machine to be stabilised. It is an example of equipment that is right in the middle of the maintenance red zone. When we are lacking a card, we will redesign it for two reasons: because it is no longer available and because the new design will get the best out of the machine. Some cards are simply modified, whereas others have to be entirely rebuilt.

All these types of repairs are developed internally because they are not available otherwise. But even though what we do in this laboratory has a lot in common with preservation of technology, it is not done with the aim of making museum pieces. Our objective is to be able to continue transferring the tapes in the best possible conditions. This is why we use modified devices, and also a lot of software to make the machines work. If we had more time and a larger workforce, we could even develop a specialised system to control the way the playback system of the machine works by computer.


Two electronic boards for helical 2" IVC players. The one at the left is lightly modified, while the components of the one on the right have been almost entirely replaced by more modern components while using the original board carrier.


PACKED: But problems of electronic origin are however harder to repair on more modern equipment?

René Paquet: Yes, problems with circuit boards are a little more complicated to repair.


PACKED: The technicians themselves also record information about the equipment?

René Paquet: Yes, for example by updating the manuals when a modification is made. When we reproduce the drawings for the new circuits, we use Autocad, but a lot of these drawings have been lost.


PACKED: Where are the manuals stored?

René Paquet: They are stored both in the laboratory where the machines are repaired, and in an internal library of the Archives called the Lanctôt Library, which is named after a famous archivist. It is a library dedicated to conservation and this is where the manuals and certain documents written by the technicians are situated.


 The Library of manuals of the laboratory.


PACKED: Is the documentation kept in this library referenced to the asset numbers of the equipment?

René Paquet: No, not really, for the time being, the technician must know where to look for the information himself. However, apart from the manuals themselves, most of the documentation is also something we are going to look for on the Internet.


PACKED: What are the problems and the breakdowns that occur most frequently?

René Paquet: The servomechanism systems, wearing of the video heads, of the pinch rollers36, there are also breaks that wear out on most of the 2”, 1” type C and U-matic equipment. The belts are not really a problem because there are companies that will custom-build belts to fit with the adequate diameter. The problems we have with belts are mainly with working out their diameter. The inventory of the equipment will also shed light on the most frequent breakdowns. But the best source of information remains the experience of the operators who have worked and continue to work in particular with a specific machine.


PACKED: What breakdowns are caused by a long period of storage?

René Paquet: What’s sure is that the capacitors will dry after a while, the dielectrics inside will become faulty because their capacitance will change, and instead of having a value of 100 microfarads it will be 40 microfarads. For belts, if they remain under tension for too long, they will have a tendency to keep a particular shape and lose their elasticity. This is why it may be advisable to loosen the tension on the belt systems slightly, or remove the belts altogether. The problem is that in terms of time and cost, doing this sort of thing is not worth the increase in life expectancy that will be gained.


PACKED: What do you use to clean the players?

René Paquet:We use 99% alcohol to clean the metallic parts. We also use this on the rubber parts such as the pinch rollers, but there is a risk of it making the rubber go hard.


PACKED: Are there also problems of a purely mechanical order?

René Paquet: One thing that can happen is that mechanical parts such as gear wheels may become difficult to get to work because the grease that was originally there has worn off or dried up. I have often had problems with mechanical parts that were difficult to get to work on ½” open-reel players: the belts, the guide rollers, the rubber rollers that drive another mechanism, etc. This affects the tracking of the tape.


PACKED: Can there be problems related to the memory of the equipment?

René Paquet: With regards to memory, if information is lost, this will cause problems for sure. This is also the case for the little 3.5 Volt batteries in old computers that lasted two or three years, we have to spot the appliances in which they are present and replace them. And as with any battery forgotten inside a device, it can end up leaking. But the advantage with digital video, contrary to analogue video, is that when the information has to be transferred, it can usually be done very quickly.


PACKED: So, once all the collections are on LTO, the next migration will be much easier and much quicker?

René Paquet: Yes, in the world of file-based video, an hour of video can be converted from one format to another in about 15 minutes. As for the migration of our analogue tapes, we foresee that our entire collection will be available on file in 2019.





  • 1. 35mm is a standard film gauge for photographic film, 35 millimeters wide, and originally created for cinema. It has remained relatively unchanged since its introduction in 1892 by William Dickson and Thomas Edison. The standard negative pulldown for movies is four perforations per frame, or about 53 frames per meter. 35mm was recognized as the international standard in 1909 and has remained by far the dominant film gauge, thanks to a good tradeoff between the quality of the image captured and the cost of the film stock. The ubiquity of 35mm makes it the only format in the film industry, analog or digital, which can be projected in almost any cinema in the world. (Source: Wikipedia)
  • 2. An internegative is a master negative print.
  • 3. An interpositive is a color master positive print.
  • 4. 8 mm is a non-professional film standard introduced in 1932 by Kodak, using a filmstrip 8 mm wide (the frame size is 4.9 mm × 3.6 mm, giving a ratio of 1.36:1). In 1965, it evolved as Super 8, retaining the same width but with smaller perforations and a 36% larger frame size, of better quality, but above all much more practical to use. (Source: Wikipedia)
  • 5. 16mm film was introduced by Eastman Kodak in 1923 as an inexpensive amateur alternative to the conventional 35mm film format. 16mm refers to the width of the film. The format was initially directed toward the amateur market and was often referred to as sub-standard film by the professional industry. But 16mm has been extensively used for television production, and is still used by experimental filmmakers and other artists. The two major suppliers of 16mm film today are Kodak and Fujifilm.
  • 6. 28 mm gauge or Pathé Kok film was introduced in 1912 by Charles Pathé. This 28 mm wide film was created to compete with 35 mm. Pathé Kok was made from inflammable cellulose acetate, rather than the nitrocellulose used for 35 mm. (Source: Wikipedia)
  • 7. 2-inch quadruplex (also called 2″ quad, or just quad, for short) was the first practical and commercially successful videotape format. It was developed and released for the broadcast television industry in 1956 by the American Company Ampex.
  • 8. 1 inch Type C is an open-reel professional videotape format co-developed and introduced by Ampex and Sony in 1976. It became the replacement for the then-dominant Quadruplex (2 inch Quad for short) open-reel format, due to the smaller size and slightly higher video quality.
  • 9. A film recording of a video image displayed on a specifically designed television monitor. Also called 'Kine’. Only means of recording TV (television) programs before video recorders and tape were invented. (Source: The National Film and Sound Archive of Australia)
  • 10. ½” open reel is an analogue video format introduced in 1965. The ½-inch tape is not inside a cassette but on an open spool. The tapes were used in combination with the first portable video recorders and were widely used by artists, lecturers and activists. Broadly speaking, there are two categories of ½” open reel: CV (Consumer Video/Commercial Video) and AV (EIAJ Type 1). Although the tapes look identical, the players are not compatible.
  • 11. D-2 is a professional digital videotape format created by Ampex and other manufacturers through a standards group of the Society of Motion Picture and Television Engineers (SMPTE) and introduced at the 1988 NAB (National Association of Broadcasters) convention as a lower-cost alternative to the D-1 format. D-2 used 19 mm (¾ inch) metal particle tape loaded into three different sized cassettes. PCM-encoded audio and timecode are also recorded on the tape. Although the D-2 tapes are similar in appearance to the D-1 format, they are not interchangeable. (Source: Wikipedia)
  • 12. Serial Digital Interface (SDI) refers to a family of video interfaces standardised by the Society of Motion Picture and Television Engineers (SMPTE).
  • 13. RTI is an American company that sells, amongst other things, machines that clean and evaluate videotapes of different formats such as 1 inch or U-matic. See:
  • 14. U-matic ¾ inch is an analog video format that was developed by Sony in the late 1960s and consisted of a ¾ inch tape inside a cassette. Its successor was the Betacam analog format.
  • 15. Betacam is a professional videotape format developed by Sony from 1982 and launched in 1983. Like with VHS, the cassettes, with a ½ inch tape, come in two sizes: S and L, which are of two different colours. It was the first professional analog format that allowed separate recording of luminance and chrominance signals.
  • 16. A program by the English company Gallery. See:
  • 17. A great number of videotapes are victim of the Sticky Shed Syndrome, which is the result of hydrolysis of the binder that attaches the particles of iron oxide to the plastic carrier.
  • 18. See:
  • 19. Digital Betacam or DigiBeta is a digital version of the Betacam tape format. For a very long time it has been considered by television and other audiovisual archives as a good archival format because there is no generational loss between two copies.Linear Tape-Open (or LTO) is an open format developed in the late nineties for storing data on magnetic tape.
  • 20. Linear Tape-Open (or LTO) is an open format developed in the late nineties for storing data on magnetic tape.
  • 21. MJPEG 2000 or Motion JPEG 2000 is the third part of the JPEG 2000 image compression standard and is applied to video. The principle is very simple: each frame of video is encoded to the JPEG 2000 format. An MJPEG 2000 video is therefore a simple concatenation of images in JPEG 2000 format, requiring just a few minor changes to the header.
  • 22. Material eXchange Format (MXF) is a "container" or "wrapper" format for professional digital video and audio media defined by a set of SMPTE standards. It supports a number of different streams of coded "essence", encoded with any variety of codecs, together with a metadata wrapper, which describes the material contained within the MXF file.
  • 23. H.264 is a digital video compression codec for high definition images and video using the MPEG-4 standard, developed by the VCRG (Video Coding Experts Group) in partnership with the MPEG (Moving Picture Experts Group), also known as AVC (Advanced Video Coding).
  • 24. See:
  • 25. Super DLT (SDLT) is a high capacity variant of DLT.
  • 26. A magnetic tape technology using ½ inch (1.27 cm) cartridges with a capacity of 10 to 70 GB. DLT technology quickly became widespread from 1995 and was widely used for local networks in companies. (Source: Grand dictionnaire terminologique, Office québécois de la langue française.)
  • 27. The 8 mm Backup Format is a magnetic tape data storage format used in computer systems, pioneered by Exabyte Corporation. It is also known as Data8. Such systems can back up to 40 GB of data depending on configuration. The tapes used are mechanically the same as the tapes used in 8 mm video format recorders and camcorders.
  • 28. René Paquet is referring to Digital Data Storage (DDS), a format for storing and backing up data on magnetic tape that evolved from Digital Audio Tape (DAT) technology, which was originally created for audio recording. In 1989, Sony and Hewlett Packard defined the DDS format for data storage using DAT tape cartridges. DDS uses tape with a width of 3.8mm, with the exception of the latest format, DAT 160, which is 8mm wide.
  • 29. The IBM System/360, released in 1964, introduced what is now generally known as 9-track tape. As with the earlier IBM 7-track format it replaced, the magnetic tape is ½ inch (12.65 mm). (Source: Wikipedia)
  • 30. U-matic spawned two variants: BVU (Broadcast Video U-matic), introduced in 1978, and BVU SP (Broadcast Video U-matic Superior Performance), introduced in 1988. These two variants were both designed to improve picture quality.
  • 31. This is performed by a Time Base Corrector, an electronic device used to correct the instability of a video signal during the playing of a tape.
  • 32. A capacitor (formerly known as condenser) is a passive electronic component consisting of a pair of conductors separated by a dielectric (insulator). When a potential difference (voltage) exists across the conductors, an electric field is present in the dielectric. This field stores energy and produces a mechanical force between the conductors. The effect is greatest when there is a narrow separation between large areas of conductor, hence capacitor conductors are often called plates. Practical capacitors are available commercially in many different forms. The type of internal dielectric, the structure of the plates and the device packaging all strongly affect the characteristics of the capacitor, and its applications.
  • 33. VHS designates a recording standard of video signal onto ½ inch tape developed by JVC in the late 1970s. Its mass distribution was launched in 1976. During the 1980s and 1990s, VHS became the standard format of general public video ahead of its competitors: Sony Betamax and Philips V2000.
  • 34. See:
  • 35. See:
  • 36. The pinch roller is a rubberized free spinning wheel typically used to press magnetic tape against a capstan shaft in order to create friction necessary to drive the tape along the magnetic heads (erase, write, read). (Source: Wikipedia)
logo vlaamse overheid