{"id":10358,"date":"2026-08-25T11:11:12","date_gmt":"2026-08-25T01:11:12","guid":{"rendered":"https:\/\/www.paradisec.org.au\/blog\/?p=10358"},"modified":"2026-08-25T11:11:12","modified_gmt":"2026-08-25T01:11:12","slug":"sg2-adventures-with-tape","status":"publish","type":"post","link":"https:\/\/www.paradisec.org.au\/blog\/2026\/08\/sg2-adventures-with-tape\/","title":{"rendered":"SG2: Adventures with tape"},"content":{"rendered":"\n

When the SG2 collection arrived at PARADISEC, it consisted of approximately 150 cassette tapes carried from Papua New Guinea to Australia in two duffel bags. They had been gathered through a community outreach project led by my colleague Steven Gagau, who invited community members to contribute cassette tapes they had lying around so he could bring them to Australia for digitisation, then return the tapes and digital copies to the communities they came from. You can read more about the first stage of this project here<\/a>.<\/p>\n\n\n\n

The recordings largely date from the 1970s through to the 2000s. Much of the collection consists of commercially produced music recorded in local studios by bands and solo performers, along with local choirs and other recordings from across Papua New Guinea. Many of the labels that originally issued this music are now defunct, and if the master tapes still exist, nobody seems to know where they are or what condition they\u2019re in. For some recordings, SG2 is the only organised digitisation effort beyond a random song or two finding its way onto YouTube.<\/p>\n\n\n

\n
\"\"<\/a>
Two duffel bags of tapes from PNG.<\/figcaption><\/figure>\n<\/div>\n\n\n

Assessing the collection<\/strong><\/p>\n\n\n\n

When Steven returned to Australia and we opened the bags, I became slightly overwhelmed by the task ahead. The first thing I noticed was the amount of dirt. It was caked onto the cassette shells, worked into the openings and sitting loose inside them. Metal components had rusted, while pressure pads and slip sheets (the thin liners that help the tape move smoothly inside the shell) were disintegrating or suffering damage. Many of the shells were also damaged or no longer suitable for playback.<\/p>\n\n\n\n

Every cassette needed to be opened and inspected because these problems were not always visible from the outside, and often the tape itself was not in much better condition. Some tape packs (the wound mass of tape on each reel) were loose or uneven, while others were wound so tightly or poorly that the reels could no longer rotate inside the shell. There were creases, stretched and broken sections, and transparent areas where the oxide layer had worn away completely.<\/p>\n\n\n\n

\n
\"\"<\/a>
Tape with large hole in the casing<\/figcaption><\/figure>\n\n\n\n
\"\"<\/a>
Missing pressure pad, plenty of dirt<\/figcaption><\/figure>\n\n\n\n
\"\"<\/a>
Worn section of tape showing where the oxide layer had come off completely<\/figcaption><\/figure>\n\n\n\n
\"\"<\/a>
Dirty tape internals with disintegrated slip sheets<\/figcaption><\/figure>\n\n\n\n
\"\"<\/a>
Badly mangled tape<\/figcaption><\/figure>\n\n\n\n
\"\"<\/a>
Dirty tape<\/figcaption><\/figure>\n<\/figure>\n\n\n\n

The LM-3032<\/strong><\/p>\n\n\n\n

As luck would have it, SG2 arrived at around the same time as the LM-3032 Tape Restorator, a bespoke machine designed and built for PARADISEC by Sam King. The machine is an updated version of a lubrication machine Sam designed and built with colleague Doug Smith at AIATSIS.<\/p>\n\n\n\n

Tape has a tiny amount of lubricant mixed into its coating to help it glide smoothly over the playback head. Over time this lubricant can dry out, creating enough friction for the tape to squeal, move unevenly or stop altogether during playback. Relubrication can make these tapes playable again, but the amount applied needs to be carefully controlled. Too little may not solve the problem, and too much can leave residue throughout the playback path.<\/p>\n\n\n\n

The LM-3032 uses a syringe pump to apply cyclomethicone (a silicone fluid used in cosmetics) evenly across the length of the tape. The tape passes over a soft ribbon, which distributes the lubricant across its oxide surface, before travelling past three ducts blowing hot air. This evaporates the volatile elements and leaves an extremely thin lubricating film before the tape is respooled.<\/p>\n\n\n\n

The fact that the ribbon ran across the oxide side initially concerned me because several SG2 tapes had visible wear or flaking. Before using the machine on the collection, I tested it on expendable tapes, including tapes with flaky oxide. The process was soft enough that treatment didn’t appear to cause further damage.<\/p>\n\n\n\n

The early stages of using the machine involved a fair amount of experimentation. The main variables were tape speed and the rate at which the lubricant was applied. Different tape formulations absorbed it at different rates, with some noticeably thirstier than others. Once I became familiar with the balance between speed and flow, the machine was intuitive to operate, although there was no single setting that worked equally well for every tape.<\/p>\n\n\n\n

I also realised early on that the restorator would inadvertently double as a cleaning device. As the tape passed over the ribbon, it deposited dirt that had remained after manual cleaning and by the end of the project the amount of debris collected on it was considerable. Seeing that build-up made it difficult to imagine achieving the same result without the machine, especially when the alternative was allowing the gunk to travel through the playback deck.<\/p>\n\n\n\n

\n
\"\"<\/a>
The LM32 tape lubricator<\/figcaption><\/figure>\n\n\n\n
\"\"<\/a>
Dirt being removed from tape as it passed over the ribbon<\/figcaption><\/figure>\n<\/figure>\n\n\n\n

The process<\/strong><\/p>\n\n\n\n

Each cassette was inspected, opened and cleaned manually. Dirt and debris were removed from the shell and as much as possible was removed from the tape pack itself. The tape was then transferred into a clean working shell, run through the LM-3032, baked at 55 degrees Celsius, then digitised using a Tascam 122 MkIII tape deck. Once the transfer was complete, the tape was returned to its original cleaned shell.<\/p>\n\n\n\n

Every tape was baked and every tape went through the restorator, including those that looked acceptable on first inspection. Visual condition helped identify obvious problems, but it was not a reliable indication of how the tape would behave once it began moving. Treating everything also reduced the likelihood of discovering halfway through a transfer that a seemingly clean tape was about to start squealing or depositing dirt inside the Tascam.<\/p>\n\n\n\n

I did not play the tapes before treatment because I didn\u2019t want dirt and debris entering the deck, which unfortunately meant there was no controlled before-and-after comparison for the restorator across the collection. Its effect was clearest when a tape continued to squeal after the first pass, as a second treatment would usually reduce the squeal considerably or remove it altogether.<\/p>\n\n\n

\n
\"\"<\/a>
Workstation<\/figcaption><\/figure>\n<\/div>\n\n\n

Getting sound from tape to computer<\/strong><\/p>\n\n\n\n

Some cassettes transferred cleanly in a single pass, while others had to be stopped repeatedly for cleaning or repair and the resulting sections joined digitally afterwards. For the more difficult cassettes, the final result was a reconstruction made from several attempts rather than a straightforward capture of one uninterrupted playback. One tape was wound so badly that it could not move through the shell at all, so I cut it in half, treated and digitised each half separately, and reassembled the recording as a digital file.<\/p>\n\n\n\n

Azimuth adjustment was necessary for every cassette and sometimes several times within the same recording. On damaged tapes, the best setting could shift as the tape played, so a passage that sounded clear on one transfer might deteriorate later, while another pass recovered that section but sounded worse elsewhere. I would sometimes make several transfers using different settings, then assemble the strongest sections into a single WAV file. For these tapes, an unsuccessful first pass wasn\u2019t enough reason to stop trying.<\/p>\n\n\n\n

This<\/em> particular song has three edit points from shifting azimuth positions (SG2-108: Mama Day by Leonard Kanya)<\/em><\/summary>\n

Some recordings contained severe dropouts that no mechanical treatment could fix. A few tapes appeared to have been stored near a magnetic field, leaving repeated losses in the signal (SG2-073<\/a>), while stretched or worn sections continued to sound poor regardless of how carefully they were treated. The LM-3032 could reduce friction, remove remaining dirt and improve movement through the deck, but it could not reverse creases, stretching, oxide loss or damage to the recorded signal.<\/p>\n\n\n\n

\n
\"\"<\/a>
Expanded, unmovable tape pack<\/figcaption><\/figure>\n\n\n\n
\"\"<\/a>
After hand-winding it to the middle, the tape was halved for digitisation<\/figcaption><\/figure>\n<\/figure>\n\n\n\n

Working out what was actually on them<\/strong><\/p>\n\n\n\n

SG2 includes recordings in Tok Pisin, Kuanua and other regional languages from Papua New Guinea. As I don’t speak those languages, my metadata mainly covered playback notes, physical condition, apparent duplicate tapes and obvious discrepancies between what was written on the case and what could be heard on the tape. The labels were often faded, water-damaged, missing or attached to the wrong cassette, and tapes and covers had frequently become separated. A cassette labelled as a particular band might contain a church choir, several unrelated recordings or material that had been partly taped over. In many cases, there was no reliable way for me to identify the performers or contents beyond noting that they did not match the information provided. Steven and his team in Papua New Guinea would work on these discrepancies.<\/p>\n\n\n\n

After several months of work spread across one or two days each week, the final tapes were completed in December 2025. When Steven travelled back to Papua New Guinea later that month, he brought with him a hard drive containing the complete digital collection so the metadata could be enriched. The identifications made by the metadata team in Papua New Guinea were added to the records, and copies were made available through both a community archive and the PARADISEC catalogue. <\/p>\n\n\n\n

Having people involved who knew the languages, music and communities represented in the recordings was essential to making sense of material that would otherwise have remained poorly identified. It also meant that the people who contributed the tapes were actively involved in its preservation and longevity. For a collection like SG2, access to that knowledge is as important as access to the recordings themselves.<\/p>\n\n\n\n

Some of the transfers still contain dropouts and anomalies that could not be repaired. You can hear in the wear how many hours of enjoyment a tape must have brought somebody. Even so, people were reunited with recordings they had not heard in decades and hearing them no longer depended entirely on the survival of the original cassette or access to a functioning tape deck. <\/p>\n\n\n

\n
\"\"<\/a>
Tape with no label<\/figcaption><\/figure>\n<\/div>\n\n\n

The music itself<\/strong><\/p>\n\n\n\n

One of the more interesting parts of working on SG2 was that much of the collection contained music, which was not only enjoyable but also gave me some indication of how international popular music had circulated through Papua New Guinea and influenced local musicians.<\/p>\n\n\n\n

SG2-151 included a version of \u201cYou Can\u2019t Do That\u201d by The Beatles. It was interesting to hear a relatively less obvious Beatles song turn up in this context and to consider the route it must have taken before being performed and recorded by a group in Papua New Guinea (Oroko Mates). British and American influences were clearly present throughout much of this material, although the musicians had filtered them through their own style.<\/p>\n\n\n\n

Listen to “You Can’t Do That” by Oroko Mates (SG2-151)<\/em><\/summary>\n
\n
\n
Listen to “Lus Man” by Jungle Image String Band (SG2-109)<\/em><\/summary>\n