Seismic data recorded decades ago is still valuable. Reprocessing, basin studies, regulatory filings and asset sales all depend on access to historical surveys. The problem is that the data often exists on tape formats that modern systems cannot read. Tape transcription — copying data from legacy media to current formats — is the only way to keep that data accessible.

Why legacy tape still matters

Operators sometimes assume that old tapes can be discarded once the data "must be" somewhere else. That assumption is often wrong. Common reasons legacy tape remains critical:

  • Sole copy — the tape is the only surviving record of a survey. No backup exists, and reacquisition is impossible or prohibitively expensive.
  • Merger and acquisition packages — when assets change hands, the buyer inherits whatever media the seller has. That frequently includes boxes of tapes with unknown contents.
  • Regulatory retention — provincial and federal rules require operators to retain seismic data for defined periods. The rules apply to the data, not the format, but the data must remain readable.
  • Reprocessing potential — modern processing algorithms can extract more information from raw field data than the original processing did. That raw data is often on the oldest tapes.

The value is real, but it has a time limit. Every year, the hardware and expertise needed to read legacy formats become scarcer.

Media types still in the field

Archives built over decades contain a mix of formats. The most common legacy seismic media include:

3480 cartridges

The IBM 3480 cartridge was a standard for seismic data in the 1980s and 1990s. It uses half-inch tape in a plastic shell, with a capacity of about 200 MB per cartridge. Drives are no longer manufactured, and working units require maintenance from specialists. Data on 3480 is often field data or early processed stacks.

9-track reels

Half-inch open-reel tape, recorded at 800, 1600 or 6250 BPI. This was the dominant seismic format before 3480. Many 9-track reels date to the 1970s and early 1980s. The tape itself is vulnerable to degradation: binder breakdown causes oxide shedding, and improper storage leads to edge damage. Functional 9-track drives are rare and require experienced operators.

LTO families

LTO (Linear Tape-Open) is a more recent format, with generations from LTO-1 (2000, 100 GB native) through LTO-9 (2021, 18 TB native). LTO is not truly "legacy" in the same sense as 9-track, but older generations face a compatibility wall: LTO drives can typically read only two generations back. An LTO-1 or LTO-2 tape cannot be read by current drives without intermediate migration. Seismic data written to early LTO is now at risk if it was never migrated forward.

Odd formats

Some archives include VHS tapes (used for observer video logs or scanned documents), DAT cartridges, Exabyte 8mm tapes, and proprietary formats from specific acquisition systems. These are less common but can hold irreplaceable supporting documentation.

Risks of waiting

Legacy tape does not fail suddenly. It degrades gradually, and by the time failure is obvious, recovery may be impossible. Key risks:

  • Binder degradation — the adhesive that holds magnetic oxide to the tape base breaks down over time, especially in humid or fluctuating conditions. This causes sticky-shed syndrome: the tape literally sticks to the heads and sheds oxide as it plays. Data loss is permanent.
  • Drive scarcity — manufacturers stopped making 9-track drives decades ago, and 3480 drives are now out of production. The remaining units are aging, and parts are hard to source. When a drive fails, there may be no replacement.
  • Operator expertise — reading legacy tape is not just a hardware problem. It requires knowledge of tape formats, block sizes, label conventions and error-recovery techniques that fewer people possess each year.
  • Unlabeled reels — many tapes have external labels that are faded, missing or wrong. Reading the tape header is the only way to identify contents, and that requires a working drive.
  • Mismatched catalogues — the database says a tape contains a certain line, but the tape header says something else. Without reading the tape, the discrepancy is invisible.

The cost of transcription today is modest compared to the cost of losing irreplaceable data tomorrow.

Transcription workflow

A practical transcription project follows a defined workflow. The steps below apply whether you are handling a single box or an entire library.

1. Intake

Tapes arrive by courier, freight or hand delivery. Each shipment is logged: number of items, apparent format, external labels, physical condition notes. Intake creates a chain of custody that follows the media through the entire process.

2. Inventory

Before reading, every tape is inventoried. This includes photographing labels, recording external markings, and noting any visible damage (cracked shells, exposed tape, water stains). The inventory is the baseline for comparing what the client thought they sent against what actually arrived.

3. Read

Tapes are mounted on appropriate drives and read to digital files. For degraded media, this step may require baking (low-temperature treatment to temporarily restore binder integrity) or multiple passes with adjusted tension. Not every tape is recoverable. A professional transcription service reports unreadable tapes in writing, with an explanation of what was attempted.

4. Validate

The digital file is checked against expectations. Does the header match the label? Does the file size correspond to the expected data volume? Is the file internally consistent? For seismic data in SEG-Y or similar formats, basic structural validation confirms that traces are present and headers are readable.

5. Catalogue

This is the critical step that many transcription projects skip. The digital file must be linked to an identity: line name, project, area, acquisition date, client. Without cataloguing, the file is just bytes with a filename. With cataloguing, it is a searchable, retrievable asset. Identity information comes from tape headers, external labels, and cross-reference to existing databases.

See How to Audit a Seismic Data Archive for guidance on checking catalogue quality after migration.

6. Deliver

Digital files are delivered by the method the client specifies: download, FTP, USB drive, LTO tape (a current generation), or cloud transfer. Delivery includes a manifest that lists every file, its source tape, and its catalogue identity. The client should be able to match what they sent to what they received.

Identity must travel with the data

A common failure in transcription projects is losing the link between the digital file and its origin. A filename like TAPE_0047.sgy tells you nothing. A filename like ProjectAlpha_Line12_1987_FieldStack.sgy, combined with a catalogue entry that includes coordinates and acquisition dates, tells you everything.

Every step of the workflow — read, validate, catalogue, deliver — must preserve and propagate identity. If the transcription service strips identity to simplify delivery, the client has traded one problem (unreadable tape) for another (unidentified file).

This is why United DataWyse treats cataloguing as part of transcription, not a separate service. The bits are worthless without the metadata.

Storage after digitization

Once data is transcribed, two questions remain: what do you do with the digital files, and what do you do with the original tapes?

Digital storage

Transcribed data should live in secure, redundant storage with defined backup policies. Cloud storage, on-premise servers or managed archive services all work, provided the storage is monitored and the data is periodically verified. A file that is unreadable because a disk failed is no better than a tape with sticky-shed. See Secure Records Storage for options.

Original media

Some operators destroy original tapes after successful transcription. Others retain them as a secondary backup, especially for regulatory compliance. The decision depends on storage costs, regulatory requirements and risk tolerance. If tapes are retained, they should be stored in climate-controlled conditions to slow further degradation.

Frequently asked questions

Can all legacy tapes be read?

No. Severely degraded tapes — those with extensive oxide shedding, physical damage, or mould — may be unrecoverable. Even with baking and careful handling, some tapes yield only partial data or none at all. A reputable transcription service will tell you upfront which tapes failed and why, not promise results it cannot deliver.

How much does seismic tape transcription cost?

Pricing depends on format, volume and condition. A box of clean 3480 cartridges is cheaper per tape than a single 9-track reel with sticky-shed syndrome. Most services quote per tape or per gigabyte, with additional charges for baking, multiple read attempts, or cataloguing. Get a written quote before shipping.

Should I transcribe everything at once or in batches?

Batches are usually practical. Start with high-priority media: tapes for active projects, tapes in poor condition, or tapes that underpin regulatory filings. This spreads cost over time and lets you learn from early batches before committing to the full library.

What if I do not know what is on a tape?

Reading the tape header is the only reliable way to identify contents when external labels are missing or suspect. This is part of the inventory and read steps. Unknown tapes are common in acquired archives; identifying them is a normal part of transcription work.