Teletext Recovery Work

In 2022 I began my journey with recovering the wonders of teletext from home-recorded VHS tapes.

This page isn't here to serve as a "what is teletext" or "how to recover teletext" (the latter will come later), but is here to serve as a public view of my progress of recoveries; and importantly as a way that people can download my recoveries if they so wish.

Table Structure

• Capture Name - An internal name used by myself, of the VBI capture.
  (NB. I won't be providing these files publicly, why? Because some of them are 20GB+ in size which is a lot to host, fear not I have the originals - if you need them please contact me via the Contact link in the heading.
• D# - Directory number, internal use.
• Date - confirmed date of the capture. (dd/mm/yyyy)
• Time - confirmed time range of the capture.
• Service - name of the service captured (e.g. CEEFAX, TELETEXT). Plus for TELETEXT the region of the capture.
• Channel - the name of the channel the capture comes from (e.g. BBC1, Channel 4).
• Programme Schedule - the identified programmes that were being broadcast during this capture. [p] means partial recording.
• J? - did the capture cover any junctions, the parts between shows (e.g. "Coming up next on BBC1...").
• A? - did the capture cover any advertisments.
• Tape ID - any identifying markers on the cassette shell/packaging, internal use only.
• Notes - any other interesting notes about this capture, such as picture quality, interruptions etc.
• VCR - the model number of the VCR used for the capture.
• Deconvolved - the process of cleaning up and preparing the original captured data stream.
• Squashed - a process unique to vhs-teletext wherein duplicate subpages and errors are removed and reduced.
• Final - a final copy of teletext data that I have manually cleaned up and is (in my opinion) as best as it can be.

Further Notes

• All files provided here are of the .T42 file type, which is generally agreed to be the teletext file format. Most teletext adjacent programs will work with these files.
• All "automated" processing of these files is completed via ali1234's vhs-teletext tool.
• The reason for providing different versions of the same file, is that there may be hidden (or yet undiscovered) data in earlier revisions.
• It is natural that the files usually get smaller with each revision.
• Capture times are approximated. Total captured pages is calculated in Meddler SE.
• Any capture should only contain pages for the service and date specified.
• As of Capture #0011 the real-time capture for each entry should be no longer than 15 minutes. Issues occur with the deconvolve process if the captures are too large.

Guides

As teletext recovery can be considered a bit of a dark art still, I have created a series of guides that explain various procedures you might go through when performing teletext recovery operations.

• Teletext Recovery Environment Setup Guide - a guide that explains how to prepare an environment for the capture and initial processing of teletext data.
• Teletext Capture and Recovery Guide - a guide that explains how to capture and process teletext data
• Guide to Speeding Up Teletext Data Procesing with NVIDIA CUDA - a guide that explains how to greatly improve the speed at which you can process raw teletext data (sometimes referred to as the VBI data).

Tools

If you're new to all this and unsure what to do with these T42 files, or if you want to play with teletext then see these links to various useful tools:

• Teletext Meddler SE My primary tool used for viewing my T42's, editing them, adjusting stuff etc. Download Page
• Teletext Recovery Editor Another tool for viewing and editing, but has a limit and won't work with large captures. Download Page
• Teletext Player A great tool for viewing teletext files in the most "natural" way on a computer. Download Page

Recoveries

These are all my recoveries, some of them are extremely poor and some are nearly perfect. Download the .t42 files from the right-hand side of the table. If you want the nicest, cleanest captures you'll want the Final column, if you want the original capture you'll want the Deconvolved column and if you want a stream that has the duplicates removed you'll want Squahsed.

If the Deconvolved column states !JUNK! then I have looked at the data and can't see anything of value. This is usually due to the tape being in such poor condition a recovery has failed. Data is still available if you want to try and repair the data given.

Some files have the extension of .tidy at the end. Fear not, they're T42 files as you'd hope, just trimmed down to remove some excess off the end.