A TCR file is a plain-text book compressed for the Psion Series 3, a palm-sized organiser from the early 1990s. The name stands for Text Compression for Reader. Halving the size of a novel mattered on a machine with a few hundred kilobytes of memory, so whole libraries of public-domain texts were packed this way, and some e-readers kept the format alive long after Psion stopped making handhelds. If you found a TCR file today, it most likely comes from an old text archive or from the memory of a PocketBook reader.
Where TCR came from
Psion's handhelds ran the EPOC operating system, the ancestor of Symbian. TCR grew out of Psion's compression scheme and competed with Palm's PalmDoc format. It compressed slightly better, while PalmDoc stored more information and had more books. Freeware readers such as EBook for EPOC opened both. With the end of Psion's consumer handhelds the format faded, though Cool Reader and older PocketBook models still list it.
What is inside a TCR file
The format is refreshingly simple:
- A 9-byte signature:
!!8-Bit!!. You can see it if you open the file in a text editor. - A dictionary of 256 entries, each a short string of bytes.
- The text itself, where every byte is an index into that dictionary.
The compressor replaces frequent pairs of characters, and then pairs of pairs, with single codes. A typical book shrinks to roughly half its size. Decompression is just table look-ups, which suited a slow processor.
What TCR does not store: formatting (no bold, no headings, no images), metadata (no title or author) and the character encoding. The text is in whatever 8-bit code page the author used.
How to open a TCR file today
- Android: Cool Reader opens TCR directly.
- Windows, macOS, Linux: calibre reads and writes TCR. The desktop version of Cool Reader (crqt-ng) also lists it. On Windows, STDU Viewer opens it too.
- E-readers: older PocketBook models list TCR among their formats. Kindle, Kobo and iPhone do not read it, so convert to EPUB.
Common problems
"TCR file shows garbage in Notepad"
That is expected: only the !!8-Bit!! header is readable, the rest is compressed codes. Open the file in a reader or convert it to TXT.
No title or author after conversion
TCR has nowhere to store them, so the result is named after the file. Fix the metadata in calibre or rename the file before converting.
Wrong characters in non-English books
Our converter guesses the encoding of the unpacked text. With short files or unusual code pages the guess can fail. Convert it in the free calibre app and set the input encoding by hand (for example cp1251 for Russian).
Converting TCR
Our converter unpacks a TCR file, handles it as plain text and works out where paragraphs begin. Useful targets:
- TXT to get the plain text back as UTF-8, readable anywhere.
- EPUB for any phone or e-reader, and from there EPUB to PDF if you need pages.
- FB2 or MOBI for readers that prefer those.
Since the source has no formatting, the result will not have chapters unless the text contained recognisable chapter headings. Converting to TCR works as well. The output is UTF-8, which Cool Reader and PocketBook handle; a real Psion Series 3 only understands 8-bit text, so accented letters would come out wrong there.
TCR at a glance
| Developer | Psion (format of the Psion Series 3 era) |
|---|---|
| Introduced | early 1990s |
| Type | compressed plain text |
| Compression | 256-entry dictionary, about 50 % |
| Signature | !!8-Bit!! |
| Formatting, metadata | none |
| Readers today | calibre, Cool Reader, older PocketBook models |
More: TCR on the MobileRead Wiki and calibre's TCR compression code.