Working with files for the Tiptoi® pen

Version on this page:
LTS Haskell 9.21:
Stackage Nightly 2017-07-25:
Latest on Hackage:1.8.1@rev:1

See all snapshots tttool appears in

MIT licensed by Joachim Breitner
Maintained by [email protected]
This version can be pinned in stack with:tttool-,2490

Module documentation for

There are no documented modules for this package.

Das tttool hat jetzt eine deutsche Webseite für Anwender:


The goal of this project is to understand the file and paper format for the Ravensburger TipToi pen. The ultimate goal is that everyone can create their own books, with their own sounds.

The current status is that we understood most of the file format (see the GME file format specification). We provide a tool that allows you to dissect these files.

The tool can also be used to generate completely new files from scratch; see below for details.

If you want to learn more please have a look into our wiki (

The tttool tool

Use the tool tttool.hs to investigate the gme files and build new ones. It supports various subcommands:

Usage: tttool [options] command

    -t <transcriptfile>
       in the screen output, replaces media file indices by a transscript

    info <file.gme>...
       general information
    media [-d dir] <file.gme>...
       dumps all audio samples to the given directory (default: media/)
    scripts <file.gme>...
       prints the decoded scripts for each OID
    script <file.gme> <n>
       prints the decoded scripts for the given OID
    raw-scripts <file.gme>...
       prints the scripts for each OID, in their raw form
    raw-script <file.gme> <n>
       prints the scripts for the given OID, in their raw form
    binaries [-d dir] <file.gme>...
       dumps all binaries to the given directory (default: binaries/)
    games <file.gme>...
       prints the decoded games
    lint <file.gme>
       checks for errors in the file or in this program
    segments <file.gme>...
       lists all known parts of the file, with description.
    segment <file.gme> <pos>
       which segment contains the given position.
    holes <file.gme>...
       lists all unknown parts of the file.
    explain <file.gme>...
       lists all parts of the file, with description and hexdump.
    play <file.gme or file.yaml>
       interactively play: Enter OIDs, and see what happens.
    rewrite <infile.gme> <outfile.gme>
       parses the file and serializes it again (for debugging).
    export <infile.gme> [<outfile.yaml>]
       dumps the file in the human-readable yaml format
    assemble <infile.yaml> <outfile.gme>
       creates a gme file from the given source
    oid-code [-d DPI] <codes>
       creates a PNG file for each given code
       scale this to 10cm×10cm
       By default, it creates a 1200 dpi image. With -d 600, you
       obtain a 600 dpi image.
       <codes> can be a range, e.g. 1,3,1000-1085.
       Uses oid-<code>.png as the file name.
    oid-code [-d DPI] <infile.yaml>
       Like above, but creates one file for each code in the yaml file.
       Uses oid-<product-id>-<scriptname or code>.png as the file name.
    raw-oid-code [-d DPI] <raw codes>
       creates a PNG file with the given "raw code". Usually not needed.
       Uses oid-raw-<code>.png as the file name.

A transscript is simply a ;-separated table of OIDs and some text, see for example transcript/WWW_Bauernhof.csv.


This program is written in Haskell and can be installed on Windows, MacOS or Linux.

For Windows users, we create zipfile containing tttool, you can find them in the releases section of the github project.

Otherwise, installation from source is not difficult either:

  1. First install the Haskell platform, see for details for your system. Users of Debian or Ubuntu simply run

    apt-get install haskell-platform
  2. Install the development packages for ncurses, i.e.

    apt-get install libncurses5-dev
  3. Install the Haskell dependencies. The Haskell platform comes with a tool called cabal, and you should run the two commands

    cabal update
    cabal install --only-dependencies
  4. Now you can build the program using

    cabal install --bindir=.
  5. At this point, tttool should be ready to go. If you run


    you should see the list of commands shown above.

If you have any problems, you can report an issue via GitHub.

Building your own gme files

Once you have installed tttool, you can create your own .gme files. The process is as follows

  1. Record the audio samples you want to include, as Ogg Vorbis files, mono, 22050Hz. I use

    arecord -r 22050 foo.wav
    oggenc foo.wav
    rm foo.wav
  2. Write a my-book.yaml file containing some general information, and especially the scripts (i.e. what to do) for each OIDs (i.e. the various fields of a book). You can use the example.yaml file as a starting point; it contains more information in its comments.

  3. Run ./tttool assemble my-book.yaml, and make sure it reports no error, i.e. finishes silently.

  4. Copy the now generated my-book.gme to your TipToi pen and enjoy!

If you need to find out what OID code corresponds to what part of the book, you can generate a debug gme using the debug.yaml file, adjusting its Product-Id to your product’s id, building it with ./tttool assemble debug.yaml and loading the resulting debug.gme on your pen. It will then read out the codes, as a sequence of english digits.

We are also collecting template files, where the OIDs are commented; these can be found in the ./templates directory. Please improve and contribute!

Again, please let us know if you have problems, but also tell us what fun things you did if you succeded.

Text to speech

If you have libttspico-utils and vorbis-tools installed, you can have tttool generate audio files from text for you, which makes developing your yaml file much easier. See text2speech.yaml for more information.

Printing your own books

With the code in this repository, you can create GME files. This is even more fun if you can also create your own books! „Pronwan“ found out how that works, as you can see in this video demonstration. He also published 30 minute video tutorial (in German).

Press Review


  • What are all the header fields? (See wip/
  • Finish decoding the games. (See wip/games.txt)
  • What is the purpose of the additional script table?

Other resources in this repository

  • oid-decoder.html allows you to manually decode an OID image.
  • scripts/ downloads all gme files from the Ravensburger server.
  • gameanalyse.c and libtiptoi.c is an alternative tool to investigate gme files. It can also replace audio files in gme files; compile and run it for diagnostic output.
  • Audio/ contains some audio files, such as digits read out.
  • docs/ collects information about TipToi found elsewhere.
  • matlab/ contains scripts to analyse gme files in Matlab
  • wip/ (work in progess) contains notes about the parts of the gme files that are not fully understood yet.
  • perl-tools contains a perl based script, to generate a PDF with all OID codes from a yaml-file as well some functions to generate PNG-files, inject pHYs-chunks with resolution hints into GD generated PNG files as result from some testing