Collection of texts from Tράπεζα κειμένων with synced audio for learning Modern Greek.
Audio is from various contributors from the Learning Greek Discord server.
-
Install Node.js.
-
Install dependencies:
npm install
-
Start web server:
npm start
-
Navigate to the URL outputted by step #1 in your web browser of choice.
- Decide on a text to add from the Τράπεζα κειμένων Google Drive folder.
- Find the text on Tράπεζα κειμένων (the URL may be used in step #3).
- Find a suitable image for the text by following instructions under How to Add an Image.
- Run
./add-new-text.js
and answer the prompts. This will create a directory for the new text in the root of this repository with a beginningconfig.json
file. - Create a
transcript.vtt
file by following instructions under How to Make WebVTT From Audio.
To add an image for a text:
- Search for an image with a creative commons license. For example, using Google image search. The orientation should be landscape instead of portrait (wider than tall).
- To reduce file size:
- If the image is large (> 1000px), then shrink and / or crop it using a free online service such as Image Resizer.
- Compress the image using a free online service such as TinyJPG.
How to Make WebVTT from Audio
-
Download the audio for a text from the Τράπεζα κειμένων Google Drive folder.
-
Download and install Audacity.
-
Open audio file in Audacity, and select Label Sounds from the menu bar (Analyze -> Label Sounds...).
-
Experiment with settings to achieve a good result.
- Set "Minimum silence duration" to 100 ms.
- Experiment with Maximum leading silence and Minimum trailing silence. (For example, 200 ms maximum leading silence and 100 ms minimum trailing silence)
-
Update labels with corresponding text by copying text chunks from the page on Tράπεζα κειμένων. Adjust labels if desired or needed.
-
When happy with the result, export the labels as text (File > Export > Export Labels....).
-
Convert the exported labels to VTT by running:
./audacity-labels-to-vtt.js path/to/audacity/labels.txt > ./path/to/transcript.vtt
For example:
./audacity-labels.to-vtt ./Τσιντιλά-audacity-labels.txt > ./Τσιντιλά/transcript.vtt