diff --git a/README.md b/README.md
index 16aefdfa1148762e26e20aaf1284776803fea2d1..2d06da9a2397e5f3a5fa81f7497e576b2b50e446 100644
--- a/README.md
+++ b/README.md
@@ -100,3 +100,16 @@ Note that while the sound is played, the interface becomes irresponsive.
 As stated above, the audio player used by pydub require additional libraries.
 If you're using this script on a remote server, the sound might be play over there or not work depending on the method you use to access said server. 
 
+### Output
+
+Once done, the script will save the label in a HdF file (whose name depend on the options used). 
+This file contains a database with one line per annotated click. 
+The column are p1_pos, ipi_sig, ipi_corr_man, ipi_corr_auto, ipi_ceps_man, ipi_ceps_auto, ind_number.
+Note that the individual number is currently not implemented and will only contained `NaN`.
+
+The current behaviour of the scripts also save unannotated clicks. This might be prone to change.
+
+Linked project
+--------------
+
+[IPI_bombyx.py](README_BOMBYX.md) is a script for annotating the IPI of the BOMBYX database.
diff --git a/README_BOMBYX.md b/README_BOMBYX.md
new file mode 100644
index 0000000000000000000000000000000000000000..4d621a1481ae3157ccbe0cc46b9c356cd1b068ef
--- /dev/null
+++ b/README_BOMBYX.md
@@ -0,0 +1,54 @@
+BOMBYX IPI extractor
+====================
+
+`IPI_bombyx.py` is a graphic utility made for the manual extraction of sperm whale inter pulse interval (IPI).
+It builds on the GUI of `ipi_extract.py`. The main features are thus explained in the corresponding [Read me](README.md).
+
+
+Option
+------
+You can obtain a list of all the available option by typing
+`python IPI_bombyx.py -h`
+
+
+Usage
+-----
+
+If you're executing this script on one of the cube, you can simply launch the script by typing
+
+`python path/to/passage_maxpred.pkl your_name_here`
+
+If you're on another server, you will need to set the `--wd` and `--done_file` options to the path of the BOMBYX database, and the shared *done_file.csv* 
+
+### Number of done files
+
+The number of currently files that have been labels is displayed on the bottom right, 
+along with the number of files that have been label by the current user. Note that labeled file that are not finished are not included in the count. 
+
+
+### Next button
+Once you have finished labeling the current file, you can click on the `Next file` button to load and annotate the next file.
+The database will be updated with the labels of the current file, and the shared `done_file.csv` will include the corresponding passage number.
+
+### Closing the window
+If you close the window, the click for the current file will be saved to the database.
+However, the current file will not be considered as a file completely labeled. 
+Thus, the shared `done_file.csv`
+
+### Output file
+
+This script will save the same kind of HdF file as `ipi_extract.py`. 
+However, unlinked `ipi_extract.py` which output one file per annotated file,
+this script will only create one database per annotators. 
+Changing the selling of the annotator will change the database associated.
+
+In addition, of the columns saved in `ipi_extract.py`, this script also save the file, passage number and annotator name.
+
+done_file.csv
+-------------
+
+This file is used to record the progression of the annotation among all the users.
+It contained the passage number of each passage that have been labeled.
+
+
+If you don't know what you are doing, <span style="color:red">__&#x26a0;DO NOT TOUCH OR MODIFY THIS FILE&#x26a0;__</span>.