Some idiosyncrasies of the program

edfAnn was designed trying to make its use as simple as possible. This implies making some choices and limiting the options of the user.

• The annotations are incorporated to the first annotation signal. If some previous annotation signal exists, it is substituted. If there aren't any annotation signal a new one is created after the ordinary signals. So to add annotations to a file that already contains some annotation signal you must export the annotations, add the new annotations and include the whole set in the EDF+ file.
• The sampling of the annotation signal is adapted to some small space over the length of the data record containing the longest set of annotations. Different lengths created by UTF-8 encoding are considered and the user should not be concerned by this issue.
• Since including annotations implies a deep modification of the structure of the file, the output can not overwrite the original file. A new file is created. You could overwrite the original file from outside the program
• The change of the onsets erases the previous annotations of the signal
• The annotations are sorted previously to be included. So, it is not necessary to sort them. If the annotations are exported and imported posteriorly they become sorted in the process
• The decision of the data record where the annotation is included is handled by the program. So if we export and import the same annotations a different structure could be created
• Time Annotation Lists (several annotations sharing the same time stamp) are broken in the conversion to the text file. Each line of the text file is included as a different annotation even if they share onset and duration
• Annotations with 0 duration are included as point annotations