Adding speaker information

How to add information about a speaker to TEVA.

Record or Open an /a/ sound, select a stable part of the recording:

Adding information about a recording

TEVA can keep a record of recordings. Click the Speaker button at the bottom left of the Main page. A window will open where you can enter an identifier (name or code) and other information like sex, age, treatment etc. There is also a field for more informal comments. The top text field is an identifier for the recording, the next lower one contains information about the speaker, ID, name, and so on. The bottom text field is for comments. Click the Continue button when you have completed the input. You can change the text later if you like.

Reading a table with speaker information

In general, it is better to collect speaker information in a table beforehand. You can open such a table with the Open button in the Main page, or with the Data button in the Configuration page. When opened, you will have to select a specific speaker or recording from the list. This is done using the Speaker button on the Main page. In the top text field, you can type in the ID code of the recording, or the line number in the table, and click Ready. Or you can step through all the records with the next (>) and previous (<) buttons. If you enter a partial ID, it will be matched to the start of the ID's in the table. The first record that matches will be selected. If you enter a non-existing recording ID in the top, ID, field, a new record is added to the table. You can remove the current record by completely emptying the top, record ID field, and replace it with a single dash, -.

You can save changes to the list with the Save button in the Configuration page. You can close and purge the current table with the Close button in the Configuration page. With Close, the current table is not saved. However, all changes made in an existing speaker table will be saved in a backup file with the same name as the original file, but with a tilde (~) added to the name (e.g., example.Table becomes example~.Table). This backup file will be kept until the table is saved to file using the Save button or all changes are purged with the Close button, after which the backup will be removed. The backup file will be overwritten the next time the table is opened from file and a change is made. When using Save to save a list, an attempt wil be made to convert all paths to audio files to paths relative to the position of the saved table.

Format of the speaker info table (.tsv or .Table)

Speaker Info tables are tab delimited (tsv) lists with five fields, starting with a line with the field names separated by tabs, i.e., ID, Text, Description, Audio, AST (the order is immaterial). The extension of the file should be .tsv or .Table.

1: ID of speaker or recording (must be unique)
2: Essential information, often starting with the ID code
3: Free form comments
4: Relative path to the audio file
5: Manually entered Acoustic Signal Type, i.e., 1, 2, 3, or 4
6: Start time of interval for which the AST was entered
7: End time of interval for which the AST was entered
8+: Any number of values for rating scales

If the table contains a path to a sound file, this file will be opened automatically when the record is chosen

Missing columns are automatically generated when the table is read. So, if a table without the Description and AST columns is read, two empty columns with the corresponding labels are created.

An example file: SignaltypeVoiceSamples.Table

ID Text Description Audio AST StartTime EndTime
Speaker1 Speaker1, M 66yo, Type I [comments] signaltype1voicesample.wav 1 0.000 1.750
Speaker2 Speaker2, M 48yo, Type II [comments] signaltype2voicesample.wav 2 1.000 2.750
Speaker3 Speaker3, F 77yo, Type III [comments] signaltype3voicesample.wav 3 0.500 2.250
Speaker4 Speaker4, F 48yo, Type IV [comments] signaltype4voicesample.wav 4 1.250 3.000

A text-only file without the .tsv or .Table extension will be read as a list of records separated by tabs. The order of the fields is like above, ID Text Description Audio AST StartTime EndTime. For instance, if three fields are given, they will be entered as ID, Text, Description. If only a single field is given, it is treated as the Audio field. The ID field will be set to numbered Item[row] values if not present. Upon reading, such a file will be converted to a a full table and saved (if backups are set, ~.tsv). Backup files too will be saved as ~.tsv files.


Links to this page

© R.J.J.H. van Son, December 20, 2011