Scripts

Praat scripts

Most used

Sp_ToBI

Eti-ToBI

An automatic transcriber for Spanish and Catalan intonation that uses the Sp_ToBI and Cat_ToBI conventions. It works better (95% accuracy) with Peninsular Spanish.

Sp_ToBI picture

Create pictures v.4

This script creates and saves pictures (PNG, PDF, wmf, eps, PraatPic) of all the sound files it finds in a folder. The pictures contain a waveform, a spectrogram, an optional F0 track and opcionally the content of the tiers of the TextGrid associated to the sound file.

Mandarin Tones

Mandarin transcriber

This is a first beta. Needs a wav sound and a textgrid with intervals. It can serve as a helper for transcription but its performance is far from perfect.

Data analysis

point_label_extraction: Works for all the files in a folder. Writes in a txt every point label and it writes its correspondent interval. In order to run it needs a TextGrid with ar least 1 interval tier and 1 point tier.
extracts_f0_from_points: Works for all the files in a folder. Writes in a txt point labels, time of point and F0 on point. Needs a .wav and textgrid with at least 1 point tier. The output is a tab separated file.
extracts_vowel_data: Works for all the files in a folder. Extracts F0, F1, F2, F3, F4, duration and intensity for all intervals that have a label with a vowel symbol.  Needs a TextGrid with an interval tier.
prosodic_data_extraction.praat v.2.1: Extracts duration of each non-empty interval, intensity and 3 F0 (Hz and st) values for each non-empty interval it finds in the tier. It can  also extract the corresponding labels for . Then it saves the values in a tab-separated txt, which you can easily open with Excel.
zero-crossings-and-spectral-moments: Script created for analysing fricatives. This script goes through all the files in a folder and gets (for non-empty-intervals or intervals that match a label) zero-crossings and spectral moments.
ltas_data: Writes Ltas values from all files in a folder in a txt that can be opened with Excel.
voice quality: Performs voice report for all wav files in folder. Writes jitter, shimmer, HNR, voice breaks.
save_perception_results: This script extracts the results of all the perception test that are in the objects list, then sorts them by stimulus and saves a txt  file with the name you indicate.

Editing, manipulation and sound segmenters

Edit sound files

This script runs through all the files in a folder and allows you to convert the stereo sounds into mono, resample the files, normalize intensity and remove noise. BEWARE! It removes noise everytime you use it, and it uses standard Praat noise removal: your sound can become metallic. Use under your responsability

Edit sound files

Modifies duration of a selected Sound neutralizes_pitch: Manipulates pitch in order to be a constante value of X Hertzs.

Modifies duration

Modifies duration of a selected Sound neutralizes_pitch: Manipulates pitch in order to be a constante value of X Hertzs.

nextract_and_save_intervals

This script extracts and saves intervals in an independent sound file, the new files will have the same name/label as the intervals.

TextGrid annotation and modification for all the files in a folder

blank_TextGrid_creation: Creates TextGrids for all the Sounds in a folder.

created_TextGrid_modification: Opens a created TextGrid with its related sound, gives you time to correct it and saves the TextGrid with the modifications.

segmentation_helper: This script goes through all the wavs in a folder and annotates them (sound by sound in IPA). It needs the ortographic transcription of the wav content. It will write the same label (sentence) in the TextGrid for all the wavs in the folder. If you want to annotate a corpus that has a different content depending on the file name, check Mietta Lennes scripts, SPASS or EasyAlign.

remove_tiers: This script runs through all the Textgrids in a folder and removes the number of tiers you select and those below it

Other tools

list_files: Writes a list with all *wav or/and  *TextGrid that finds in a folder.

 

Utilities for Amper members

segmentador.praat: Automatic segmenter for the Amper's project fixed corpus. It identifies sentences out of a large .wav sound and extracts it sentence with its corresponding code. In order to use it, you need to have a txt with the codes in the order that you have recorded them.

copy_Amper_folder_without_expansion_files.praat: This script copies all the files belonging to the sentences without expansions from the "permanent" Amper corpus.

Python scripts

Corpus tutorials: Two seminar sessions for introducing Python to Applied Languages students.

AMPER Forensic: The script suggests a series of geo-prosodic affiliations for the unknown speaker, arranged in order of likelihood. AMPER_Forensic has been tested with AMPER data from Italian, Friulan and Catalan. It only works with AMPER-based database files.

R scripts

Amper_txt_Hz_to_St.R: This script goes through all the Amper files in a folder and converts the F0 in Hertz data in St data.

If you want to get Matlab scripts (Amper06 or ProDis), contact us.