Tools for Speech Analysis How do we choose

Tools for Speech Analysis

How do we choose? • What kind of data? • Which task? 2

Data • Speech content (noise, multivoice, …) • Data File – Sound/Transcription/Pitch. Contour – Sampling/Quantization 16 k 12 k 8 k 4 k 8 bit – Size: how much data? – Format • Sound: wav, wma, mp 3, ogg, aiff, aifc, au, vox, raw, sd, CSL, Ogg/Vorbis, NIST/Sphere • Transcription types 3

What tasks do we want to perform ? • Visualization and Editing: – Record, play, edit, mix, add effects • Analysis: – spectral, pitch, intensity • Speech manipulation: – Filtering, mixing, adding effects, prosodic manipulation • Annotation: – segmentation, labeling • Scripting: – Batch, communication with outside 4

Sample Tasks • Create stimuli for an experiment (i. e. hybridization) • Create a database for TTS • Create a prosodic database • Analyze a speech corpus from experiment or ‘real’ recordings • Verify/correct an automatic segmentation or pitch track 5

No Unique Speech Tool • • No piece of software does everything There are usually many ways of doing the thing you want to do 6

Features to Look For • • • Visualization/Edition Analysis Speech manipulation Annotation Scripting Plotting • • • Supported formats Platform/installation Evolution/community Accessibility Price 7

Possible Options • • Goldwave (audio editor) Esps Xwaves (routines + visual. ) Praat (speech analysis) Wavesurfer (speech editor) Transcriber (annotation tool) Matlab (general purpose soft) OGI speech tools (routines + app. dev. ) …winpitch, pitchworks, phonedit, cooledit…. . 8

Links • • • www. goldwave. com www. speech. kth. se/software/#esps www. praat. org www. speech. kth. se/software/#wavesurfer www. cse. ogi. edu/toolkit www. mathworks. com (Matlab) • • www. lpl. univ-aix. fr/~sqlab/ (phonedit) www. sciconrd. com/pworks. htm (Pitch. Works) www. winpitch. com (Win. Pitch) www. adobe. com (Cool. Edit > Audition) 9

Praat • Developed by Paul Boersma and David Weenink at the Institute of Phonetic Sciences, University of Amsterdam • General purpose speech tool : editing, segmentation and labeling, prosodic manipulation 10

11

Praat • Pros: designed for speech analysis (not only sound edition or spectrogram visualization), nice GUI, scripting, active development and community, prosodic manipulation • Cons: limited scripting language, native format of transcription and pitch files 12

File Management • Recording files and saving them – New menu • Opening files – Read menu • Long and short sound files • Other file types – Write menu 13

Editing Options from Objects Window • View – Navigation • Spectrum: spectral slice, spectrogram • Pitch: settings, pitch information • Intensity: settings, intensity information • Formant: display controls, information 14

Modifying the Data • Stylizing the pitch contour: – From Praat objects, Go to manipulation – Edit (the new object) – Pitch stylize pitch (2 st) – Then …. • Modifying pitch • Modifying duration 15

Annotation: Textgrids • From objects – Annotate To textgrid • Labeling • Point vs. interval tiers • NB: remember to select the interval or point first in the waveform or spectrogram before trying to insert a label 16

Scripting • Automatic, from history – Ctrl new Praatscript Edit Paste history – NB: you can run all or part of the script • Writing scripts 17

Help • Online help, FAQ, manual • Links from http: //www. praat. org • Additional tutorials, scripts, resources, user groups 18

Files to Play With • http: //www. cs. columbia. edu/~julia/cs 4706/sound s 19