#1 title
some background information
I'm an EE. specialized in speech and language processing.
7yrs in geneva, sw (psycholing and comp ling lab)
1yr at AT&T, NJ

i'll start with an overview of some speech analysis tools.
then we'll do together some manipulations on speech with a software I like, called Praat

#2 two questions
important to choose your software

#3
content:
where was is recorded? street, room, booth (echo, unwanted sound)
which material ? quality, number of channels
size: memory disk is now cheap, but not RAM. must be well designed

#4which taks
*analysis: spectrogram, spectrum, wavelets, spectrum tilt, various methods of pitch tracking, formant, intensity, duration (phone,word,pause),  breathe
*manipulation: band filtering (low-band filter, high-band filter, stop band , pass band), inverse filtering (Linear Predictive Coding) and get the residual/source signal (glottal signal)
*annotation: chunck turns, words, phonemes . auto methods exists

#5examples of tasks
*stimu: edition, maybe analysis for details
*TTS-db : correct labeling, normalize intensity
*prosoDB: save many IP curves in a format
*record: digital transfer from a DAT to computer, trim useless parts,make all the possible analysis

#6two rules
you cannot chose ONE tool for life
stay open to other tools (new ones or improvment)
don't spend time on redoing what another tool do