dev2.praat

# Start with getting...
# 1. mean of intensity in final section of IPUs, section lasting 200 ms/300 ms/500 ms/1000 ms DONE
# 2. mean of pitch in ^^ DONE
# 3. python for CSV of IPU_Stops that preced a Smooth Switch
#       PRAAT SCRIPT NEEDS TO KNOW WHICH WAV FILE TO SELECT  >>DONEE
#       AND missing Defining a Smooth Switch in the Python
# 4. jitter, etc., in ^^
# 5. pitch slope in ^^

# CSV files...
#   IPO Task for Quiet and 70 (2 total)
#   FTO Task for Quiet PositiveFTO and 70 Positive FTO (2 total)


# desired Praat CSV output file architecture
# 1. create separate CSVs for each of mean intensity, mean pitch, jitter, etc.
#   do something like: init last_filename = first name, get current_filename, for loop through all rows but if last != current use current filename
# 2. create new columns as appropriate, eg. mean in last 200 ms, last 300 ms, etc., for each CSV
#   add those to the right of WAV filename , Original CSV Row Num , Condition , Type , Group , Replicate, Talker, IPU_Duration , IPU_Stop


# Input CSV file format, preprocessed with a Python script
#
# Columns in input CSV:
#  1. WAV_Filename: the corresponding audio recording that each IPU belongs to
#  2. Original_IPU_Row
#  3. Condition
#  4. Type
#  5. Group
#  6. Replicate
#  7. Talker
#  8. IPU_Start
#  9. IPU_Stop
# 10. IPU_Duration
# 11. NumSyll
# 12. LeveldB
# 11. isOverlapWithin
# 13. Post_IPU_FTO: the FTO that follows each IPU, as obtained from the FTO CSV
# 14. Original_FTO_Row
# 15. isTargetIPU: whether the IPU in this row is one we should analyze
# 16. postIPU_Switch_Type: the nature of the switch, either "Smooth" or "Hesitant"



#clear old messages in the info window 
clearinfo


# select files
form IntensityExtraction
    sentence directory "C:/Users/Ludens/Desktop/data/"
    sentence wav_file "brownfox.wav"
    sentence csv_file "q211fox.csv"
endform

sound1Path$ = 'directory$' + 'wav_file$'
csv1Path$ = 'directory$' + 'csv_file$'


# create Sound Object
sound1 = Read from file: sound1Path$
sampling_rate = 22050


# read CSV file
csv1 = Read Table from comma-separated file: csv1Path$


# store in an array the IPU ending times
ipu_ends# = Get all numbers in column... IPU_Stop


# define IPU-final sections to be analyzed
sectionDuration = 0.2
sectionEnd# = zero# (size (ipu_ends#))
sectionStart# = zero# (size (ipu_ends#))
for i from 1 to size (ipu_ends#)
    # Define the start time for a final 200 ms section
    sectionEnd# [i] = ipu_ends# [i]
    sectionStart# [i] = sectionEnd# [i] - sectionDuration
    if sectionStart# [i] < sectionDuration
        sectionStart# [i] = 0
    endif
endfor



# 1. create the Intensity Object
selectObject: sound1

intensity_pitchFloor = 75
intensity_timeStep = 1/sampling_rate

intensityObj = To Intensity: intensity_pitchFloor, intensity_timeStep

selectObject: intensityObj

# 1.1 obtain the mean intensities in the IPU-final sections
for i from 1 to size (ipu_ends#)
    # make into a ##
    meanIntensity = Get mean: sectionStart# [i], sectionEnd# [i], "energy"

    # print to console for now, but put into CSV
    # appendInfoLine: "1: ", sectionStart# [i], "   2: ", sectionEnd# [i], "   MEAN INTENSITY: ", meanIntensity
endfor



# 2. create the Pitch Object
selectObject: sound1

# filtered autocorrelation is recommended by Praat documentation for measuring intontation
pitch_timeStep = 1/sampling_rate
pitch_floor = 50
pitch_top = 800
pitch_maxCandidates = 15
pitch_veryAcc$ = "no"
pitch_attenAtTop = 0.03
pitch_silenceThres = 0.09
pitch_voicingThres = 0.5
pitch_octaveCost = 0.055
pitch_octaveJumpCost = 0.35
pitch_voicedUnvoicedCost = 0.14

pitchObject = To Pitch (filtered autocorrelation): pitch_timeStep,
                                                    ...pitch_floor,
                                                    ...pitch_top,
                                                    ...pitch_maxCandidates,
                                                    ...pitch_veryAcc$,
                                                    ...pitch_attenAtTop,
                                                    ...pitch_silenceThres,
                                                    ...pitch_voicingThres,
                                                    ...pitch_octaveCost,
                                                    ...pitch_octaveJumpCost,
                                                    ...pitch_voicedUnvoicedCost

# 2.1 obtain the mean pitch in the IPU-final sections
meanPitchUnit$ = "Hertz"
for i from 1 to size (ipu_ends#)
    # make into a ##
    meanPitch = Get mean: sectionStart# [i], sectionEnd# [i], meanPitchUnit$

    # print to console for now, but put into CSV
    appendInfoLine: "1: ", sectionStart# [i], "   2: ", sectionEnd# [i], "   MEAN PITCH: ", meanPitch
endfor