CODEWITHSUNDEEP
pythonnumpyaudiowav
Dr.Cappo
Dr.Cappo

Reputation: 1

How to took middle time range from wav . Python

I need to extract exaclly 8 seconds from the middle of the audio data from the wav. file with lenght 0:27 sec.

--All what I did already, it took the middle 9 sec by divided wav. file on 3 parts and took the middle one, but it's 9s I need 8s.

And how find a number of bits in that numpy array?

import scipy.io.wavfile
import pyaudio
import numpy as np

(samplRate,data)=scipy.io.wavfile.read('Track48.wav')

print

CHANNELS=2

p= pyaudio.PyAudio()
                        #
nine_sec=len(data)/3
eight_sec=2*len(data)/3

stream = p.open(format=pyaudio.paInt16,
                         channels=CHANNELS,
                         rate=44100,
                         output=True
                         )

cuted_data=data[nine_sec:eight_sec]

newdata = cuted_data.astype(np.int16).tostring()
stream.write(newdata)

print(cuted_data)

Thank you for your help.

Upvotes: 0

Views: 2263

Answers (2)

ephraimbuddy
ephraimbuddy

Reputation: 331

I know this question is old but someone may like the solution i want to offer which uses numpy only. No need of pydub.

import scipy.io.wavfile as wavfile

fs, data - wavefile.read("Track48.wav")

# number of samples N
N = data.shape[0]

# Convert seconds to samples
eight_secs_in_samples = float(fs)*8 # time = number_of_samples / rate

midpoint_sample = N//2 # Midpoint of sample

# substract 4 seconds from midpoint
left_side = midpoint_sample-(eight_secs_in_samples//2) 

# Add 4 seconds from midpoint up
right_side = midpoint_sample + (eight_secs_in_samples//2)

# The midpoint of samples is therefore:
mid8secs = array_data[int(left_side):int(right_side)] # this range contains the required samples
# Save the file
wavfile.write("eightSecSlice.wav",fs,mid8secs)

Upvotes: 0

Anil_M
Anil_M

Reputation: 11473

You can use pydub to slice middle 8 seconds very easily. Details on pydub are here
And you can install as pip install pydub

I had a wav file of 348 sec duration whose middle 8 seconds are sliced.

>>> song.duration_seconds
348.05551020408166

You can also use different file formats such as wav, mp3, m4a, ogg etc. for import (convert to data-segments) and export.

Source Code 

from pydub import AudioSegment
from pydub.playback import play

song = AudioSegment.from_wav("music.wav")

#slice middle eight seconds of audio
midpoint = song.duration_seconds // 2

left_four_seconds = (midpoint - 4) * 1000 #pydub workds in milliseconds
right_four_seconds = (midpoint + 4) * 1000 #pydub workds in milliseconds

eight_sec_slice = song[left_four_seconds:right_four_seconds ]

#Play slice
play(eight_sec_slice )

#or save to file
eight_sec_slice.export("eight_sec_slice.wav", format="wav")

As you can see length of middle 8 seconds slice is exactly as desired.

>>> eight_sec_slice.duration_seconds
8.0

Upvotes: 2

Related Questions