making wav file after beamforming

[gav...@gmail.com]

Hi,

i have a recorded wav file from 2 microphones (someone speaking to one mic and noise in the second mic), converted to h5, also wrote xml mic positions, and used beamforming as in the demo. I get a mapping of the sources.

My question is how to convert the data that have been "beamformed" (using delay-sum-algo) to a wav file, so that i can hear that the noise has been reduced and the speaker is now louder.

Thanks for your help.

[ndimuba...@gmail.com]

Hello,
You can convert a data array into a wav file using Python *wave* library ( in your case the signal array resulting from beamforming summing and scaling).
Assuming you have the "wave" library installed, then the code at the following link would save you data array as a wav file: https://pastebin.com/embed_js/w1YBnfR8

Best regards,
Boris

[gav...@gmail.com]

Hi,

thanks for your answer. I know how to make a wavfile, i use wavfile.write() from scipy.io package. my question was where is the output of the beamforming so i can use this function to make my new wav file. I need a numpy array with the new samples. (BeamformerBase return some pointer..)

In fact i have a bigger issue. I see that no matter what input file i give to work on, the graph will always look the same.

I made 2 files where i talk from different angles, but the output sources seems identical..

Here is my code:

import sys

import numpy as np

import h5py

from os import path

import acoular

from pylab import figure, plot, axis, imshow, colorbar, show

from scipy.io import wavfile

from acoular import __file__ as bpath, td_dir, MicGeom, WNoiseGenerator, PointSource, Mixer, WriteH5

if __name__ == "__main__":

    if len(sys.argv) <= 1:

        print ("You must supply a filename to generate")

        exit(-1)

    for fname in sys.argv[1:]:

        fo = open(fname, "r")

        

arr = np.fromfile(fo, dtype=float, count=-1, sep=',')

arr = np.resize(arr, (len(arr)/2,2) )

max_value = np.max(arr)

arr = (arr/max_value) - 0.5

hdf = h5py.File('data2.h5', 'w')

time_data = hdf.create_dataset('time_data', data=arr, dtype='float32')

time_data.attrs['CLASS'] = 'EARRAY'

time_data.attrs['EXTDIM'] = '0'

time_data.attrs['TITLE '] = ''

time_data.attrs['VERSION'] = '1.1'

time_data.attrs.create('sample_freq' , '20000.0',None,float)

hdf.attrs.__setitem__('CLASS', 'GROUP')

hdf.attrs.__setitem__('PYTABLES_FORMAT_VERSION', '2.1')

hdf.attrs.__setitem__('TITLE' , '')

hdf.attrs.__setitem__('VERSION', '1.0')

hdf.close()

hdf = h5py.File('data2.h5', 'r')

micgeofile = path.join(path.split(acoular.__file__)[0],'xml','array_2.xml')

mg = acoular.MicGeom( from_file=micgeofile )

ts = acoular.TimeSamples( name='data2.h5' )

ps = acoular.PowerSpectra( time_data=ts, block_size=128, window='Hanning' )

rg = acoular.RectGrid( x_min=-0.7, x_max=0.7, y_min=-0.7, y_max=0.7, z=0.7, increment=0.01 )

bb = acoular.BeamformerBase( freq_data=ps, grid=rg, mpos=mg )

pm = bb.synthetic( 250, 1 )

Lm = acoular.L_p( pm )

imshow( Lm.T, origin='lower', vmin=Lm.max()-20, extent=rg.extend(), \

interpolation='bicubic')

colorbar()

figure(2)

plot(mg.mpos[0],mg.mpos[1],'o')

axis('equal')

show()

hdf.close()

here are my sample file (test.txt - mic1[sample], mic2[sample]..etc) and my xml file (2 mics). https://files.fm/u/9rnqqsmm

In this test.txt i said "left" , "right" , "center"

i run python myfile.py test.txt

please help me !

thanks a lot

[samantha...@gmail.com]

hi,

sorry,but how can i convert one file wav in a file .h5????

thanks for your help

sammy

[carlos.p...@gmail.com]

Hi, I think I try the same, but have the problem, how get samples filtered after beamforming?

"where is the output of the beamforming so i can use this function to make my new wav file. I need a numpy array with the new samples. (BeamformerBase return some pointer..)"

Thanks!

[stephe...@gmail.com]

Hello -- maybe you can help me on my problem? 

I have collected a field recording (birds, vehicles, etc.) using a 4-microphone array, and would like to map the sounds using acoular, which requires an h5 file.

I am very new to Python -- starting with wav files & array positions, how do I get the data into an h5 file?

Thank you!

Steve Caird