Brown noise - Python script - Human Speech - Raspberry Pi Forums
good morning everyone!
i'm developing script plays human voice in spanish, understand how our voice works started 0 , not app.
right i'm playing vowels a, e, i, o , u, different values in main variables, trying different functions in new step want add noise air generated make sounds.
first, i'm creating 1 called white noise instruction success!:
after got random value array need added sound array, taking care on bit depth limits... sounds little good, want try brown noise.
characteristic of brown noise higher frequencies comes lower amplitude white noise. sounds our air when comes out of mouth or waterfall, perfect!. white noise work s & z.
didn't found information of python , noise in web... questions are:
there in python function works low-pass filter? if yes, it?
there way generate random sampling?
taking advantage in time process information... 2 seconds! creating , adding noise increased 12 seconds...
best regards!
i'm developing script plays human voice in spanish, understand how our voice works started 0 , not app.
right i'm playing vowels a, e, i, o , u, different values in main variables, trying different functions in new step want add noise air generated make sounds.
first, i'm creating 1 called white noise instruction success!:
code: select all
speaking_freqlength = int(round(sample_rate/fundamental_frequency)) speaking_resolution = int(round((2**(bit_depth-1)-1)*(1-noise_level))) speaking_noise = int(round((2**(bit_depth-1)-2)*noise_level)) # ---------------------------------------------------- print 'creating noise data' # ---------------------------------------------------- for fx in range(int(round(sample_rate*sound_duration))): speaking_noise_sample[fx] = int(round(random.uniform(-speaking_noise, speaking_noise))) # ------------------------------------------ print 'filling audio array fourier data' # ------------------------------------------ speaking_time_sample = 0 for speaking_loop in range(int(round(fundamental_frequency*sound_duration))): for speaking_wave_form in range(speaking_freqlength): speaking_data_sample[speaking_time_sample][0] = speaking_voice_sample[speaking_wave_form] + speaking_noise_sample[speaking_time_sample] speaking_data_sample[speaking_time_sample][1] = speaking_voice_sample[speaking_wave_form] + speaking_noise_sample[speaking_time_sample] speaking_time_sample = speaking_time_sample + 1
characteristic of brown noise higher frequencies comes lower amplitude white noise. sounds our air when comes out of mouth or waterfall, perfect!. white noise work s & z.
didn't found information of python , noise in web... questions are:
there in python function works low-pass filter? if yes, it?
there way generate random sampling?
taking advantage in time process information... 2 seconds! creating , adding noise increased 12 seconds...
best regards!
have @ scipi , numpy.
http://docs.scipy.org/doc/scipy-0.14.0/ ... ignal.html
i'm not sure if work on pi support kind of complex features looking for.
http://docs.scipy.org/doc/scipy-0.14.0/ ... ignal.html
i'm not sure if work on pi support kind of complex features looking for.
raspberrypi
Comments
Post a Comment