Wavefile IO

Nsound includes a simple Wavefile class that can read RIFF wavefiles that are PCM encoded. This is a very basic wavefile format. Nsound currently does not support any other audio formats, but there are plenty of other free tools on the web that can convert audio formats to PCM wavefiles.

Reading Wavefiles

As seen in previous sections, a wavefile can be read by passing its filename to the Buffer or AudioStream constructor:

import Nsound as ns

b = ns.Buffer("california.wav")
a = ns.AudioStream("california.wav")

Nsound has also overloaded the left shift operator, concatenation from wavefiles is easy:

import Nsound as ns

b = ns.Buffer()
b << "california.wav" \
  << "walle.wav"

Writing Wavefiles

Warning

Values outside the range of -1.0 and 1.0 will get clipped in the output wavefile. To avoid this, call the function normalize() on the Buffer or AudioStream.

The Buffer and AudioStream class includes a function to write wavefiles to disk.

Buffer.writeWavefile(filename)

AudioStream.writeWavefile(filename)

The Wavefile class is used by Buffer and AudioStream to perform the actual writing. Note: the Buffer class does not know the sample rate of the data contained in it. The Wavefile class uses a default sample rate setting. The Wavefile default sample rate and sample size can be changed by calling:

Wavefile.setDefaultSampleRate(sample_rate)

Wavefile.setDefaultSampleSize(sample_size)

Where sample_size is 8, 16, 24, 32 or 64.

When an AudioStream writes to disk, the sample rate is known by the AudioStream. Like the Buffer class, the sample size is not known. Use the Wavefile.setDefaultSampleSize() to set this option:

import Nsound as ns

a = ns.AudioStream("california.wav")

ns.Wavefile.setDefaultSampleSize(32)

a.writeWavefile("california-32bit.wav")

Nsound has also overloaded the right shift operator for writing wavefiles, the example above can be rewritten as:

import Nsound as ns

a = ns.AudioStream("california.wav")

ns.Wavefile.setDefaultSampleSize(32)

a >> "california-32bit.wav"

Nsound will write wavefiles using integer data types by default. This can be changed to floating point.

Floating Point Wavefiles

Warning

Few programs know how to read the IEEE floating point format. Audacity 1.3.12 is able to read mono floating point formats, but does not seem to be able to read multi channel files.

Note

Since the data is stored as floating point, there is no need to normalize the data first.

Nsound can read and write wavefiles that store their samples in IEEE floating point format. To write files using this format, call:

Wavefile.setIEEEFloat(flag)

And set the flag to True. The Wavefile sample size must be set to 32 or 64.

Resampling Wavefiles

Sometimes a wavefile won’t be at the sample rate that we wish it were. For example, you downloaded a sample sound but it is not at the sample rate of your project.

With Nsound it’s easy to resample the file to the sample rate you need:

import Nsound as ns

a = ns.AudioStream("wav_at_16KHz.wav")

a.resample2(48000.0)

a >> "wav_at_48KHz.wav"

The Python script below can be used on the command line to change a wavefile’s sample rate:

#! /usr/bin/env python

import Nsound as ns

from optparse import OptionParser

parser = OptionParser(
    usage = "resample target_sample_rate input.wav output.wav")

(options, argv) = parser.parse_args()

argc = len(argv)

if argc != 3:
    raise RuntimeException("Expecting 3 arguments!")

target = float(argv[0])
f1 = argv[1]
f2 = argv[2]

print "Reading %s" % f1

a1 = ns.AudioStream(f1)
source = a1.getSampleRate()

print "source: %d" %(source)
print "target: %d" %(target)

ratio = target / source

print "ratio:  %f" %(ratio)

print "Resampling ..."

a2 = a1.getResample(ratio)
a2.setSampleRate(int(target))

print "Writing %s" % f2

a2 >> f2