- Setup using a Jupyter notebook.
- Installed scipy, sounddevice, and libportaudio2.
- set timespace variable t.
- from 0 to 1.
- number of samples to generate (sample rate * frequency).
- use that to generate a sine wave from -1 to 1 using floats.
- Convert to 16 bit signed ints (
sine16 = np.int16(sine_wave * max16)). - Write to
not-clipped.wavusingscipy.io.wavfile.write. - Here is a screenshot of what the
non-clipped.wavlooks like:
- Using
numpy.clip(sine16//2, -8192, 8192)you can quickly create a clipped version. That looks like this and matches the assignment:
- Saved this file as
clipped.wavand the notebook will automatically playback the audio using the python librarysounddevice. - If need be, I can make this into a quick and simple python script, but I
am using a python
venvand Jupyter notebook with arequirements.txt. - This assignment is complete to the specs laid out. There's no further work for me to do, but it does give me some ideas to play with when it comes to generating all kinds of sounds.
- I wonder, is this easier than using tools to get what I want? Just raw creation vs tweaking and adjusting things I know nothing about a la music editing software...?
hw1-clipped
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|
parent directory.. | ||||
