/dev/audio
I've spent the last half hour playing around with /dev/audio, the audio device in Linux. Because Linux is based on Unix, it treats some pretty complex things as normal files. The audio device is an example. You can write to it like any other file, but instead of storing the information on disk, it interprets each byte you send it as an amplitude (from 0x00 to 0xff, I guess).
You can then use the `cat` command to pipe other system devices its way. For instance, `cat /dev/urandom > /dev/audio` produces completely random noise. That's not too interesting though. What's better is catting the raw contents of a hard drive (something like `cat /dev/hda1 > /dev/audio`. You hear the data structures pass through in audio form. Different filesystems apparently have their own characteristic sounds.
Also interesting, and probably more useful, is `sudo tcpdump > /dev/audio`. It gives a low growl whenever a packet is sent or received. And you can listen to the system's RAM - `cat /dev/mem > /dev/audio`. Some people have heard that pass through audio files in memory and actually play the songs. ASCII text can be identified because it only uses the last 7 bits of each byte (so the audio is quiet).
If you're lucky enough to have a /dev/mouse (I don't know what my one is), you can try that. Also, super-highly recommended is creating a new bitmap in Gimp, saving it somewhere and then doing:
You can then use the `cat` command to pipe other system devices its way. For instance, `cat /dev/urandom > /dev/audio` produces completely random noise. That's not too interesting though. What's better is catting the raw contents of a hard drive (something like `cat /dev/hda1 > /dev/audio`. You hear the data structures pass through in audio form. Different filesystems apparently have their own characteristic sounds.
Also interesting, and probably more useful, is `sudo tcpdump > /dev/audio`. It gives a low growl whenever a packet is sent or received. And you can listen to the system's RAM - `cat /dev/mem > /dev/audio`. Some people have heard that pass through audio files in memory and actually play the songs. ASCII text can be identified because it only uses the last 7 bits of each byte (so the audio is quiet).
If you're lucky enough to have a /dev/mouse (I don't know what my one is), you can try that. Also, super-highly recommended is creating a new bitmap in Gimp, saving it somewhere and then doing:
while true; do cat whatever.bmp > /dev/audio; done
as you edit it. The size of the bitmap will determine the length of the sample you create. The colours work as you might expect - lots of contrast gives interesting tones. A primary (RGB) colour will give a high pitched tone, and the stronger the colour the stronger the tone. Splitting the image in half horizontally is like working on half the time of the sample. Bands vertically can provide tone. A messy textured image produces a messy, staticish sound. Loads of fun.