Runs a process inspired by Alvin Lucier's "I Am Sitting In A Room" for a sound in a room defined by its impulse response.
Essentially, it just applies a convolution reverb over and over again, reinforcing the room's resonant frequencies and smoothing out irregularities present in the original sound.
My voice after being processed for 20 iterations.
python sitting-in-a-room.py {input filename} {IR filename} {options}
-n, -num_passes {int} : The number of iterations to compute. (Default 10)
-l, -level {float} : The amount of "room" you want for each pass. (Default 0.5, or -6dB)
-full : Output all of the iterations side-by-side, much like the real thing.
The two sounds must be in the sound/ directory, and should be 1-channel, 16-bit .wav files, ideally with matching sample rates.
- Get a microphone and some recording software.
- Make a recording of yourself speaking, or making any noise. This is your input sound.
- Place your microphone on the other side of the room and make a recoding of you clapping or hitting something very hard. This is your impulse response.
- Place your two files in the
sound/folder. - Run the script with your sounds!
Or just experiment with any two sounds you think would be interesting, of course.
The example sound provided is from freesound.org by user digitopia.