warss

Wrapper script for ARSS (Analysis & Resynthesis Sound Spectrograph).

ARSS is one of a very few only open-source pieces of software that will let you generate sound based on a spectrogram (image).

Other two I know of are:

• https://www.coppercloudmusic.com/enscribe/
• https://github.com/plurSKI/imageSpectrogram/blob/master/imageSpectrogram.pl

ARSS is pretty neat, but the issue is - usage of this software is not trivial. I have created this script to make it easier for myself, and decided to share it.

Dependancies

• ARSS - download binaries or compile the program yourself (it's a single C file): http://arss.sourceforge.net/download.shtml
• ImageMagick - most Linux distributions will come with it preinstalled, or available in software repositories. This is used to convert any image file you may provide into a format that ARSS will read (as it's only supporting a very specific variant of BMP images).

Usage

1. Drop the arss binary and warss.sh into the same directory as your image file to process
2. Run the script:

./warss.sh image.png

This will create an audio file image.png.wav based on your input.

I am providing sane defaults that result in good quality and fast processing, if you want to tune the synthesis parameters, reference ARSS help:

./arss --help

This will give you the following:

The Analysis & Resynthesis Sound Spectrograph 0.2.3 Usage: arss [options] input_file output_file [options]. Example:

arss -q in.bmp out.wav --noise --min-freq 55 -max 16000 --pps 100 -r 44100 -f 16

--help, -h, /? Display this help --adv-help Display more advanced options --version, -v Display the version of this program --quiet, -q No-prompt mode. Useful for scripting --analysis, -a Analysis mode. Not req. if input file is a .wav --sine, -s Sine synthesis mode --noise, -n Noise synthesis mode --min-freq, -min [real] Minimum frequency in Hertz --max-freq, -max [real] Maximum frequency in Hertz --bpo, -b [real] Frequency resolution in Bands Per Octave --pps, -p [real] Time resolution in Pixels Per Second --height, -y [integer] Specifies the desired height of the spectrogram --width, -x [integer] Specifies the desired width of the spectrogram --sample-rate, -r [integer] Sample rate of the output sound --brightness, -g [real] Almost like gamma : f(x) = x^1/brightness --format-param, -f [integer] Output format option. This is bit-depth for WAV files (8/16/32, default: 32). No option for BMP files.

You can read more on the project's website: http://arss.sourceforge.net/doc.shtml