image2sound

image2sound is a utility that accepts an image file, converts the RGB values of each pixel to a frequency, and saves the result to three separate WAV files.

Requirements

See requirements.txt for specifics

Setup

Create and enter a virtual environment then install the requirements:

python3 -m venv .venv
source .venv/bin/activate
python3 -m pip install -r requirements.txt

To run

Simply running python3 main.py from inside the src folder will generate audio using the test image and default settings. Running it with --help now provides dynamic and more verbose assistance by pulling from the parameter definition files.

Arguments

The following optional arguments may be set:

• -p for a path to an image
• -o for path to save the output file to
• -key for musical key (defaults to C )
• -t for tempo (defaults to 60 bpm)
• -ts to set the time signature (defaults to 4/4)
• -w to use specific waveform types (sine, square, triangle, sawtooth, or piano)

The "clicky" audio found in older versions has now been solved by introducing ADSR envelope filters. This also negates the need to use Blackman smoothing, which is now an option rather than the default behavior. If you prefer to use smoothing, pass --smooth.

• Use -adsr followed by a template type to choose a preset for attack, decay, sustain, and release values. This defaults to piano.

"Reveal" mode

Adding --reveal will override the key, tempo, and movement type with data derived from the image itself, "revealing" the music within the image

"Reveal" mode plus overrides

You can specify arguments as overrides in conjunction with "Reveal" mode. For example, if you want to make sure that the key is D-Major, but you want the other parameters to be derived from the image, run python3 main.py -key D-Major --reveal

Experimental new conversion method

Adding --method2 will utilize an experimental new conversion method that limits the left and right channels to specific frequency ranges, simulating " left-hand" and "right-hand" keyboard parts. Please note that this method does not currently support "Split" mode.

"Quartet Mode"

Using CMYK images will automatically trigger "quartet mode." This will create four separate, mono WAV files with each being limited to the sonic range associated with the four instruments used in a traditional string quartet

MIDI export with Composition Engine

Adding --midi will export the converted audio as a MIDI file.

Differences from the standard conversion method

• MIDI now uses a composition engine which arranges phrases and sections into songs. You can specify a movement type with -mt (sonata is the default).
• It also incorporates variable note lengths. The note lengths are determined by the difference between the color value of the current note in the list and that of the following note.
• Composition styles are defined in movement_definitions.py and new variations can be added there.
• Use of the composition engine means that time parameters passed by the user no longer have any effect since the track length is determined by the movement style and the tempo.
  • Red, cyan, and magenta = violin
  • Blue and yellow = viola
  • Green and black = cello

Examples

Example 1:

python3 main.py -p image.png -key D-minor -t 80

Example 2:

python3 main.py -p image.png --reveal

Example 4:

python3 main.py -p image.png -key G-Major -t 96 -ts 3/4 -adsr cello -w sawtooth

Example 5:

python3 main.py -p image.png -key A-Major -t 80 --midi -mt rondo