veggiedefender / sounds

Turn images into sounds viewable on a spectrogram!

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

sounds

Turn images into sounds viewable on a spectrogram!

The idea isn't that new—Aphex Twin did this in 1999, and people have been doing it manually with whistle/voice since at least 2009.

This program is written using Processing and depends on the processing-sound library, which you will need to install before running.

For visualizations, I played sounds through my PC speakers and used the Spectroid Android app.

Examples

a clip-art chinese dragon a scene from the music video for "Take on Me"
a clip-art dog surprised pikachu meme

Logos in particular end up looking really awesome and dystopian:

Aperture Science's logo from the video game Portal The logo of Overwatch
Enron's logo Fat Llama's logo

Tunable parameters

Parameter Description
timeDelta Milliseconds to spend on each column of the image; scales the image horizontally
resolution Number of vertical "pixels" the image should have. Higher values tend to have more noise
logFreqStart Lower bound, as a power of 10, of the frequency the program can emit
logFreqEnd Upper bound of the frequency

Note that this is built with a log-scaled spectrogram in mind. If yours is linearly scaled, then edit this line to accommodate your needs.

About

Turn images into sounds viewable on a spectrogram!


Languages

Language:Processing 100.0%