This is potentially the most useless post I could make, it has no real use or application beyond satisfying the inner nerd in all of us. BUT, it does (probably) teach us a little about the spectrogram. Now then, it all began earlier today when I was sitting quietly on the computer. I was getting a bit bored of the internet and my thoughts took over. While usually this is a very bad thing, today the thoughts proved benign. My main wondering turned out to be a realisation. That an image could be converted into sound and played back.
So I took to the internet, and decided to learn a little about it. It turns out (thankfully) I'm not the first to think about it and so there was software available to do the job. I also found out that it has been done before, most notably by Aphex Twin, this is a spectrogram of the section of one of their songs:
So how does this work? A spectrogram is a tool to analyse the intensity of different frequencies over time. The x-axis represents time, the y-axis represents frequency. The colour intensity displays the amplitude of that frequency at that time. See below:
What I had to do was get the image coded so that when analysed, it came up with a spectral representation that matched the picture. I used two bits of software for this, Coagula was the software used to convert the image into the sound and Sonic Visualiser was used to view the resulting spectrogram. I found that the y-axis was clearest on a logarithmic scale and so I assumed that Coagula encoded the image with that in mind. My observations of the spectrograms also showed that the bottom was always less clear so with some colour, scale and window adjustment, it was improved upon.
Naturally I wanted to write a post on it and who better to choose as an image subject than Kim Jong Un so this is the final result:
See the video at the top for the sound recording.
I thought I'd leave you all with a cute one.