Award Sounds Music, Photography & Consultancy

Christian, thanks for the great comments.

You've provided an important alternative perspective. My viewpoint is tainted by thinking of the application of signal processing to images and audio files. The processing doesn't necessarily have to change the file, but could be used to view properties of the file, such as assessing how close to black and white is it.

With that perspective, I was trying to think of what properties would mean that an audio file could be classed as equivalent to a black and white image. I had assumed music rather that audio in general. Your mention of soundscapes made me realise my assumption. Good point.

I also like the point about the function of time, I've seen a few spectral analysers that will average out the values across the time and show a "signature" for the song as a whole. I think even the match plug-in in Logic does that to some extent.

I agree with you about texture. That makes me think that using an averaging function would probably remove the differences in texture of sound. How would we be able to tell the difference between the instruments if we couldn't hear the timbre?

The reason it's taken me a couple of days to respond is that I had trouble finding an article. Your response sparked a memory of an article a couple of years ago about passing audio through image processing. I found it again at SoundonSound. Unfortunately it doesn't look like the MonaLisa files are available at the site. Shame.

The only other relevant article I can find is From Magnitude Spectrum to Sound. There must be more out there, but I haven't been able to find any yet.