Nathan Marcus


electronic music | hard- and software

Spectral audio player demo

Spectral Audio Player is a piece of software I developed because the lack of simple audio players that give more details about the file you are playing. Most audio players only display a simple timeline. However, as I work frequently with very long audio files I would like a piece of software that will help me find the important parts of an audio file.

The web version of this software was developed in Chrome and might not work with other browsers.

Loading audio file...
Song title here

This software gives the user more details about the loaded audio file by displaying a spectrogram. This displays the presence of frequencies over time. For example you can tell if there is high or low pitched sound or any sound at all at any point in time. This is very useful for quickly analyzing large audio files as it gives you a simple overview of the content of the loaded audio file.

The most common ways of displaying audio are or simply a timeline or using a waveform.


Both of these ways of displaying don't tell you much about the context of the audio file. Even though the waveform tells you something about the volume of an audio file it does not say anything about the actual content of it. The second screenshot (Audacity) could be containing from an orchestral recording to a field recording of a city.

However, when looking at an spectrogram you can immediately see and tell something about the actual context of an audio file. Below is an example screenshot of a spectrogram. Without listening to this audio file you should be able to tell that it consists of a computer generated sine tone that falls from high to low in about 9 seconds and then travels back again. You can tell that this file is computer generated by the lack of background noise in this file. There is almost no frequency present except the main one.


The example below is a screenshot of a spectrogram of a field recording. There are some interesting details about this file that can be observed immediately. What catches attention is the bar of of high pitched noise throughout the first half of this file (marked yellow) and some short spikes that cover the whole frequency range (marked blue). Through the audio file there is some low pitched sound that is active around approximately the same frequency and sometimes includes some overtones (marked green).


By looking at this spectrogram you can already guess that this low pitched sound might be a voice that is talking or maybe singing. The spectrum-wide spikes will probably some noisy interferance. These spikes are in fact passing motorcycles. The high pitched noise throughout the first half of the file are a group of very active crickets.