Detecting MP3-Sourced and Mini-Disc Audio – Part 2: Spectral Analysis

A spectral view of a wav file plots the sound frequencies vertically and the timeline of the track horizontally. The intensity of the sound is shown by the colors. Here is an example, once again using Prove It All Night from Winterland Night because it is somewhat complex

One easy tell-tale to look for is if the frequency spikes above 19 khz or 15 (or even less on old mp3s) have been given a nice neat haircut. Instead of being random, they are artificially even, as in this LAME VBR encoded mp3

Note that you can see far more differences than you hear, because MP3 is a “perceptual” encoding system based on what humans actually hear. In fact, in all likelihood you would not be able to hear the differences between the wav in Figure 5 and the mp3-sourced wav used for Figure 6.

It’s not always this clear-cut (;-). Also look for areas where it looks like there are unnatural lines or areas blocked out. A conventional recording should show little or no area that is completely black. For example, compare the 256 CBR mp3-sourced file in Figure 7 to the original wav in Figure 8

(Again, looking at Figure 6 and Figure 7, you might think the track in Figure 7 may sound better because the effect of mp3 compression appears less severe than in Figure 6. However, in fact they are roughly comparable in sound quality – both very good, and the VBR uses fewer bits, which is the point of mp3. These two mp3s would likely each be better or worse than the other in different ways, with the VBR better in almost all of the ways that matter. For example, this kind of graph does not really tell you how well the mp3 is capturing the important mid-range frequencies. So, this only illustrates again that looking at audio is in no way a substitute for listening to it – the purpose of this exercise is to detect source issues, not evaluate quality.)

If you see something like this, it’s easy. Note that the cutoff for frequencies is around 16khz – a substantial difference from the other examples

But remember, you can’t always be sure what accounts for an unusual frequency profile. It may be that the original source has issues. In particular, mini-disc (MD) recordings will show an even more pronounced effect, as they also use a compression approach roughly similar to mp3 but with more compression. The following figure shows an MD recording transferred to wav (but never encoded to mp3) and illustrates the even more severe compression which MD recordings undergo, which affects the frequencies from around 13-14khz and up

And keep in mind there are a lot of possibilities, you have to look at number of files to start to know what they should look like. The differences may have to do with the source of the original recording; whether it was mixed or processed; and so on.

Note that the level of sound will affect the graph display, so keep an eye on the Peak Level number to get a feel for how that affects the display. Be careful with interpreting the colors. It’s not about whether a region is orange or blue, necessarily, but whether a region is empty or has an unnatural look to it.

Scroll to top