Music frequency distribution is a crucial factor in speaker development. It's pointless to design speakers for frequencies that don't exist in music, but equally important to ensure that all essential frequencies are covered. While the concept seems straightforward, comprehensive statistics on music frequency distribution are surprisingly scarce. This blog mentions some frequencies that rarely occur in music often. Good enauch for blog, but a more robust data source is needed for speaker development.

The Technical University of Berlin has undertaken a commendable project and recorded 22 frequently used musical instruments to perfectly map their sound image (all notes, all frequencies, all directions). This is not exactly the same as the frequency distribution in music, because there are certainly more and less used notes there as well. However, it is even better input, because it also covers those frequency ranges that are rarely needed in music. As good a guide as this is for a speaker builder, it is also for a speaker buyer.

Before going into frequency, we also set a scale to look at. A 6 dB decrease is a fourfold decrease in sound intensity considered a significant change in sound pressure. A 40 dB decrease in sound pressure is a 100-fold decrease in sound intensity, considered inaudible compared to the main sound.

Tuba, double bass, and drums were chosen from the instruments for bass. Accordingly, -6 dB requires frequencies of 24, 31, and 62 Hz. The last one also includes guitar, but the following instruments are already above 100 Hz. Looking at this picture, the bass of a normal speaker starts at 60 Hz, a good speaker at 30 Hz, and a great speaker at 24 Hz. There is nothing very surprising about the bass and the previous knowledge was confirmed.

The midrange notes must all be covered until we reach the end of the straight section at about 5 kHz. From there, the drop begins and -6 dB is at 8 kHz. Now the sound pressure drops very quickly and at 17-18 kHz we have already reached the -40 dB level. This leads to two guidelines. Firstly, a very good speaker can also be one whose frequency range ends before 20 kHz. Secondly, the need for high sound levels also decreases with increasing frequency, meaning that the power required for loudspeakers at high frequencies is negligible at the upper end of the frequency scale.

It is surprising that there is a uniform lack of higher harmonics. Regardless of whether it is a string instrument, wind instrument or human voice, frequencies above 10 kHz disappear very quickly. This uniform drop also gives a better idea of why speakers are often analyzed up to 10 kHz in comparisons and the last octave is mostly ignored.

If you are one of those who recently bought a speaker with a supertweeter that can reproduce sound up to 40 kHz, you might be disappointed to see this data. It seems like it was a pointless investment. Maybe, or maybe not. Here too, the devil and the angel are in the details. More on this in the next blog post.