Just as a facemask can hide the identity of its wearer, so can a sound wave hide the identity of another. Masking is a psycho-acoustic phenomenon that occurs when one sound conceals another sound. The effects of masking can render a sound completely inaudible, but more likely, a certain element to the sound will change.
For example an instrument may sound superb in isolation, but when combined with others in the recording mix it might lack warmth, body, clarity etc. This article aims to illustrate the basics of the phenomenon and offer several methods that may reduce the problem.
The frequency range of the human ear extends from 20Hz – 20,000Hz (20kHz) depending on the individual. This spectrum can be divided into approximately 24 sections, each with a limited bandwidth (Zwicker, 1961). These sections, known as Critical Bands, are traditionally modelled as a series of overlapping bandpass filters (or sections) as pictured in Fig.1. (Fletcher, 1940). Each band is capable of hearing a limited amount of information – basically reducing the total information processed by the brain – and when two signals fall into the same band, the softer signal is likely to be masked to some degree.
The existence of masking can be demonstrated easily by listening to two separate sine waves that fall within one critical band. For example create a sine wave of 450Hz using a signal generator and set the volume to a comfortable listening level (which incidentally is quite low for sine waves). Then create a second sine wave of 470Hz at a level approximately 30dB less than the first. Each on its own can be heard without difficulty, but when combined, the second is inaudible.
If the second sine wave is raised in pitch so that it instead falls within the next critical band, it instantly becomes audible and we can hear the two sine waves as two distinct and separate auditory events.
In popular music there are many degrees by which masking may affect the sound, and often the conflict is unavoidable. Essentially, masking is likely to occur whenever two or more instruments with similar frequency content are blended together. Classic confrontations include the kick drum vs. bass guitar or a rhythm piano vs. a rhythm guitar.
Taking the piano vs. guitar example, if they play in a similar register, they will have many frequencies in common, causing masking mayhem. One possible solution is to use a dedicated psycho-acoustic processor such as the SPL Vitalizer. This device adds a slight time delay to the loudest sounds in order to unmask the softer sounds. Another way to alleviate the problem is to use an equaliser to raise or lower the frequencies in the problematic area of the signal. This could be either to increase the amplitude of soft signals that are being masked, or conversely to decrease the louder frequencies that cause the masking to occur.
Since these solutions are essentially ‘Band-Aids’, each with their own set of disadvantages, the best approach is to avoid the problem in the first place. For instance, we could alternatively place the sounds in different areas of the stereo field using the pan pot. The loss effect will be subdued because the instruments come from different speakers. Of course, panning is sometimes undesirable or mono compatibility may be relevant.
Perhaps the best possible solution is to look at the instrumentation and compositional aspects of the music. To avoid similar frequencies fighting for the same critical band, it is possible to separate the instruments using pitch and timbre.
Imagine two guitars playing identical parts. It will sound thicker as the same frequencies combine and reinforce, but it’s hard to distinguish what each guitar contributes to the overall sound. As soon as one of the guitars is played in a higher register then it becomes easy to tell the two guitar parts from one another.
When writing music, consider the frequency response created by the instrumentation used and aim to fill the spectrum across all the critical bands. For example, in a given song perhaps a harmony part would be better suited to the song if performed by a woodwind rather than a bowed instrument (or vice versa), or perhaps performing the part a perfect fifth higher would create space.
Ultimately, to achieve a piece of music that is perceived as being full-bodied with clarity requires careful use of instrumentation and composition. Consider your goals, but remember the foundation needs to be set before tools such as the Vitalizer or equalisers are even considered. Try experimenting with different chord voicings, playing in different registers and explore the art of arrangement before taking a swim in the muddy waters of masking.
Reuben Rowntree is the Diploma Program Co-ordinator at SAE Institute in Parnell. He also offers various courses for SAE Online specialising in mastering and music production using Apple Logic, as well as lecturing part time at Auckland University’s School of Music. He can be reached at firstname.lastname@example.org.