From the course: Audio Foundations: EQ and Filters

The perception of frequency and amplitude

From the course: Audio Foundations: EQ and Filters

Start my 1-month free trial

The perception of frequency and amplitude

Together, frequency and amplitude make up the pitch and loudness of the sound we experience. At this point, it's important to note that humans do not perceive amplitude levels at all frequencies equally. That is to say, a 1000 Hz tone played at 80 dB SPL would actually seem louder to us than a 10,000 Hz tone played at 80 DB SPL. Take a listen to this 500 Hz tone. Headphones work best for this example. (music playing) Now listen to this 10k tone played back at the same exact level. (music playing) Notice that the 10K tone, while noticeably higher in pitch, seems a bit quieter. Now, listen to the 500 Hz tone played back to back with the 10K tone. This time the 10K tone will be increased by approximately 10 dB. (music playing) You see, in order to hear the 10K tone at the same perceived level as the 500 Hz tone, it had to be increased by about 10 dB to sound roughly the same. But why? Our human ears are tuned to be most sensitive between 2 and 5K, where the resonance of our ear canal is the strongest. This is by design. Curiously, a human baby's cry is centered around 3500 Hz, and the intelligibility and detail of most speech occupies the same frequency range. Our ability to perceive various amplitude levels at different frequencies is best described by an equal-loudness contour graph. Equal-loudness contours were first measured by Fletcher and Munson in 1933. Because of this, the graphs are commonly referred to as Fletcher-Munson curves. This graph measures the SPL that a tone cycling from lower to higher frequencies would have to be adjusted for a human to perceive that tone at the same level throughout the entire frequency range, just like we heard earlier with the 500 Hz and 10K example. Notice on the graph that lower- and higher-pitched frequencies have to be played back at higher SPL levels to maintain a constant perception of equal loudness, while the dip around our sweet spot of 2 to 5K requires less SPL. It is also important to note that this graph changes as the reference level of the 1K bass tone increases, becoming slightly more even at louder SPLs. This is measured in phons, a scale designed to compensate for the effect of frequency on the perceived loudness of tones. In other words, our ears have a flatter frequency response at louder listening levels, which is why you will often hear engineers discuss the importance of calibrated monitoring environments, where they can achieve a consistent listening level, or SPL from day to day, mix to mix. While I might take advantage of the flatter frequency response my ears provide at 80 plus SPL listening levels, I personally do not hard-fix my monitoring levels at any one SPL for the entire mixing process. For one, I find mixing at that high of an SPL to be especially fatiguing, so I've learned to mix at lower levels, and adapt to changes in my hearing in favor of being able to mix for longer periods of time between breaks. Also, I know the end listener will ultimately be listening at a variety of levels, so I do the same, listening to the mix at both loud and soft SPL levels, and everywhere in between. Understanding the equal-loudness contour and how humans perceive different frequencies at different amplitude levels is important when learning how to use EQs and filters, since we will be using these tools to control and alter the frequency and amplitude relationship of sounds.

Contents