In the early 1930’s, two researchers at Bell Labs named Harvey Fletcher and W.A. Munson uncovered perhaps the most significant psychoacoustic phenomenon in history. They found that the frequency of a sound affects our perception of its loudness. In other words, two different frequencies at the same mechanical level may not sound the same level to us because of the way our hearing mechanism processes them. In the researcher’s experiments, they played dozens of tones for listeners against a 1 kHz tone at the same level then asked the listeners what seemed louder.

 

DIFFERENCE BETWEEN LEVEL AND LOUDNESS

 

These two tones are equal in level (same dB). iZotope Ozone’s EQ reads equal peak levels. Yet, the higher pitched tone seems much louder.    

 

ELoud Spec Eq Lev

 

 

 

The two tones below are the same tones from above but are equal in perceived loudness (same Phons).  Although their peaks are nearly 20 dB apart, they sound equally loud because of hearing sensitivity. I simply attenuated the higher tone by about 20 dB with a bell filter and bounced the output into a new analysis so you could see.

 

ELoud. Spec Equ Loud

 

 

 

ANALYSIS

 

After analyzing the data, the researchers came up with the fletcher munson curves – a graphical representation of what levels sound EQUALLY LOUD for all frequencies and stages of loudness. This graph has since been perfected using results from a bigger pool of testers and renamed – the equal loudness curves.

 

ELoud. Graph

 

How do I read this graph?

 

Frequency is on the X axis. Level (dB SPL) is on the Y axis. The lines represent equal loudness across the spectrum for different ranges of loudness. Phons is a unit of loudness – not to be confused with dB, which measures level. The higher a frequency falls on the curve, the more gain it will need to sound as loud as 1 kHz at the same mechanical level – and visa versa. Equal loudness curves are inversely proportional to how our ears treat input. If our ears were an equalizer and you were to flip these curves upside-down, the settings on the equalizer would match the upside-down curves.

Equal loudness curves reflect the settings you would need on an EQ to in order to counteract what our ears do. If our ears were an equalizer and you were to flip these curves upside-down, the settings on  that equalizer would match the upside-down curves.

 

For Instance…

 

The curves tell us that 250 Hz at 50 dB sounds as loud as 2 kHz at 38 dB. I know this because these are the levels the two tones line up with when on the same loudness curve. Look at the picture below… REMEMBER – THE CURVES REPRESENT EQUAL LOUDNESS

 

ELoud. Graph Drawn 1

 

They tells us that 63 Hz at 100dB sounds as loud as 4 kHz at 80dB. This example relates directly to and validates my audio samples in the introduction of this article.

 

ELoud. Graph Drawn 2

 

So what is this REALLY telling me?

 

  1. It validates the cliche, EARS OVER EYES – WHAT YOU’RE SEEING IS NOT ACTUALLY WHAT YOU’RE HEARING. This is why very balanced mixes tend to have a visually unbalanced spectrum – often a downward slope because low frequencies need more energy than higher frequencies to produce the same loudness.
  2. At low levels, WE ARE NOT VERY SENSITIVE TO FREQUENCIES BELOW 250 Hz and are VERY SENSITIVE TO FREQUENCIES BETWEEN 1 and 8 kHz.
  3. FREQUENCIES ARE PERCEIVED MORE EQUALLY THE LOUDER THE SOUND GETS. This is one reason all music is more effective louder – the low and high ends are not rolled off as much. Make sure you’re doing the majority of proper mixing at at least 80dB but no more than 100dB. Listening too loud causes a phenomena called temporary threshold shift (TTS) – where the outer hair cells in your hearing mechanism  naturally dampening the basilar membrane to protect you, throwing off your sense of balance.
  4. BOOSTING MIDS CAN HELP ADD WEIGHT TO THE MIX WHEN IT IS PLAYED AT LOWER LEVELS.
  5. A MIX THAT SOUNDS GOOD AT LOW LEVELS WILL PROBABLY SOUND GREAT LOUD – BUT, NOT THE OTHER WAY AROUND