Monday, August 13, 2012

Fletcher-Munson Equal Loudness Contour Explanation

Hi there!

Recently I saw Peavey refer to one of their powered speakers as having “Automatic equal loudness (Fletcher-Munson) compensation.” That is quite a mouthful of something that probably doesn’t mean much to knuckleheads like me that buy speakers. I have seen other references to equal loudness curves and equal loudness contours, so I thought I would dig into Fletcher and Munson and see what is going on with them.

Fletcher and Munson were Bell labs engineers who performed extensive listening tests in 1933 with the goal of ascertaining how the human ear perceives loudness over the audible spectrum. This was done by having test subjects decide when the tones of two different frequencies were as loud as each other (using headphones, by the way). Listeners were presented with tones at different frequencies and with changes in volume in 10 dB increments (intensity). The listener compared each different frequency and intensity to a 1000 Hz reference tone, which was adjusted until it was perceived to be of the same loudness as the test tone.

Of course, asking people to describe sounds is very subjective, but the researchers made up for this (somewhat) by averaging the results from a large pool of listeners, who were mostly young people with good hearing. As there are not very many average people out there, these results are not exactly rooted in stone, and may not be indicative of how you or I may perceive sound.

Anyway, these averages were plotted to a series of contours (the Fletcher-Munson curves) that graphically show how people perceive loudness of sound by frequency. The lowest equal-loudness contour shows the quietest audible tone (“absolute threshold of hearing”), and the highest contour shows the “threshold of pain.” Which would be a great name for a band. These contours cover the generally accepted range of audible frequencies which are from 20 Hz to approximately 20,000 Hz.

Time does not stand still, and there has been quite a bit more audio research over the past 79 years, so others have built upon Fletcher and Munson’s work. In 1937 Churcher and King presented results that did not exactly follow the original study, and in 1956 a seemingly more accurate study was completed by Robinson and Dadson (do these guys always work in pairs?). They completed the experiment using speakers instead of headphones, so the sound was presented to the ears in a more natural fashion. Their work was the basis for ISO 226, which stood until 2003.

What happened in 2003? The International Organization for Standardization (ISO) revised the standard curves based on the results of studies from research groups in North America, Asia and Europe. The new set of curves is included in ISO 226:2003, and amazingly enough the new curves more closely match the original Fletcher-Munson curves, and not the later and more widely accepted Robinson-Dadson results. Go figure.

Anyway, despite advances from later research they are still called "Fletcher-Munson curves”, which is a lot catchier than ISO 226:2003.


No comments:

Post a Comment