It is crucial to use studio headphones and monitors with flattest frequency response. These headphones and monitors reduce the variation that listeners hear because of their stereo systems, equalization preferences and the influence from their listening rooms.
These industries have seen billions of dollars and thousands of hours spent on research and development.
Two physicists entered the fray and completely changed the course of events.
Fletcher and Munson disrupted both the recording and home stereo/entertainment system industries in such an impactful way that we still discuss the implications and their consequences 84 years later.
They laid the foundations for international standards in audio-related research and manufacturing.
Let’s start by looking at the history of perceived loudness curves. Next, let’s look at how they impact mixing and mastering workflows to produce better results for artists and fans.
Let’s first define this curve and then look at the history of this fascinating discovery.
What is the Fletcher Munson Curve and how does it work?
The Fletcher Munson Curve was the first of many measurements that became equal loudness contours. These contours are visual diagrams that show the effects of loudness and frequency on the human hearing.
While all the attention has been on the frequency response diaphragms, and other mechanisms that reproduce sound like speakers, no one considered the biological organs that hear sound.
The scientists aforementioned considered the problem to be of higher order and measured how the human ear perceives volume, pitch and relative loudness levels of each frequency range. The first psychoacoustic study he ever conducted was then initiated.
History of Fletcher & Munson’s Loudness Curves
It all started when Harvey Fletcher, the lead researcher, and Wilden A. Munson published the Journal of the Acoustic Society of America titled “Loudness, definition, measurement, and calculation”.
Harvey Fletcher was an American physicist who was born September 11, 1884, and died July 23, 1981. He had a distinguished career in acoustics and television and many other engineering and scientific disciplines. His achievements include:
- The invention of the 2-A microphone
- One of the earliest hearing aids
- Contributing to experiments that won the Nobel Prize for Physics
- Membership in many scientific societies
- He is also known as the father stereophonic sound.
The acoustic research team developed the idea of the equal-loudness curve, which is a grouping where their original curve is only one subset. Both are incorrectly and commonly used interchangeably.
This paper published the methods and results of measuring the sensitivity of human ears in relation to frequency and loudness.
The diagrams and conclusion showed that frequency perception is strongly influenced by the amplitude of each frequency.
These findings reveal that humans are most sensitive to frequencies in the range between 3 kHz – 4 kHz.
This means that all frequencies outside the 3 kHz- 4 kHz range must be at least 10 decibels louder than the average volume to be perceived as being at one volume.
The Fletcher Munson Diagram
Below is a larger image of Fletcher Munson diagram with all the data reported.
The original data was reported on the blue lines. The red lines are the results of Robinson and Dadson’s measurements, which we will discuss next. They eventually became the international equal loudness contour standard.
Both plots are plotted logarithmically using the horizontal access across human hearing’s frequency spectrum. The vertical access shows sound pressure levels ranging from -10 to 130 dB. This allows us to view six sets of measurements simultaneously at 20 dB increments.
As mentioned above, the dip between 3 and 4 kHz corresponds to an increased sensitivity in the human ear within that frequency range. Between 1 kHz to 2 kHz there is a slight increase in sensitivity of between 2 kHz to 2 kHz. However, between 4 kHz- 11 kHz there is a linear increase in sensitivity.
This is called an inverse relationship. It means that these frequencies require less volume to be perceived by humans as the same volume as louder frequencies.
While the curves appear to be very similar above 500Hz, they are quite different in the sub-bass and bass regions.
Although it is now accepted that the Robinson Dadson Curve has a greater accuracy, the original curve remained more in line with all other standards until 2003 when the ISO 226 standard changed.
Although the frequency ranges of sub-bass and bass frequencies have shown discrepancies, no explanation has been found that is satisfactory.
- These regions were not properly calibrated for the measuring devices
- Over the years, the judging criteria has varied for different frequency.
- Before testing, the subjects had been suffering from excessive exposure to low-frequency sounds.
- Race of participants had an impact (inclusion of Japanese data).
A valid experiment that is accepted as an international standard will have tested for listener fatigue. This can be eliminated. Sub-bass frequencies can be felt more than heard. This is partly why the discrepancy was glossed over.
The Robinson Dadson Curve
The ISO 226 current standard is based upon the Robinson Dadson curves, as reported in the British Journal of Applied Physics in a paper titled “A redetermination of equal-loudness relationships for pure tones.” D. W. Robinson’s and R. S. Dadson’s work were surprising, to say the most.
Between the original report and Robinson and Dadson’s, different standards were accepted. The published data of the latter differed more from the current standard and average results of all other results than any other.
The original 40-phon measurement that was used as the basis of the A-weighting standard was a good match for the Robinson Dadson 40 phon measurement. This gave confidence in all subsequent measurements.
Zillow Zillow Home Loans can help you finance your next home.
Transparent process to get the home of your dreams Zillow is the best place to find your dream home. Explore
These are the current standards. The ISO 226 standard is based on Robinson’s and Dadson’s data. However, it has been adjusted using recent assessments made by researchers around the world in 2003. The blue line represents the original ISO standard 40-phons. It is not to be confused the original 40-phon curve.
To everyone’s relief the adjustments brought the data back in line with the original paper, particularly in the bass regions.
This measurement is considered the most precise and widely used by professionals worldwide, despite the fluctuating results in sub-bass and bass frequency ranges.
Equal Loudness Contours
While we have been focusing on the results, how can we measure an equal loudness curve? Because subjective data cannot be objectively compared, we can’t rely solely on listeners’ self-reporting. How is a human hearing curve constructed?
It is a simple method that involves playing two pure tones at different frequencies using sine waves. The volume of the sound waves can be adjusted so that the listener reports when they feel the two are equal.
Participants must be average youths with no hearing impairment. A person cannot participate in the experiment if they have a hearing impairment.
The human auditory system is capable of detecting frequencies between 20 Hz and 22 kHz. Adults can perceive frequencies as low as 20 kHz, whereas young people can only hear 22 kHz.
Researchers began to understand resonances between cavities. They saw that there would be nulls and peaks in the frequency response due the shape and length the ear canal, and the functionality of middle ear. They were correct. A pure tone sine wave shown over one with a more natural timber
The Loudness paper reported the first measurements using a reference tone at 1000 Hz. This tone was then fluctuated in volume until participants claimed it was the same volume of the experimental tone. They did this again for all frequencies of human hearing and across many participants, and then averaged the results.
The absolute threshold of hearing was used. After that, the test was increased by 10 dB to reach the threshold of pain.
In 1937, Churcher and King tried a second set. However, their results were so far from the accepted standard that they were eventually disregarded. Robinson and Dadson published their results in 1956 using the same methodology, but with a more precise and detailed approach.
SIDE VS. FRONTAL PRESENTATION USING LOUDSPEAKERS VS. HEADPHONES
Robinson-Dadson used headphones to communicate with participants, while Fletcher-Munson used loudspeakers to speak directly in front of participants. Two key differences are evident here:
- Frontal versus side presentation of sound
- Distance between the sound source and the human ear
While headphones present sound only from one side, loudspeakers can project sound from a greater distance away from the source. This is because they are able to show sound in a more natural angle due to reflections in the room or in nature.
Although the ISO report states that the team used “compensated headphones”, the differences in bass measurements could not be explained. We are still not sure what this means so it is yet another excuse. The problem of acoustics, removed by headphones or acoustic treatment
Headphones offer a flat frequency response and do not require acoustic treatment. They can avoid all the problems of constructive and destructive interference caused by the room’s box shape.
They avoid any issues related to the shape and position of the head and ear, and how they influence certain frequencies.
No matter what method a research team employs, the other method of listening is still available and used by listeners every single day.
The Fletcher Munson Curve and its Implications
This is the simplest way to reduce all this down into a manageable explanation.
Mixes that are left unaltered with their tonal balance intact will sound different at lower volumes than at higher volumes. The bass and higher frequency ranges will sound much more quiet at lower volumes than the mid-range. High volumes will make bass more prominent, high frequencies may become too forward and the mid-range will remain relatively the same, with some slight variations.
How can a mastering and mixing engineer deal with this problem? What can we do to achieve a pleasing mix of frequencies for our listeners, regardless of the volume at which they are listening to music or movies?
Mixers new to the field will feel frustrated when they have to check their mixes at different volumes. They need to boost and cut and do and un-do the same changes repeatedly, without understanding the effects of the original loudness curves.
It is logical to say, “Well, we should mix at loud volumes because that’s when our listeners are paying the greatest attention anyways.” They can turn the music down if they need background noise.
While it makes sense, you should accept that you will experience ear fatigue and be unable to make sound decisions during long mix sessions. Mixing at a moderate volume is important for most of the progress. It’s best to mix in the 80-85 dB range.
A calibrated mixing level is one that you can always refer to and know better than any other volume. To adjust at a higher volume, we can use our studio monitors and subwoofer.
We will make minor adjustments to faders during this time, but no more than 2 dB to 5. We should pay attention to this time and adjust our equalizers.
Do not keep the volume too high. You can reduce the volume, and then check it again at an average volume to make sure it sounds great. We begin to experience ear fatigue with extended acoustic overload
We should also lower the volume and refer to our mix via our headphones. The bass and upper frequencies should be slightly quieter than they were previously, but they should still sound amazing.
You can make slight adjustments here if you wish. Your volume will increase by two to three.
Do not try to compensate consumers for their actions. They have their own solutions. We buy subwoofers to still hear the bass in our favorite songs or movies at normal volume.
To compensate for low volume, our stereos include bass boost buttons. You can compensate by trying to compensate first. Then there will be too much bass and double compensation.
The goal is to achieve the best possible result in the lowest frequency response environment. Once the audio has been distributed, this minimizes any potential problems.
Mixing the mix may require you to do it again two or three more times, making smaller and smaller changes until you find a balance where the mix sounds great at a medium-to-high volume but still sounds good at a high volume.
But don’t let this drive you crazy. There is no mixer or mastering engineer that can change the way humans hear at different volumes.
You want to reduce the differences by creating a pleasant mix at all volumes. Although the mix may sound different at different volumes, it is still pleasing.
You’ll love this discussion about the RIAA curve. This equalization curve is similar but not directly related to our ears, but to turntables, and vinyl records. The pre-emphasis curve is applied to the record and reversed when it’s played back to cancel each other but reduce noise.
The Fletcher Munson Curve Conclusion
We would never have imagined that our research would need to be taken to a higher level. It is difficult enough to develop and study microphones and speakers in a way that doesn’t affect the frequency response of sounds being recorded or produced by them.
We also have to consider how our brain and ears perceive sound across the frequency spectrum.
We have the advantage of equal-loudness contours that help us to understand the changes needed in our manufacturing processes. We remember and honor the Fletcher Munson Curve.