I am writing this post about the loudness war as from training in mastering audio and from my experience as a live engineer I have noticed that there is a common misconception that Loud = Good. I am not saying that this can't be the case. There are plenty of loud tracks that work well. However, I would like to draw your attention to the good and the bad of what has historically become referred to as 'The Loudness War'. In the 1940's, Jukeboxes started to gain popularity. These Jukeboxes were set to a predetermined level and tracks that were mastered louder than others would found to stand out and to have more success. By the 1950's, producers started requesting louder 7 inch singles so they'd stand out on the radio and mastering engineers started to compete to produce the loudest LP's and vinyl. These requirements led to a further improvement of the tools that were needed for mastering that allowed subjective loudness to be increased. These tools work as follows; Compression; This reduces the dynamic range of an audio signal. This processor starts to restrict the level of sound over a set threshold and at a set ratio. The overall level of the music can then be increased as there is then more room to do so. This often increases the average level of the track and this will make the music appear subjectively louder. Limiting; This is a more extreme version of compression. Same principle, but with a higher ratio of sound reduction applied. EQ; This is the process by which the tonality of the sound is manipulated. So for example, on a simplified consumer system, Bass can be increased or attenuated. On more sophisticated equipment you can increase the subjective loudness by using EQ to pull out a particular resonance and then increasing the overall volume. By the 1990's, the Loudness War was a known phenomenon and was in full swing. In 1995 Oasis released “What's the Story morning Glory' on which their average song reached loudness levels of -8dbFS throughout the track. (db Full Scale – This is a digital scale of loudness, 0dbFS being the loudest possible sound before clipping)
In 1999 the Red Hot Chilli Peppers released 'Californication'. The mastering of this album received criticism on it's mastering as it had noticeable clipping throughout the album. Here is an interesting review of the mastering of this album...
(This article talks about Brickwall limiting; This is where an incredibly high ratio of dynamic reduction is applied to a track. (infinity:1) This prevents any sound from being louder than the given threshold. But, when the overall level is increased the dynamic range is dramatically squashed. This can affect the tonality of the track and can cause added distortion.)
It is interesting to note how both 'What's the Story, morning Glory' and 'Californication' both appear to have set a precedent for the loudness of future of popular music. However, I remember both albums coming out as a young teenager and I was blissfully unaware of any issue with the production. I remember just playing both albums on repeat- so maybe here's a direct example of how the album loudness worked!?
Psychoacoustics Of Loudness
I think that it is worth mentioning Psychoacoustics of sound here. This is the fascinating study of how humans perceive sounds. It's a hard subject to quantify accurately as each human will have their own perspective and different physical make up.
'The Psychoacoustic Problem is that when two identical programs are presented at slightly differing loudness, the louder of the two often appears 'better' but only in the short term of listening' (Bob Katz mastering engineer)
It is easily possible to determine the physical intensity of sound however it is not so easy to determine perceived loudness as it is subjective. There are however trends and the most notable factors in determining how a person perceives loudness are;
Our ears are more sensitive to sounds at different frequencies.
This blurry graph above is called the Fletcher-Munson Curve and it represents our hearing perception level (Measured in Phons) at different loudness' and at different frequency levels.
From this graph we can see that the lower bass frequencies need to be played much louder than other frequencies in order to appear equally loud as other sounds. It also shows that humans are much more sensitive to frequencies between 2 – 5kHz which is interestingly enough at the pitch where a lot of the articulative sounds of speech are placed.
At low listening volumes – mid range frequencies sound more prominent, while the low and high frequency ranges seem to fall into the background.
At high listening volumes – the lows and highs sound more prominent, while the mid range seems comparatively softer.
As the actual loudness changes, the perceived loudness our brains hear will change at a different rate, depending on the frequency.
These findings also tie back in with how EQ can be used to manipulate perceived loudness. Psychoacoustically you could give an impression of loudness just by increasing the highs and lows of the music.
2.Bandwidth (Frequency Span)
Studies have shown that sounds that possess a bigger bandwidth are perceived as louder than narrow bandwidth sounds.
Wide Bandwidth Here is a depiction of a sound with a wide bandwith (Frequency range is shown horizontally on this graph)
Low Bandwidth Here is a sound with a low bandwidth
By compressing a sound you are increasing the perceived bandwidth (Even if it's just by adding noise!) by increasing the lower level frequencies across the spectrum.
'Tones of short duration at a fixed level will not sound as loud as tones of longer duration at the same level. The effect is observed up to a duration of about 200 to 300 msec, beyond which the loudness will remain the same. ' (5)
Compression can also increase the perceived length of a sound by again increasing the loudness of the notes release.
Making Music Louder
One problem with the loudness battle is that by listening to a track without a huge amount of dynamics (Loud and quiet parts) can warp the listeners perception of the music.
Here in Example 1, is an original clip of Gershwin's Rhapsody in Blue;
Below in Example 2 I have reduced the dynamic range of the excerpt by limiting it. It completely changes the musical concept as when the orchestra comes in after the piano solo not far off the same volume it has less impact.
In Example 3 below I have limited Rhapsody In Blue even further and now you can hear parts distorting. The piano at the beginning is now appearing to have less of a solo than the original example as you can hear more clearly the woodwind playing in the background. I hope you can now hear the impact that limiting and reducing the dynamic range can have on a piece of music.
Having said that, I would like to give you an example of a modern style track without limiting.
And here is the same track with some limiting, compression and a little EQ;
I personally find 'Eugene Master' to be punchier and less flat sounding and dare I say.. more exciting to listen to!
So, in this case I would say that reducing the dynamic range has benefitted the music and made for a more pleasant listening experience. Conclusion I think personally that the use of limiting can make certain tracks in certain styles exciting or ear catching. For example if you're dancing in a club you most likely want consistent energy, or if you're playing an advert you want it to stand out to the listener. However, Classical music or Jazz stylistically benefits from having dynamics. So it really is platform dependant and taste dependant. However, having been a live sound engineer for years I can see both sides having worked with bands who want it as loud as possible to get energy back from the audience and then realising that 'loud' doesn't necessarily equate with 'good'.