I have hesitated to mention the L word (Loudness) because most people have a strong preconceived picture of what that is at least subjectively, maybe like how it was with subwoofers once, it’s mostly an incomplete view.
From the technical point of view, Mark hit on one key issue that separates live sounds with recorded sounds. Most folks think of loudness like the classic “VU” meter, VU meaning Volume Units fwiw.
That indicator was developed to roughly represent the subjective loudness in the dynamic realm. This IS NOT an indicator of actual loudness, just how loud it sounds.
It works because the meter movement integrates out he peaks to a large degree, the meter shows something like an average over some time. This was needed back in the days of tape where the dynamic range was very limited relative to now with digital, recording at but not above max was important for noise. Unlike digital, tape was not a brick wall at 0dB but could tolerate >0dB levels if short (making the distortion inaudible).
Switching away from the subjective to the operation and technical requirements;
If one looks at the instantaneous voltage from a proper microphone, one finds that many things produce very high but short peak levels. For example, throwing a table spoon on to a tile kitchen floor, produced a peak level of 134dB SPL, being in a car with the windows rolled up and then closing the door produced a peak SPL of 142dB SPL.
Now, these sound like “deafening” levels but because they are short, VERY FEW are aware they exist and instead associate “loudness” with the averaged SPL readings. Now, as Mark suggests, drums are VERY loud AND they do not contain the normal distribution of harmonics but are more like an energy envelope occupying a short time.
To make recordings that “fit” on to a records dynamic range or through AM radio, it was necessary to compress this kind of signal.
With the peak maximum level set by the equipment involved, compressing the dynamic range then makes the drum sound louder because it’s average level is increased. In the extreme, one has the sound of big cymbals on some pop recordings that are SO compressed they actually sound like a high hat or something else entirely.
So, while much music is processed so that it sounds pretty good through ear buds while jogging, even a good home system is often incapable of approaching “live” even when a recording of same is used.
To reproduce something trivial like the spoon hitting the floor, one can say well with Sound level meter set on fast, it produced a peak in the 90’s.
Or, one can say, well, on an oscilloscope, the peak microphone Voltage corresponded to a peak SPL in the mid 130’s.
With one way requiring about 10,000 times more energy and being similarly more difficult, guess which way the industry goes, which kind of system can be added to an elegant room without being noticed.
In the case of a drum, one must also reproduce a wide range of harmonics at the same instant which is an entirely separate "hard task" which gets more difficult the large the acoustic power is required.
Best,
Tom Danley
Danley Sound Labs
Here is a fun link to a "level indicator" which helps illustrate what a difference time makes. Look at your most and least dynamic recordings with it. Remember loudness and dynamics.
http://orban.com/meter/
