How ironic.
The question of band limiting or not may had been a wrong question all those decades.
There never was band limiting if we look at production and reproduction not from microphone to loudspeaker. But from the vocal chords of the singer or actor to the listener's brain.
We know, see reference, that correct production and reproduction is not by having flat frequency response from microphone to loudspeaker but flat frequency responce from vocal chords of singer to listener's brain.
This is to a large part explained by the Fletcher-Munson equal loudness curves at the brain of the listener according to what acoustic power enters the ear and voice effort curves of the frequency spectrum produced by the vocal chords of a singer according to what acoustic power is output out of the singer's mouth. These levels could be the same in what is called classical singing but almost always they are very different in pop music singing.
For example: When I Fall in Love - Nat King Cole
https://m.youtube.com/watch?v=91bQyER32GY
If Nat King Cole was recorded with a mic that had a roll off after say 14KHz, this does not mean that higher frequency information is lost. It may be just attenuated a few dB. And these few dB may the a reason why When I fall in Love sounds correct to our brain.
Similarly the low end the spectrum was attenuated. If an RCA ribbon microphone was used, an inductor (Pleiades filter) would have been inside the mic in parallel with its output to cut the proximity bass effect. Electronic engineers might have decreased the inductance of this filter further by adding for example another one in parallel so that more low cut was produced and he could come even closer to the mic. Additional roll off might have been added before or after the preamplifiers, for example with a Pultec so that his voice sounds again correct to the listener's brain, ie flat frequency responce to listener's brain. The microphone may had been an RCA ribbon, a Neumann condenser, or an Electro-Voice moving coil. We may not know but we get the idea.
And this band "limited" signal was enhanced by electron tube preamplifiers adding subtle second harmonics etc giving more magic to the magic of Nat King Cole's voice. And at the same time instantaneously trimming off peaks. So that the average signal can be now increased and be large without distorting many of the lesser equipment that we have on 2017. And so that it sounds great so many decades after on the great medium of YouTube.
To turn the coin upside down...If he had sung to a microphone flat from 20Hz to 20KHz the following things would have happened:
The bass sounding like mud. The high frequencies like spitting. Hiss all over. The extreme dynamic range in watts of the acoustic signal from his voice would produce transient intermodulation distortion on most of the lesser production electronics of today. So the sibilant problems would be magnified by say 100. The signal would still sound small and quiet as it is full of spikes and of not much average level.
We might have to put it to a digital plug in to make it loud. All the sibilant problems and other artifacts would be magnified.
Let's fall in Love again. Or rather let's get up in Love again.
References:
(Flat frequency responce from chords to brain) Motion Picture recording and reproducing characteristics - Loye, Morgan - Journal,of the Society of Motion Pictures
Tube vs Transistors, is there an audible difference? - Russel O. Hamm - Journal of the Audio Engineering Society
Interviews of Mike Oldfield on how he produces his guitar sound (first spectrum limiting it to almost a sinewave, then increase the HF content? then distorting it, then low pass filtering it etc as an extreme example, or papers of Bob Orban on FM Optimod processing for radio stations
No comments:
Post a Comment