AnalogSteph
Major Contributor
Apples to oranges or pear-to-pear networking, that is the question.
https://www.lewitt-audio.com/microphones/lct-recording/lct-540s
Now what would have happened if they chose another analysis bandwidth than 1/3 octave for their self-noise? Much the same as when changing FFT order, I assume - larger bandwidth, graph goes up, smaller bandwidth, graph goes down. Which in turn would mean that the graph could be placed almost arbitrarily. In order to line them up correctly, one would need to know what the analysis bandwidth (bin width) for human hearing is - is 1/3 octave really a good approximation?
What do you folks think? I would hate to bother the good folks at Lewitt for nothing and make a fool out of myself.
There is little doubt that this mic with its 4 dB(A) self noise rating is very, very quiet either way (way less than I'd ever need, in all likelihood), but that's not the point. I would normally expect 4 dB(A) to be right around the threshold of audibility, not so decisively below. Fun fact, I do remember reading that when you get into this kind of terrain, the noise of air molecules hitting the diaphragm (Brownian motion and all) ultimately becomes the limiting factor over electronic noise. Pretty wild.
https://www.lewitt-audio.com/microphones/lct-recording/lct-540s
So that on top is the bog-standard hearing threshold curve for sinusoidal stimuli, I assume.How is this mic better
Self-noise values below the threshold of human hearing.
Our hearing threshold is very frequency-dependent. It is less sensitive towards low and very high frequencies and most sensitive around 2 – 5 kHz. With our specific and unique approach towards circuit and capsule design, we were able to realize self-noise that'll stay below the threshold of hearing across the entire frequency spectrum.
Now what would have happened if they chose another analysis bandwidth than 1/3 octave for their self-noise? Much the same as when changing FFT order, I assume - larger bandwidth, graph goes up, smaller bandwidth, graph goes down. Which in turn would mean that the graph could be placed almost arbitrarily. In order to line them up correctly, one would need to know what the analysis bandwidth (bin width) for human hearing is - is 1/3 octave really a good approximation?
What do you folks think? I would hate to bother the good folks at Lewitt for nothing and make a fool out of myself.
There is little doubt that this mic with its 4 dB(A) self noise rating is very, very quiet either way (way less than I'd ever need, in all likelihood), but that's not the point. I would normally expect 4 dB(A) to be right around the threshold of audibility, not so decisively below. Fun fact, I do remember reading that when you get into this kind of terrain, the noise of air molecules hitting the diaphragm (Brownian motion and all) ultimately becomes the limiting factor over electronic noise. Pretty wild.
Last edited: