From: Patent Tactics, George Brock-Nannestad
a comment from me at the end of Göran Finnberg's most useful repetition of
the fundamental relationships:
> dB for voltage ratios is:
> 20 log10 V1/V2
> 3 dB = 1.414
> 6 dB = 2
> 10 dB = 3.33
> 20 dB = 10
> 40 dB = 20dB X 20dB = 10 X 10 100
> 26 dB = 20dB X 6 dB = 10 X 2 = 20
> For power ratios:
> 10 log10 V1/V2
> The math fo r replaying a mono disk and summing to mono instead of
> in stereo:
> L&R channel outputs a unit of 1
> Summing 1+1 = 2 = 6 dB in voltage gain for in phase signals.
> The noise, clicks, hiss, crackles, rumble appearing in a horizontal
> direction are random with no phase correllation thus sums to 1.414 in
> voltage level thus increases 3 dB.
> Thus we see a net gain + of 6 dB for the wanted correllated in phase audio
> signal but the uncorrelated noises only sums to + 3 dB.
> So a net gain of + 3dB for the wanted audio signal in respect to the
> unwanted signal.
> Those that dreams about using L or R only listening intently for the
> noise and distortion can keep on dreaming while I just sum to mono as this
> ALWAYS produces the best result.
> Also distortion caused by the pinch effect, rumble, disk warp producing
> massive LF signals etc and other vertical only "distortions" are removed
> completely ONLY IF SUMMED TO MONO.
> One can hear this very clearly if matrixing to M/S and only listen to the
> difference,S, vertical component that only has all the noises and rumble
> with zero music content while the, M, correlated signal contains ONLY the
> best of the wanted content sans all of the dirt that can be removed by
> playing a mono disk.
----- however, for all the good things to happen over the whole frequency
range of interest, the channel separation of the stereo cartridge needs to be
good, and it is usually no more than 20 dB at the best, which is generally at
1000 Hz. One reason the Ortofon mono pickup (invented and produced for a
decade from 1948, and taken up again with unchanged design about 5 years ago)
is that it is a moving coil pickup which is linear even for large amplitudes.
For distortion in a magnetic field, is amplitude, not velocity that counts.