Re: 24 bit vs. 16 bit to CD, dithering etc....

makolber@xxxxxxxxx writes:

On Dec 14, 6:05 pm, Randy Yates <ya...@xxxxxxxx> wrote:
Les Cargill <lcarg...@xxxxxxxxxx> writes:
[...]
log2(N)*6 , right?

PS: Pretty darn close, yeah.

  2 ~= 10^(3/10),

therefore

  N ~= (10^(3/10)^log2(N)

and

  20 * log10(N) ~= 20 * log10((10^(3/10)^log2(N))
                 = 20 * (3/10) * log2(N)
                 = 6 * log2(N)

(Had to scratch my head for a few minutes to get this...)
--
%  Randy Yates                  % "Watching all the days go by...    
%% Fuquay-Varina, NC            %  Who are you and who am I?"
%%% 919-577-9882                % 'Mission (A World Record)',
%%%% <ya...@xxxxxxxx>           % *A New World Record*, ELOhttp://www.digitalsignallabs.com

Hi Randy,

I think I disagree with you.. when you record multiple tracks, yes
the noise is uncorreltated but the signals are also uncorreltatd,
usually you record different mics on different tracks.

Hi Mark,

I think it just depends, but, in general, there is going to be some
significant amount of correlation. Definitely not 6 dB every doubling of
the tracks, but something.

It depends on the tracks. For example, I would not expect the drum track
to correlate veriy muchwith the instrument tracks (guitar, keys,
etc.). On the other hand, a trumpet and sax playing the same melody are
going to correlate pretty well, I would expect.

So when two tracks are added the noisees will add 3 dB and the AVERAGE
power of the signals will add 3 dB but the PEAK power of the signals
will add 6 dB.

I have no idea what you're talking about here.

It's not simple...

That's a true statement, and I believe I have been stressing that by
repeatedly stating that IT DEPENDS ON THE SIGNALS.
--
% Randy Yates % "Though you ride on the wheels of tomorrow,
%% Fuquay-Varina, NC % you still wander the fields of your
%%% 919-577-9882 % sorrow."
%%%% <yates@xxxxxxxx> % '21st Century Man', *Time*, ELO
http://www.digitalsignallabs.com
.

Relevant Pages

  • Re: synesthesia Re: Doing The Narrative Drains
    ... i.e. don't correlate with anything else. ... So *if* you had a group of cells specialised for cone ... 1, and another for cone 2, then signals from cone 1 ...
    (rec.arts.sf.composition)
  • Re: Upsampling and Interpolation for TDOA Correlation
    ... as to realize the couple-of-nanoseconds delay between the two arrivals? ... when I construct 1MSps signals and attempt to upsample and ... the signals fail to correlate to the ... estimate fractional-sample time ...
    (comp.dsp)
  • Re: Precise Measurement of Phase Angle Between 2 Signals w/ Noise
    ... i do not understand why error components (what we ... would call "noise") would be correlated. ... This is going to depend a lot on the path the signals have taken. ... The noise here should correlate as much as the desirable parts of each ...
    (comp.dsp)
  • A bit of confusion, xcorr output vs. corr output
    ... It seemed my post from yesterday fell into a black hole.. ... I have two time-series that I cross correlate to find the optimal ... i.e. the delay of one relative to the other where the signals ...
    (comp.soft-sys.matlab)
  • Re: 24 bit vs. 16 bit to CD, dithering etc....
    ... the noise is uncorreltated but the signals are also uncorreltatd, ... usually you record different mics on different tracks. ... tracks are added the noisees will add 3 dB and the AVERAGE power of ... the signals will add 3 dB but the PEAK power of the signals will add 6 ...
    (rec.audio.pro)