Richard Hess:

> As to the click energy, Goran, think of this. If the audio goes to
> say 12K and you digitize to 20K that gives you 8K of scratch energy.
> If you digitize to 40K it gives you 28K of scratch energy to use to
> differentiate...but perhaps even a better way of looking at it is
> that the scratch is more faithfully reproduced so it's easier for the
> software to find if you're digitizing with a 40K rather than a 20K bandwidth.

In the NoNoise process from Sonic Solutions one has the option to
prefilter the audio by a declick filter, LP 12 kHz @ 6 dB/octave, or
chose not to.

The recommendation is not to use it as it is stated by Sonic Solutions
that the particular algorithm used can discriminate between music
and/or clicks even better without any filtering.

Having spoken to Dr Moorer of Sonic Solutions he claimed that there were
no particular problem even on bandwith limited material to discriminate
between clicks and music.

And my use of the production declicker in the Sonic NoNoise system bears
him out.

BTW, thereīs a fallacy in what you state Richard.

Most cartridges used acts as LP filters with an electrical cutoff no
higher than 20 kHz, MM cartridges, and those that have a wider
electrical response, MC cartridges, have in most cases seen a mechanical
cutoff at close to 20 kHz which MM cartridges have too.

So even if the higher sampling rate gets you information above 20 kHz it
isnīt coming off the record due to either the electrical filtering
effect of the coils and cabe capacitance giving close to 20 kHz cutoff
at 12 dB octave together with the mechanical cutoff due to the paricular
cartridge use.

In the case of 78 shellac disk reproducers, EMT OFD and Ortofon MC
cartridges, the mechanical resonance is at between 15 to 20 kHz in my
experience rolling off after that at greater than 12 dB/octave.

Shure M44s and others of the same type, MM, will ultimately roll off at
greater than 24 dB/octave so its even less rise time here.

That filtering effect in the cartridge used is THE major low pass filter
rolloff and moving to greater than 44.1 sampling frequency will not make
much difference to the final risetime as digitized and thus will not
make it easier for the software to discriminate between music and

The same argument as above can be said about music recording where most
mics roll off at 12 dB/octave minimum above 12 to 20 kHz, Neumann 67/87
at 14 kHz, so in those cases itīs an absolute waste using anything
higher than 44.1 from a frequency content information view.

So my point is simply that thereīs very little content above 20 khz
coming off mechanical transducers and due to this a greater sampling
frequency than 44.1 kHz will not add anything useful for the software to
chew on.


Goran Finnberg
The Mastering Room AB

E-mail: [log in to unmask]

Learn from the mistakes of others, you can never live long enough to
make them all yourself.    -   John Luther