cliveb wrote:
> eCo Wrote: 
> 
>>Pat...normalization is apparently not the same in this software.  I've
>>been assured there is no expansion or compression when it is used.
> 
> I am 99.99999% certain that Pat has not suggested that normalisation
> will involve compression. He understands fully what the implications
> are. The simple fact of the matter is that if you try to apply a linear
> amplitude adjustment to a digital recording, then unless the adjustment
> factor is an exact power of 2, you *will* incur some degree of rounding
> error. This is nothing to do with compression: it is in principle
> *impossible* for any software to perform normalisation that avoids
> rounding errors.
> 
> That said, let's be pragmatic about the issue. The audibility of the
> rounding errors that are introduced by normalisation is vanishingly
> small compared to the much grosser effect of actually changing the
> volume. I have over a decade of experience in digitising LPs, and I
> would be very surprised to hear that anyone was able to successfully
> ABX normalised v. unnormalised recordings of an LP (after level
> matching, of course). Nobody should lose sleep over it: in practice,
> normalisation is a benign operation.

Pat is right.

Normalisation involves scaling the values that represent the waveform by
a simple process of multiplication.

I think the confusion has arisen because eCO said:

"Normalizing in my software simply adds a nominal gain throughout
the recording"

That is also true, if eCO is tlaking about the addition of a dB value.

dB use logarithmic scale, relative to a reference value, e.g.:

dB = 20 log10(V/V0), where V0 is the reference value.

The additional of a "nominal gain" in dB is equivalent the multiplying
the actual voltage values by a constant.

Now, the waveform is AC, i.e. there are +ve and -ve values as well as
some 0 values; so, multiplying by a constant has the effect of
stretching the waveform about the zero axis.

Here's a simple example:

Input:  4 2 0 -2 -4
Gain:   2
Output: 8 4 0 -4 -8

HTH,

R.

_______________________________________________
audiophiles mailing list
[email protected]
http://lists.slimdevices.com/lists/listinfo/audiophiles

Reply via email to