Marinos
you can believe what ever you wish
But my admittedly very (maybe, over) simple statement is for sure 100% true=
.
But of course in the "real World" of actual A-D converters this implies an=
input filter with an infinite "slope". IE impossible, & for sure even a ver=
y
hi Q (or slope) filter will almost for sure sound pretty rough.
Oversampeling techniques to a great degree overcomes this limitation ( IE a=
s
I say what is happening between samples is in no way a guess it is known
fact, so it is quite possible to Up-sample your say 44.1k samples to any
amount you wish then filter it with a decent sounding filter, & all the
derived values will be very, dam near correct. IE near zero aliases (aka
dodgy values).
This done well is IMO the way to go, but another approach (with great
marketing & BS possibilities) to sample at an elevated sample rate and
apply the same near correct nice sounding anti-imaging filters).
John L
" <>
To: <>
Sent: Thursday, August 20, 2009 7:23 PM
Subject: Re: [Nature Recordists] Re: High Sample Rates
well
as I told I am not sure whether this is the case or not... :-)
On 20 =CE=91=CF=85=CE=B3 2009, at 6:16 =CE=9C=CE=9C, John Lundsten wrote:
> Marinos Koutsomichalis wrote
> when recording @ 44,1, the machine will measure
> 44100 times per second the amplitude of the signal and
> it will store these values somewhere. During playback these values
> will be converted back to an analogue that WILL NOT be the same due
> to the fact that only if we take infinite measurements we can have a
> true digital representation of the signal. That is because has
> absolutely no clue of what happens between those 44100 samples.
>
> This is quite simply not true. Assuming the sound that is sampled
> 44.1k/sec
> has NO sine wave components greater than 22.05kHz then it is totally
> predictable what happens between each measurement a sine wave is Very
> predictable.
>
>
|