On Wed, Apr 21, 2010 at 08:51:22AM +0200, Arnold Krille wrote:
> The highest frequency possible to reproduce -with correct amplitude- is half
> the sampling-rate _only_ if the phase is aligned to the sampling-clock so that
> minima/maxima of the sinus are correctly sampled. If its out of phase, the
> amplitude is not reproduced correctly.
Correct.
> It is easy to understand that this correlation between phase and correct
> amplitude also affects frequencies below half the sampling-rate. Might be as
> low as quarter of the sampling-rate, which in case of the CD is 11kHz. Below
> that you will have more then four samples to reproduce the sinus wave.
> That is in fact another reason to do the recording, mixing and mastering in
> more then 44kHz...
This is popular belief but not correct. Any frequency below FS/2,
not matter how little, is represented exactly and independent of
phase by the samples. If there is any phase dependency that is the
result of imperfect antialias filtering. What happens is that the
signal 'beats' with its alias, and that makes its appear as if its
amplitude depends on the phase.
Its quite unlikely that this would go down in any significant way
down to e.g FS/4. For example if the antialias filter has gain = 0dB
at FS/4, and -60 dB at the alias (a *very bad* filter) the amplitude
variation is less than 0.018 dB (= 1.001 / 0.999).
Ciao,
-- FA O tu, che porte, correndo si ? E guerra e morte ! _______________________________________________ Linux-audio-user mailing list Linux-audio-user@email-addr-hidden http://lists.linuxaudio.org/listinfo/linux-audio-userReceived on Wed Apr 21 12:15:03 2010
This archive was generated by hypermail 2.1.8 : Wed Apr 21 2010 - 12:15:03 EEST