Re: The Difference Betweeen 96khz & 192khz

"Tim Martin" <tim2718281@xxxxxxxxxxxx> wrote in message
news:OikZe.14506$wm3.2555@xxxxxxxxxxxxxxxxxxxx
> "Arny Krueger" <arnyk@xxxxxxxxxx> wrote in message
> news:Fuadnfpd4_wEDKjenZ2dnUVZ_smdnZ2d@xxxxxxxxxxxxxx
>
>> If you want to compare different sample rates, that's all
>> that should change.
>
> I think I disagree.

> Suppose you want to digitize audio in the frequency range
> up to 20kHz.

> If you digitize with a sampling rate of 48000, you need
> to use a filter which stops all frequencies above
> 24000Hz, but allows through all frequencies below 20000Hz.

> However, if you digitize with a sampling rate of
> 192000Hz, you meed a frequency which stops all
> frequencies above 96000Hz, but allows through all
> frequencies below 20000Hz.

> This is, in principle, an easier filter to build; and
> it's possible that may result in the quality of the
> filtered analog (pre-digiization) audio being better for
> the 192000Hz sampling than it is for the 48000Hz
> sampling.

No, the low-pass filter for 96 KHz sampling is generally no
easier nor harder to build than the one for 44 KHz sampling.
In fact its often the very same filter!

For about the last 10-15 years the filter in question is
*always* implemented as a digital filter. In most cases
this filter implemented as a filter that has very flat
frequency and ideal phase response up to about 0.95 of the
Nyquist frequency, and also has tons of rejection at the
Nyquist frequency and above. If you operate this filter at
any reasonble frequency it does what is needed.

It is true that some people have thought that a more gentle
slope from 20 KHz to 0.95 of the Nyquist frequency for
higher frequencies might be a neat thing. This has been
implemented as a digital filter, as well. The design of
these filters is stored as sets of parameters on the
converter chip. It's up to the designer of the converter
product to pick which parameter set he wants to use when
multiple parameters are used.

IOW, its entirely feasible to design a converter that
operates very nicely at any sample rate from say 44,100 to
192,000 without any changes other than changing the clock
rate.

> So I think there are two separate questions:

> 1) do devices sampling at 192000Hz result in better
> digital files than devices sampling at 48000Hz? To
> answer this question, you need to use different sampling
> devices, as well as different sampling rates.

This hinges on how you define *better*. I don't know a
global definition for *better* but I do know a globally
applicable means for testing equipment to see if it sounds
different.

The question at hand is - does a good job of sampling at
192,000 Hz sound any different doing a good job of sampling
at 44,100? Quite a bit of evidence has been carefully
collected, and the answer seems to be: "No". Most if not all
contrary reports seem to be due to lack of appropriate care
while collecting data.

> 2) if you sample at 192000Hz, does it sound better if you
> store and play back the data at 192000Hz compared with
> downsampling to 48000Hz and storing and playing back at
> 48000Hz? To answer ths question, only one device is
> needed.

I think we can figure this one by inductive reasoning. If a
good job of sampling at 44,100 Hz sounds no different than
doing a good job of sampling at 96,000 Hz, increasing the
upper sample rate to 192,000 is probably not going to sound
any different.

> I think in practice your comment addresses the second
> question, but not the first.

If the converters are performing ideally or at least well
and consistently, then both situations should sound sound
the same.


.

Relevant Pages

  • Re: Hellp: Sampling logic
    ... I have a sound blaster 16 which I use for capturing data on everything ... is the sampling performed evenly over a 1 ... The sample rate is always even, ... of the sample rate clock, but the problem then is that it is very hard ...
    (comp.sys.ibm.pc.soundcard.tech)
  • Re: Envelope Detector using Hilbert Transform
    ... >>> I have working with ultrasound signals from a solid state ... >>> Khz) are digitized through a sampling card and imported into MATLAB, ... >>> implementing it in real time on the DSP processor. ... >>fit for the filter to make the passband ripple small. ...
    (comp.dsp)
  • Re: Anti-Aliasing filter
    ... The antialiasing filter problem basically goes away with a sigma delta converter since the actual sampling rate is very much higher than the effective sample rate. ...
    (comp.dsp)
  • Re: Digital processing of analog TV broadcast
    ... FPGA clock - fixed and stable. ... I regenerate the chroma subcarrier using a numerically controlled ... numerically controlled oscillator, and low-pass filter. ... carefully plotting out the sampling components and their aliases. ...
    (sci.electronics.design)
  • Re: The Difference Betweeen 96khz & 192khz
    ... it's possible that may result in the quality of the filtered analog audio being better for the 192000Hz sampling than it is for the 48000Hz sampling. ... the low-pass filter for 96 KHz sampling is generally no easier nor harder to build than the one for 44 KHz sampling. ... IOW, its entirely feasible to design a converter that operates very nicely at any sample rate from say 44,100 to 192,000 without any changes other than changing the clock rate. ... If a good job of sampling at 44,100 Hz sounds no different than doing a good job of sampling at 96,000 Hz, increasing the upper sample rate to 192,000 is probably not going to sound any different. ...
    (rec.audio.pro)
Loading