[vorbis] 48000 Hz in vorbis rc3?

[John Morton]

On Tuesday 26 March 2002 08:13, Ben Pearre wrote:
> On Wed, Mar 20, 2002 at 09:30:30PM +0100, Moritz Grimm wrote:
> [munch]
>
> > Apart from that, there are still many people who can't playback music
> > with sample rates higher than 44.1kHz. You'll propably get the best out
> > of resampling to 44.1kHz, in both quality and size - just use a good
> > piece of software to do the resampling with antialiasing filter. Those
> > additional frequencies are pretty much inaudible and matter only in
> > postproduction. [munch]
>
> Not entirely true.  It's well-known that we can't consciously hear
> sustained sine waves above 18KHz or so,

Er... I can hear sustained tones to about 20kHz if the volume is right, and 
there are enough people out there who are a standard deviation or two away 
from the norm who can here out to as much as 25kHz (enough to whine about 
44.1kHz sampling rates being too low, anyway). As for being able to here 
tones above 18kHz when there's plenty of other things going on, that's a 
different story.

> but it's now thought that
> there are other situations in which much higher frequencies make a big
> difference.  Transients?

Cite? I keep hearing of these claims that frequencies higher than 20kHz 
matter, but no explainations as to why.

> The DVD-Audio-type standards sample much higher (96KHz? 192?) just to
> fix this kind of problem.  I haven't heard demos, but people I trust
> fairly well have told me that there's a huge difference, even on
> otherwise normal (admittedly nice) stereos.

I understand that the increase in word size is intended to reduce 
quantization noise rather than increase bandwidth, and increased sample rate 
has the advantage that it makes it much cheaper and easier to build adequate 
low pass filters as there's a lot more frequency range outside of normal 
hearing that can be safely rolled off. Being able to cut out the downsampling 
phase in the mastering process, and having enough audio data to actually fill 
the DVD are probably factors, as well.

John

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.