[linux-audio-user] Re: 192kHz

fons adriaensen fons.adriaensen at skynet.be
Sat Jan 28 10:49:12 EST 2006

On Sat, Jan 28, 2006 at 02:11:50PM +0100, Wolfgang Woehl wrote:

> Ok, filter quality. Esben, Fons, on another aspect of 
> samplerates higher than 48k: Is it possible that what is 
> audible from an orchestra for example stems in part from 
> interference or intermodulation of harmonics from above the 
> audible band? Relevant for the reproduction had the 
> performance been recorded to discrete channels?
> I don't know how to phrase my question better. Gene said Yes 
> to that if there was "something non-linear in the mixing 
> process". I didn't understand that though.

There are certainly instruments that produce sound well 
above 20 kHz. Now to get 'beats' between two of those 
requires something non-linear - just summing them is not
enough. Our ears are non-linear at high sound levels, so
in a real performance this could happen. But I don't
think it will be noticeable at any sane sound levels. 

As to non-linearities in the mixing process, that is
something that should *not* happen. A properly dithered AD
or DA conversion has *no* non-lineraties resulting from the
quantisation (until you drive it into saturation).
Even at 16 bits it's difficult to get audible intermodulation
from the quantisation, unless you are working with pure
sinewaves. But some cheap equipment shows very clear
non-linear effects, often from the badly designed analog
parts, or in the early days of digital, from the converters.
It's a long time ago, but I remember having once tested a 
AD converter that had large 'gaps' - some ranges of values
were never used. 


More information about the Linux-audio-user mailing list