Subject: Re: [linux-audio-dev] 24 bits, 2's complement - help!
From: Benno Senoner (sbenno_AT_gardena.net)
Date: ma tammi 17 2000 - 05:12:26 EST
On Sun, 16 Jan 2000, Bill Gribble wrote:
> Juhana Sadeharju <kouhia_AT_nic.funet.fi> writes:
> > Tables are probably no no; specially if used in driver.
>
> I agree. Dithering should not be done in the driver at all, I think.
> If an application wants to support high-quality word-length reduction
> using RPDF dither, noise-shaped dither, or any other fancy technique,
> that application should open the audio device in the
> higher-word-length mode and perform the dithering at the user level.
>
> It would be a big, big plus for Linux pro-audio developers to have an
> open-source library of high-quality word-length reduction algorithms
> handy. In commercial systems, the availablity of dither is a
> significant price booster and the availability of multiple dither PDFs
> is (while perfectly useful and easy to do) only something you see in
> the really high-end (with a few exceptions, like the Symetrix 620).
>
> Bill Gribble
yea, an opensource dithering/shaping library would be great,
do you know if there are some papers around the net which speak about
high quality dithering/noise-shaping ?
But explain me please why tables are a no-no ?
if you generate a few secs of white noise, and let the noise play in a loop,
you can't almost not detect that there is a loop, you should hear
a continuous white noise.
The dithering would be reduced to a mere table lookup and a sum.
I don't know a way faster than this ..
Benno.
This archive was generated by hypermail 2b28 : pe maalis 10 2000 - 07:23:26 EST