🚀 go-pugleaf

In <2lhpr6$1t4@ugle.unit.no> hafting@Lise.Unit.NO (Helge Hafting) writes:
>In article <xktru*Ey1@dynam.adsp.sub.org>, jmpor@dynam.adsp.sub.org (Jean-Marc Porchet) writes:
> > In article <2laeps$dik@news.cs.tu-berlin.de>, Dirk Kuhlmann writes:
> > 
> > 
> > > 
> > > The problem:
> > > 
> > > Audio data of 8kHz, 8 bit ulaw shall be converted to 44.1kHz 16 bit linear
> > > representattion. There is no DSP support. The conversion is part of a
> > > multi-channel audio mixing driver and therefore has to be as fast as
> > > possible. A FFT is not possible due to timing restrictions.
> > 
> > You can forget to do a "good quality" sample rate converter without a
> > DSP.(I asked our company's signal processing guru and his answer was that
> > there is no clean way that is also a fast way (ie. you need a lot of cycles
> > to do a clean job).

>A friend of mine has written an assembly program that does real-time
>fourier transforms on sound data (transform on approx. 240 data items)
>as the sound is sampled.   His program doesn't go up to 44kHz, but a 486 
>should be able to do more than an amiga500.  So I think it is possible to do 
>such conversion, if you don't need anything more complicated than a fourier
>transform.  It just isn't easy, he used lot of hand-optimizing and
>special mathematical tricks in the program.


Any transform is a bad idea to do such a job.

Go for oversampling and filtering.

-- 
  +-------------------------------------------------------------------+
  | another post from : Jean-Michel MERCIER   jmm73@pern.frmug.fr.net |
  |           PERN : Where the Dragons and the Knights live           | 
  +-------------------------------------------------------------------+
  |  Please, no email >100Ko without permission (I pay hot for it)    |
  +-------------------------------------------------------------------+