Re: [ASD-devel] Two things..

[Lennart P. <poe...@us...>]

On Wed, 15.05.02 18:52, Sam Clegg (sa...@su...) wrote:

> On Wed, May 15, 2002 at 06:34:53PM +0200, Lennart Poettering wrote:
> > > In fact, perhaps we should rename the library to 'libpcm'.
> >=20
> > This is a good idea.=20
> >=20
> > What is the exact definition of "pulse code modulation"? Do uLaw and
> > aLaw fall under this notion? If this is not the case, libpcm might not
> > be the ideal name.
>=20
> Well, technically PCM is a simple linear scale and ulaw and alaw are
> non-linear.  Do we need to support ulaw and alaw?  If so libpcm is
> not correct.   ulaw and alaw don't show up much these days.

I'd like to see ulaw/alaw support in the libary since it is still used
widely. (e.g. sun sparcs, isdn telephones, .au-files) And, the most
important reason: it is nearly no work for us to implement this, only
an additional entry in an enumeration is needed - in distinction to
ADPCM based sample formats. (And the Java Sound API supports uLaw as
well) The name "libpcm" should not force us to limit the support to
real PCM formats. I think the name "libpcm" is warrantable, alltough
it is an "Etikettenschwindel".

lp

PS: ding translates the german word "Etikettenschwindel" with "bogus
claim" or "fraud"

--=20
name { Lennart Poettering } mail { poe...@us... }
    icq { 11060553 } loc { hamburg, germany } pgp { 0x1A015CC4 }
sourceforge.net { asd, battstat, befound, libsample, libshbuf, ivam }