Re: Apple Computer changes its name

[bieling@terra.es]

Jeff Blakeney ha escrito:

> To: bieling
> On Mon, 15 Jan 2007 05:39:10 -0800, "bieling" <bieling@terra.es>
> wrote:
>
> >Let's say that we've got a music file with a bitrate x. We burn this
> >file into a CD (AIFF), then we encode it from the CD's AIFF into
> >another mp3 file with the same bitrate x as the original. The question
> >is, have we really lost any quality in the process? I ask because I'm
> >not so sure... ?
>
> Look at it this way, you start off with the original full quality,
> uncompressed data on a CD.  That is compressed into whatever format
> (AAC in the case of iTunes) which throws away some of the data (almost
> all popular audio compression technologies are lossy) so you get a
> lesser quality but smaller version of the data.
>
> If you uncompress that you still don't get the full quality original
> back but a version that is good enough.  Now, if you compress that
> again, this recompression, which is again lossy, will throw away more
> data so you end up with a lesser quality version of the already lesser
> quality version.
>
> The only way to avoid this is the only get uncompressed data or data
> that has been compressed with a lossless algorithm.  However, I
> believe if you get data with higher bit rates you can lessen this
> effect.
>
> Jeff Blakeney

Yes, mp3 encoding is about throwing data away, but it always looks for
and throws away the same "kind of data". Once you've encoded a file for
the first time, if it's decompressed to an AIFF and encoded again back
to an mp3 (with the same encoding settings as the original mp3), I
don't think that the data to be thrown away was there anymore, so it's
not thrown away and the second encoding's product is of no less quality
than the first one was... At least that's what I think that happens.
(given reasonable values for samples/s, bit depth and bitrate, in every
step of the process)

Jorge.