From: Michael O'Quinn <michael_at_oquinn.info>
Date: Wed, 30 Jul 2003 14:01:07 -0700 (PDT)

On Wed, 30 Jul 2003, TP Diffenbach wrote:

> Quoting Michael O'Quinn <michael_at_oquinn.info>:
>
> > On Wed, 30 Jul 2003, TP Diffenbach wrote:
> > > What I'd really like is to have three testers per model (FM, Recorder,
> > > Player) each of whom tests one of the given bit rates:
> > >
> > > 128 160 192
> >
> > Since you are being so exhaustive, you might also want to do something
> > really low -- say 32 or 48 -- for the audio book folks. FWIW, that one
> > should show the largest improvement.
> >
> > Michael
>
> Thank you for volunteering, Michael! I didn't want to assign that, as I
> suspect it could take a nice long time. But I'll be glad to hear your
> results!

Sure, no problem.

Is 2007 gonna be O.K.?

>
> Here's the important thing: the test needs to be done using constant bit
> rate MP3s. So get out your ruipping software.

Hhhmmm...

Why is that?

I would think that as long it's the same file being played over and over,
and that as long as you don't make it too long (say not over 5 minutes)
then it really shouldn't matter, since that one file will be played over
and over and over and over and over and... The VBR differences will be
averaged out, and I would expect that you are going to see at least 5
minutes variance between (seemingly) identical runs anyway.

OTOH, if this means you are *also* volunteering to supply the actual music
files themselves, who am *I* to argue? Besides, that'll save wear and
tear on *my* music files.

Michael
Received on 2003-07-30