From: Stuart Tedford <stuart.tedford_at_piresearch.co.uk>
Date: Mon, 13 Jan 2003 18:01:23 -0000

The "buffer time" can be thought of as the estimated play time left in the
buffer before another disk read is done. So if the time is long, the disk
is read well before the buffer is empty. If the time is too short, the disk
might be read too late, so the buffer could become empty before new data is
inserted from the disk, which causes a skip.

The optimum time is therefore the shortest time you can enter before you get
skips due to not reading the disk before the buffer is completely empty.
The problem is that the disk read time depends on many external factors so
the "buffer time" has to be a compromise based on your circumstanses.

Stu.


> -----Original Message-----
> From: Bryce Benton [mailto:me_at_brycebenton.com]
> Sent: 13 January 2003 17:30
> To: rockbox_at_cool.haxx.se
> Subject: Re: anti-skip buffer question
>
>
>
> can you (or someone) be more specific?
>
> do i understand correctly: longer buffer time means less battery
> consumption because the disk doesn't have to spin up as often?
>
> if that's true, my question remains... in what cases would
> someone want
> less than the maximum buffer time?
>
> thanks,
> bryce
>
> > > how does the anti-skip buffer affect performance?
> > >
> > > or... is there any reason to select less than the longest time for
> >this
> > > setting?
> >
> >Eats more battery, since it reduces the time between disk
> reads by the
> >buffer amount.
> >
> >v
>
>
Received on 2003-01-13