Rockbox mail archiveSubject: Re: Faster skip forward
Re: Faster skip forward
From: Mike Soultanian <msoultan_at_csulb.edu>
Date: Tue, 10 Sep 2002 23:48:46 -0700
> spindown timer to perhaps 10 seconds, and see if it helps. Don't worry about
> the spindown timer being high, it only waits 10s after user-initiated disk
> operation, not when loading MP3 data while playing.
I just tried it now and it's the disk read that's slowing the song
transition, not the disk spinning down. When I change tracks, it's
within the first 1-2 seconds of a track so it hasn't even had a chance
to finish filling the buffer.
> > Has anyone thought of
> > loading the first 5 seconds of the track following the next track such
> > that if you hit forward again, it already has the next track loaded.
> That would slow down the disk loading. So, no, we don't want to do that.
If you don't mind me asking, how would this slow down the disk loading?
I'm not sure how clear my idea is, so here's a sample scenario just to
-Hit play on song 1
song 1 starts playing
-loads 5 seconds of song 1 (for current playback) and 5 seconds of song
2 (for anticipated song change)
-after 3 seconds loads more of song 1 (assuming normal playback)
-hit ff within first 5 seconds
song 2 starts playing
-begins playback of already loaded song 2 (from previous read) and reads
5 seconds of song 3
-after 3 seconds loads more of song 2 (assuming normal playback)
-hit ff within the first 5 seconds
song 3 starts playing
-begins playback of already loaded song 3 (from previous read) and reads
5 seconds of song 4
-after 3 seconds loads more of song 3 (assuming normal playback)
user decides to continue playback so it fills the buffer with song 3 as
Now, I'm not sure how many seconds it takes to do the reads, so I just
guessed, but I figure you'd get the general idea. What I'm suggesting
here would, in the perfect world, allow near instantaneous transitions
from song to song. This is probably a somewhat similar idea to reading
the next song before the current song has ended to allow a gapless
transition (assuming there's no recorded silence). Sound crazy?
-- [...]Received on 2002-09-11