Rockbox.org home
release
dev builds
extras
themes manual
wiki
device status forums
mailing lists
IRC bugs
patches
dev guide



Rockbox mail archive

Subject: Fw: Re: Iriver: 48kHz Ogg turns on radio?

Fw: Re: Iriver: 48kHz Ogg turns on radio?

From: Pedro Vasconcelos <pbv_at_st-andrews.ac.uk>
Date: 2005-07-16

[ Ryan - I intended sending this to the whole Rockbox list. Sorry for the duplicate reply ]

On Fri, 15 Jul 2005 18:36:52 -0600
Ryan Jackson <rdjackso@gmail.com> wrote:

> http://www.cs.utah.edu/~rdjackso/problem.ogg
>
> This one is actually just a test case. I encoded several tracks from
> different albums as Q10 just to see how they'd perform. Most play
> without any problems. This is one of the few exceptions.

Thanks Ryan. According ogginfo and xmms, this 30s section averages
583kbps with peaks above 600kbps (!!), so it doesn't surprise me that
playback stutters in the current rockbox/ihp combination.

One option that *might* make decoding fast enough for these very high
bitrates is the 'low accuracy' Tremor mode. This uses 8-bit coefficients
for a lot of the tables, reducing memory bandwith at the expense of some
extra shifting operations and (of course) reduced precision of the
decoded stream. (Conjucture: maybe this is what the stock iriver
firmware does?).

However, I don't think it make much sense to use lower acuracy if it
reduces actual sound quality. After all, what's the point of using
Q10 if the compromises required to make it play on the Coldfire hardware
result in WORST sound quality than, say a Q6-7 at full accuracy?

But while this is all conjecture, I have to say that I'd much rather
have be better accuracy at "sensible" quality levels than to strive for
full compliance with extreme cases. For what it is worth, my oggs in
Rockbox sounds much better to my ears than with the iriver firmware.

Pedro

_______________________________________________
http://cool.haxx.se/mailman/listinfo/rockbox
Received on Sat Jul 16 18:37:32 2005


Page was last modified "Jan 10 2012" The Rockbox Crew
aaa