From: Michael Sevakis <jethead71_at_sbcglobal.net>
Date: Tue, 24 Feb 2009 19:51:04 -0500

> I've come to the conclusion that, without preserving decoder state
> information, maybe that's as good as it's going to get. There just
> doesn't appear to be sufficient information to construct faithful audio
> from an isolated initial mp3 frame.
>
> I'll probably just file a low-priority bug report at leave it at that.
>
> As an aside, in the process of converting my collection to vorbis, I've
> already discovered an annoying little glitch in one of my CDs that had
> been masked by the "pop", so it's not all bad. :-)

See, the problem is that the decoder should know its own latency and the
encoder knows its own and so can store that information in the XING header.
MP3 will splice gaplessly and with the editing software I have, I can edit
down to the sample and have done it visually with sample-perfect results.
What's intriguing is that this has any issues at all.

So, I'm not sure who is the one that didn't think it through enough-- the
designers or myself by just being unaware of something that becomes obvious
when getting into the nitty-gritty.


-------------------------------------------------------------------
List admin: http://cool.haxx.se/cgi-bin/mailman/listinfo/rockbox
FAQ: http://www.rockbox.org/twiki/bin/view/Main/GeneralFAQ
Etiquette: http://www.rockbox.org/mail/etiquette.html
Received on 2009-02-25