[Vorbis] Re: Vorbis granule position

[Ian Malone]

James M. Rotenberry wrote:
> Ralph Giles
> 
> Thank you for your comments.
> 
>> Decreasing granulepos is a problem, and shouldn't happen. What stream
>> are you seeing this from?
> 
> Based on my experience and the results of the tests I have appended
> to this message, I believe that EVERY rip of streaming ogg vorbis
> will have this problem.
> 
> The tests I ran used three different ogg vorbis streams, three
> different ripping programs (wget, streamripper, and fetch), and
> two different operating systems. In all three cases there is a
> granulepos problem and, consequently, incorrect values for the
> Playback length and the Average bitrate.
>

I can't tell from the ogginfos you supplied: do you know if
they have the decreasing granulepos problem mentioned in the
link from your earlier email or is it just missing (= -1)
granulepos? (Your ogginfo output doesn't appear to contain
the decreasing granulepos issue.)  I don't think the missing
pages bit is a cause for concern, this has probably happened
as you have the header pages and then the stream starting
from whatever time you joined, though it may be causing
the length calculation problems you're seeing.  Running
vorbiscomment on them may fix that.

-- 
imalone