[Ffmpeg-devel] OT: cdrecord
Sun Jul 31 15:21:57 CEST 2005
On Saturday 30 July 2005 18:45, Rich Felker wrote:
> realtime priority is not useful; joerg just claims it is. i've nver
> used realtime priority, even when i had an old burner, and all modern
> burners have "burnproof" functionality which turns off the laser when
> the buffer drains and resyncs and resumes burning when more data is
> available, all in hardware with no software intervention.
Joerg claims this burnproof functionality degrades the disc quality so much
that you would rather want to be notified of this fault, fix the cause, throw
away the coaster, and get a new blank disc and burn again. Infact, this is
the default behaviour in cdrecord.
It's partly true, the resyncs create a few bytes of corrupted data, but it's
happily fixed by one of the many layers of ECC on CDs...
Add dodgy cheap media, high burn speeds, careless storage and handling
of the discs, and you might get enough extra corruption for the damage+
resync to become uncorrectable. You hear teh insane laughter of Joerg
when you discover your data is gone :)
Btw, anyone ever tried readcd -c2scan after burning CDs at different speeds?
> the "cd burning software must run realtime" thing is a myth born out
> of joerg's ego, whereby he thinks every program written by him must be
> given godlike permissions just because he is a god...
Another effective way of creating coasters is to put source drive and burner
on same IDE cable. Now, if you are running without realtime, and you've got
heavy system load which makes cdrecord lose CPU for a long time, it might
free up the bus from burner load long enough for more data to arrive from the
source, and if you are lucky, cdrecord gets CPU again just in time before burner
buffer is empty :))
With cdrecord at realtime priority, I imagine it would be sending one sector at a time
to the burner, and if you have burnproof enabled, lots of linking/resyncing causing
more c2errors :)
JK, who bought an extra IDE controller to dedicate to his burner
More information about the ffmpeg-devel