castalla wrote:
> file
Progress of sorts.
It is clear something is going wrong with cdda2wav. cdda2wav is issuing
a read command and it fails - twice. If the errors are coincident with
"scatterlist" errors, it sort of implies a kernel issue. AFAIK the CD
drive kernel module is generic - no difference between CD drives.
Code:
--------------------
0%Errno: 5 (Input/output error), ReadCD MMC 12 scsi sendcmd: cmd timeout
after 61.439 (60) s
CDB: BE 04 00 00 00 21 00 00 34 10 00 00
resid: 122290
cmd finished after 61.439s timeout 60s
0%Errno: 5 (Input/output error), ReadCD MMC 12 scsi sendcmd: cmd timeout
after 61.440 (60) s
CDB: BE 04 00 00 00 55 00 00 34 10 00 00
resid: 122290
cmd finished after 61.440s timeout 60s
--------------------
Since it has been possible to execute the command standalone - it seems
to point to a CPU, load or timing issue rather than memory.
Not sure about current drives but CD drives used to be very CPU
sensitive. The drive had to have its data read within a certain time
period otherwise error would follow. To minimise the chances - cdda2wav
ups its scheduling priority if it can (probably as no message otherwise)
"#icedax version 1.1.11, real time sched., soundcard, libparanoia
support"
I did tests many years ago and found that default settings and "speed 4"
was the best compromise of speed with high rate of success. Drives have
improved hugely since those tests - I don't know if the drive is
internally too fast for the CPU or too slow.
There are lots of cdda2wav parameters that can be "tweaked" just not
sure which ones actually make a difference. I'll have to do some tests.
That said a Pi/PCP expert opinion would be welcome.
Code:
--------------------
icedax
(-S) speed=speedfactor set the cdrom drive to a given speed during
reading
(-P) set-overlap=sectors set amount of overlap sampling (default is 0)
(-n) sectors-per-request=secs read 'sectors' sectors per request.
(-l) buffers-in-ring=buffers use a ring buffer with 'buffers' elements.
--------------------
and cdparanoia extra options "-paraopts= " which I have never gone
near
Code:
--------------------
retries=amount set the number of maximum retries per sector.
overlap=amount set the number of sectors used for statical paranoia overlap.
minoverlap=amt set the min. number of sectors used for dynamic paranoia
overlap.
maxoverlap=amt set the max. number of sectors used for dynamic paranoia
overlap.
--------------------
------------------------------------------------------------------------
bpa's Profile: http://forums.slimdevices.com/member.php?userid=1806
View this thread: http://forums.slimdevices.com/showthread.php?t=47288
_______________________________________________
plugins mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/plugins
