On Mon, 04 Sep 2000 21:51:27 +0200, Andreas Mueller wrote:
>Do you have an Intel CPU? I've got some reports that my pre compiled
>binaries contain some Intel optimized code (I'm not sure why) that
>fails on AMD CPUs. Maybe you can try to recompile the package on your
>host and check if the crash still occurs.
>I checked the code if there is bug in the driver table reading function
>but it looks OK to me.
Ah.. thats probally it.. I have an AMD. I will get the source and try a compiled
version
tonight.
>> #3 DAE is very slooooow. Much slower than cdparanoia.
>
>Cdrdao reads the sub-channel data in addition to the main channel
>audio data. Maybe this slows down the extraction process. Especially,
>if you get lots of CRC errors, your drive is probably not able to
>retrieve the sub-channel data correctly. BTW, the CRC errors are only
Its on a TEAC R58S. How would one go about figuring out if it has issues with
sub-channel
reading?
I did a little testing over the holiday weekend and found that --paranoia-mode 0 is
quite
fast and mode 1 (jitter only) is only about 3 minutes or so slower than using
cdparanoia.
And I seem to have a good audio copy. Or at least I haven't heard any artifacts yet.
So I gather from this that cdparanoia dosen't read sub-channel data? The man page
dosn't
say anything about it. Also the man page of cdparanoia seems to suggest to me that by
default its in a "full paranoia" mode as I only see options to disable various
checking
and not to enable it. Is this the case?
>marks this way. It's also worth trying to use options
>'--driver generic-mmc:0x10' or '--driver generic-mmc:0x13' to select
>different sub-channel reading modes.
I will play with those and see if I can make any difference.
>data source definitions. Do you think I should point out this more
>explicitly? However, I just noticed that the description for the location
>of the global CD-TEXT data block is misleading.
Well let my tell you how I went through the man page. The read-cd stuff is pretty
self
explanitory. So I had a image and a toc file going in short order. Then I started
looking for info on how to add CD-TEXT. The first info I came across for cd-text
stated
that it was explained in the cd-text block section. So I jumped to that section and a
quick cut and past later was trying test burns. So I kinda glossed over the section
talking about where the placement should be.
Now perhaps its RTFM on my part but I very seldom read an entire man page. Maybe a
short
blurb in that section showing the overall structure of where cdtext info should go
would
be a good addition.
>I decided to not have such a magic automatism to avoid unwanted results
>caused by a missing entry for a track. Unfortunately, I missed to update
>the documentation.
Ahh.. that explains things... Also I noticed if you use gcdmaster to create cd-text
info
from scratch it indeed does not add the performer info if disable is selected. But if
you
have a PERFORMER statment in the file when loaded an then selecting disable performer
still writes performer statements.
In addition, gcdmaster does not create a LANGUAGE MAP block either. If a LANGAGE MAP
is
not specified cdrdao will not write any cd-text info. It also dosen't give any error
or
warnings or such. The only way to see if CD-TEXT info is being written is to enable
the
advanced debugging output.
>Usually, the source audio data is already corrupted. Did you check that?
>Otherwise it might have been a temporary problem with the CD-R media.
No, and I deleted the image so I can't go back and see. However my second attempt was
completely sucessful.
How hard would it be to add a --verify option that went back and double checked the
audio
stream so I can play with various modes and see what works best?
My sucessfull burn used mode 1 (jitter correction only) for the read and then burned
at 4x
and created what seems to be my first flawless CD-TEXT CD.
As I have about 200 cd's I want to do this with so I will have ample opportunity to
experiment.
--
Richard A. Smith Bitworks, Inc.
[EMAIL PROTECTED] 501.846.5777
Sr. Design Engineer http://www.bitworks.com
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
