From http://bugs.musicbrainz.org/ticket/4397, paraphrasing discussion:

I wondered why the tracknumber of the first track is part of the discid... now I am trying to rip and tag just such a disc, the second disc of a 2-disc set where the first track is #11... unfortunately, picard does not seem to generate the right toc. http://musicbrainz.org/bare/cdlookup.html?id=1RqCOj.Qic5LqYIxqjKUVDzmL4U-&tracks=20&toc=1+20+332466+150+150+150+150+150+150+150+150+150+150+150+36122+68078+84202+129986+156150+165005+176025+223006+265254

I believe this should be: toc=11+20+332466+150+36122+68078+84202+129986+156150+165005+176025+223006+265254 ?

anyway the disc already has a discid, which appears exactly the same except for the numbering... the disc really does have no track 1-10, did not know such existed.

Output from "cdrdao read-toc"

Reading toc data...

Track Mode Flags Start Length

11 AUDIO 2 00:00:00( 0) 07:59:47( 35972) 12 AUDIO 2 07:59:47( 35972) 07:06:06( 31956) 13 AUDIO 2 15:05:53( 67928) 03:34:74( 16124) 14 AUDIO 2 18:40:52( 84052) 10:10:34( 45784) 15 AUDIO 2 28:51:11(129836) 05:48:64( 26164) 16 AUDIO 2 34:40:00(156000) 01:58:05( 8855) 17 AUDIO 2 36:38:05(164855) 02:26:70( 11020) 18 AUDIO 2 39:05:00(175875) 10:26:31( 46981) 19 AUDIO 2 49:31:31(222856) 09:23:23( 42248) 20 AUDIO 2 58:54:54(265104) 14:56:12( 67212) Leadout AUDIO 2 73:50:66(332316)

luks asks user:

Can you try the attached file, please? You will need to install the development version of libcdio for that (on Ubuntu/Debian it's libcdio-dev). To compile it, use:

gcc cdtoc-dbg.c -o cdtoc-dbg -lcdio

and then run for example:

./cdtoc-dbg /dev/cdrom

It will output a log of text, can you please attach it to this ticket?

.. output is attached by user