#399517 cuetools: cueconvert doesn't correctly convert TOC with no START tag for a track

Package:
cuetools
Source:
cuetools
Description:
tools for manipulating CUE/TOC files
Submitter:
Andy Hawkins
Date:
2015-10-03 18:54:08 UTC
Severity:
important
#399517#5
Date:
2006-11-20 10:05:31 UTC
From:
To:
When using cueconvert to convert a TOC file to a CUE file, the INDEX points
are incorrectly numbered if the TOC file doesn't contain START tags for
every track.

For example, the following TOC file:

CD_DA

CATALOG "0042284361422"

// Track 1
TRACK AUDIO
NO COPY
NO PRE_EMPHASIS
TWO_CHANNEL_AUDIO
FILE "cousteau-2006-11-19.21:11:15-13112.bin" 0 08:23:50
START 00:00:33
INDEX 04:11:07
INDEX 05:09:37


// Track 2
TRACK AUDIO
NO COPY
NO PRE_EMPHASIS
TWO_CHANNEL_AUDIO
FILE "cousteau-2006-11-19.21:11:15-13112.bin" 08:23:50 07:10:50
INDEX 04:05:30


// Track 3
TRACK AUDIO
NO COPY
NO PRE_EMPHASIS
TWO_CHANNEL_AUDIO
FILE "cousteau-2006-11-19.21:11:15-13112.bin" 15:34:25 06:29:35


// Track 4
TRACK AUDIO
NO COPY
NO PRE_EMPHASIS
TWO_CHANNEL_AUDIO
FILE "cousteau-2006-11-19.21:11:15-13112.bin" 22:03:60 46:54:63
START 00:01:20

is converted to the following CUE file:

CATALOG 0042284361422

FILE "cousteau-2006-11-19.21:11:15-13112.bin" WAVE
TRACK 01 AUDIO
INDEX 00 00:00:00
INDEX 01 00:00:33
INDEX 02 04:11:07
INDEX 03 05:09:37

TRACK 02 AUDIO
INDEX 00 08:23:50
INDEX 01 12:29:05

TRACK 03 AUDIO
INDEX 01 15:34:25

TRACK 04 AUDIO
INDEX 00 22:03:60
INDEX 01 22:05:05


As you can see, the INDEX 01 for track 2 is actually the index of the first
INDEX point in that track, not the index of the start of the track.

I can work around it by adding a 'START 00:00:00' to any track that doesn't
already have a START tag, but that shouldn't be necessary should it?
#399517#12
Date:
2008-05-22 11:10:40 UTC
From:
To:
Hello,

while your bug reports are done with older versions and many changes
have been made I wanted to ask, if you still could reproduce these bugs?

Thanks in advance.
#399517#17
Date:
2008-10-31 15:45:51 UTC
From:
To:
* Patrick Matthäi <patrick.matthaei@web.de> [2008-05-22 13:10]:

Given that Andy included a test case, it's pretty easy to verify that
this is indeed still there.

With current cueconvert you get (for track 2):

TRACK 02 AUDIO
INDEX 00 08:23:50
INDEX 01 12:29:05

where with toc2cue you get:

  TRACK 02 AUDIO
    INDEX 01 08:23:50

(so toc2cue ignores the 2nd index point, but at least it gets the
first one right)
#399517#26
Date:
2015-10-03 18:50:37 UTC
From:
To:
I just had a look at this in jessie with cuetools 1.4.0-2

I copied the sample TOC into a file called test1.toc and executed:

$ cueconvert test1.toc > test1.cue

and observe the same fault in test1.cue

Then I tried adapting the TOC with the sed script[1]

$ cat test1.toc | sed -f tocfix.sed > test1-with-START.toc

$ cueconvert test1-with-START.toc > test2.cue

and test2.cue appears different (is it correct?):

CATALOG 0042284361422

FILE "cousteau-2006-11-19.21:11:15-13112.bin" WAVE
TRACK 01 AUDIO
INDEX 00 00:00:00
INDEX 01 00:00:33
INDEX 02 04:11:07
INDEX 03 05:09:37

TRACK 02 AUDIO
INDEX 01 08:23:50
INDEX 02 12:29:05

TRACK 03 AUDIO
INDEX 01 15:34:25

TRACK 04 AUDIO
INDEX 00 22:03:60
INDEX 01 22:05:05


I feel the script name tocfix.sed is slightly misleading, the TOC is not
actually broken, maybe toc-workaround.sed may be more precise.

Upstream cueconvert is now on Github[2] so maybe it is possible for
somebody to make a pull request to finally fix it

1.
https://anonscm.debian.org/cgit/pkg-multimedia/flactag.git/tree/tocfix.sed

2. https://github.com/svend/cuetools