Help - Search - Members - Calendar
Full Version: Pregap detection discussion
Hydrogenaudio Forums > Hosted Forums > foobar2000 > Support - (fb2k)
Fandango
Feb 27 2006, 07:33
QUOTE(Garf @ Feb 27 2006, 01:23 PM)
0.9 RC1 handles pregap information perfectly.
*



Yeah I know. I believe you, but what I meant was that it isn't supported in cue sheet creation/handling. Internally it supports it very well.... good enough for ripping and encoding properly. But as long as fb2k can't write cue sheets that are exact representations of the original CD layout, it's of no use for backup purposes.
Peter
Feb 27 2006, 08:17
You are simply wrong and your info is outdated.
Fandango
Feb 27 2006, 08:52
Then enlighten me, please.
Peter
Feb 27 2006, 09:10
Which part of my post above you did not understand?
Fandango
Feb 27 2006, 09:18
Are you serious? All of it.
Peter
Feb 27 2006, 09:20
Which version are your conclusions about pregaps/subindexes getting lost during cuesheet creation based on?
Fandango
Feb 27 2006, 09:25
v0.9 RC (2006-02-24)

And this is the way I rip a CD (might be the reason why I get no pregaps, maybe I do it wrong):

Insert CD, Open audio CD, Setting up my CD-ROM drive (paranoid + offset correction), Add to playlist, select all tracks, Detect CD pre-gaps, Convert to album images with cuesheets or chapters (WAV format)

Here are the results:

ExactAudioCopy v0.95b4

CODE

REM GENRE Rock
REM DATE 1977
REM DISCID 8A08970A
REM COMMENT "ExactAudioCopy v0.95b4"
PERFORMER "Pere Ubu"
TITLE "The Modern Dance"
FILE "Pere Ubu - The Modern Dance.wav" WAVE
TRACK 01 AUDIO
TITLE "Nonalignment Pact"
PERFORMER "Pere Ubu"
INDEX 01 00:00:00
TRACK 02 AUDIO
TITLE "Modern Dance"
PERFORMER "Pere Ubu"
INDEX 00 03:18:26
INDEX 01 03:19:45
TRACK 03 AUDIO
TITLE "Laughing"
PERFORMER "Pere Ubu"
INDEX 00 06:47:65
INDEX 01 06:49:57
TRACK 04 AUDIO
TITLE "Street Waves"
PERFORMER "Pere Ubu"
INDEX 00 11:24:69
INDEX 01 11:27:00
TRACK 05 AUDIO
TITLE "Chinese Radiation"
PERFORMER "Pere Ubu"
INDEX 00 14:31:33
INDEX 01 14:33:68
TRACK 06 AUDIO
TITLE "Life Stinks"
PERFORMER "Pere Ubu"
INDEX 00 18:01:28
INDEX 01 18:03:12
TRACK 07 AUDIO
TITLE "Real World"
PERFORMER "Pere Ubu"
INDEX 00 19:55:38
INDEX 01 19:56:07
TRACK 08 AUDIO
TITLE "Over My Head"
PERFORMER "Pere Ubu"
INDEX 00 23:55:63
INDEX 01 23:57:63
TRACK 09 AUDIO
TITLE "Sentimental Journey"
PERFORMER "Pere Ubu"
INDEX 00 27:46:42
INDEX 01 27:47:25
TRACK 10 AUDIO
TITLE "Humor Me"
PERFORMER "Pere Ubu"
INDEX 00 33:53:06
INDEX 01 33:55:23


foobar2000 v0.9 RC (2006-02-24)

CODE

FILE "_.wav" WAVE
TRACK 01 AUDIO
TITLE "CD track 01"
INDEX 01 00:00:00
TRACK 02 AUDIO
TITLE "CD track 02"
INDEX 01 03:19:45
TRACK 03 AUDIO
TITLE "CD track 03"
INDEX 01 06:49:57
TRACK 04 AUDIO
TITLE "CD track 04"
INDEX 01 11:27:00
TRACK 05 AUDIO
TITLE "CD track 05"
INDEX 01 14:33:68
TRACK 06 AUDIO
TITLE "CD track 06"
INDEX 01 18:03:12
TRACK 07 AUDIO
TITLE "CD track 07"
INDEX 01 19:56:07
TRACK 08 AUDIO
TITLE "CD track 08"
INDEX 01 23:57:63
TRACK 09 AUDIO
TITLE "CD track 09"
INDEX 01 27:47:25
TRACK 10 AUDIO
TITLE "CD track 10"
INDEX 01 33:55:23

Peter
Feb 27 2006, 09:26
Detection of CD pregaps is confirmed buggy (I am still looking into it), does not apply to creation of cuesheets when converting from CD image that already has correct pregap/subindex info.
Fandango
Feb 27 2006, 09:54
Ok, good to know.

And I did a quick test with the same CD (EAC single image). I loaded the rip into both beta 6 and RC. Converted tracks to single files (WAVs) - all the same except for one file: Track 7 - same size but different hash. I converted that track again, just to be sure, but: same outcome.

And then I noticed that post-beta 6 releases got non-compliant cue + multiple files support back, that's a good thing IMHO, I'm happy you guys decided to put it back in. Nevertheless when doing bit-to-bit compares of single tracks (between the same tracks once of non-compliant and of the single image rip respectively) in RC I get mismatches on the tracks with pre-gaps, did this with another CD of which I already deleted the rips, so sorry, can't tell you which kind of cue entries this happens with. I tried the same with beta 6, always "No differences in decoded data found.". Maybe you should look into this matter, too.
Peter
Feb 27 2006, 10:44
"Non-compliant" cue sheets are not supported - making them work as intended would require too many structural changes (multiple referenced files per track) that go against the standard; if you somehow got them to play, you are most likely getting incorrect results. Whether this will be changed or not in the future is yet to be decided.
Fandango
Feb 27 2006, 11:16
Thanks for the clarification.
Peter
Feb 27 2006, 12:44
QUOTE(Peter @ Feb 27 2006, 04:26 PM)
Detection of CD pregaps is confirmed buggy (I am still looking into it), does not apply to creation of cuesheets when converting from CD image that already has correct pregap/subindex info.
*


The bad news is that current CD pregap detection is flamingly insecure compared to properly configured EAC according to my testing today and I might pull the plug on it for 0.9 release as I don't want people to use something that I know is not as reliable as it should be.

The good news is that I have a test case CD and will try to improve this situation when I have time.

Interestingly enough, my favourite subindex read test disc (Marillion - Misplaced Childhood, EMI, 1998 remaster) would read fine a few months ago at the point I was writing subindex read routines, but does not anymore (tried three different LG drives in different machines), and also gives nonsense when you use EAC with default settings. I initially thought I found some kind of bug in the code that was introduced since last time this was tested, but I got similar results with old versions pulled from source backup archive.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2008 Invision Power Services, Inc.