IPB

Welcome Guest ( Log In | Register )

EAC vs. ImgBurn gap detection, Two rippers giving me different values in cue sheet
Tony1337
post Oct 5 2011, 22:51
Post #1





Group: Members
Posts: 4
Joined: 5-October 11
Member No.: 94162



I ripped a CD of Ashkenazy's Chopin Etudes with each of them, and got the following cue sheets:

*** EAC ***

CODE
REM DISCID A60EC70C
REM COMMENT "ExactAudioCopy v1.0b2"
CATALOG 0028941412720
PERFORMER "Unknown Artist"
TITLE "Unknown Title"
FILE "Image.wav" WAVE
TRACK 01 AUDIO
TITLE "Track01"
PERFORMER "Unknown Artist"
ISRC GBF077230601
INDEX 00 00:00:00
INDEX 01 00:00:32
TRACK 02 AUDIO
TITLE "Track02"
PERFORMER "Unknown Artist"
ISRC GBF077230603
INDEX 00 03:20:72
INDEX 01 03:23:42
TRACK 03 AUDIO
TITLE "Track03"
PERFORMER "Unknown Artist"
ISRC GBF077230605
INDEX 00 09:52:19
INDEX 01 09:53:65
TRACK 04 AUDIO
TITLE "Track04"
PERFORMER "Unknown Artist"
ISRC GBF077230607
INDEX 00 15:41:42
INDEX 01 15:42:47
TRACK 05 AUDIO
TITLE "Track05"
PERFORMER "Unknown Artist"
ISRC GBF077230609
INDEX 00 19:34:69
INDEX 01 19:36:20
TRACK 06 AUDIO
TITLE "Track06"
PERFORMER "Unknown Artist"
ISRC GBF077230611
INDEX 00 24:14:59
INDEX 01 24:17:72
TRACK 07 AUDIO
TITLE "Track07"
PERFORMER "Unknown Artist"
ISRC GBF077230701
INDEX 00 29:29:64
INDEX 01 29:39:07
TRACK 08 AUDIO
TITLE "Track08"
PERFORMER "Unknown Artist"
ISRC GBF077230703
INDEX 00 34:27:17
INDEX 01 34:29:65
TRACK 09 AUDIO
TITLE "Track09"
PERFORMER "Unknown Artist"
ISRC GBF077230705
INDEX 00 37:56:39
INDEX 01 38:00:22
TRACK 10 AUDIO
TITLE "Track10"
PERFORMER "Unknown Artist"
ISRC GBF077230707
INDEX 00 43:43:24
INDEX 01 43:46:17
TRACK 11 AUDIO
TITLE "Track11"
PERFORMER "Unknown Artist"
ISRC GBF077230709
INDEX 00 50:39:17
INDEX 01 50:42:40
TRACK 12 AUDIO
TITLE "Track12"
PERFORMER "Unknown Artist"
ISRC GBF077230711
INDEX 00 56:24:67
INDEX 01 56:28:30

*** IMGBURN ***

CODE
CATALOG 0028941412720
FILE "Image.wav" WAVE
TRACK 01 AUDIO
ISRC GBF077230601
INDEX 00 00:00:00
INDEX 01 00:00:32
TRACK 02 AUDIO
ISRC GBF077230603
INDEX 00 03:20:73
INDEX 01 03:23:42
TRACK 03 AUDIO
ISRC GBF077230605
INDEX 00 09:52:20
INDEX 01 09:53:65
TRACK 04 AUDIO
ISRC GBF077230607
INDEX 00 15:41:43
INDEX 01 15:42:47
TRACK 05 AUDIO
ISRC GBF077230609
INDEX 00 19:34:70
INDEX 01 19:36:20
TRACK 06 AUDIO
ISRC GBF077230611
INDEX 00 24:14:60
INDEX 01 24:17:72
TRACK 07 AUDIO
ISRC GBF077230701
INDEX 00 29:29:65
INDEX 01 29:39:07
TRACK 08 AUDIO
ISRC GBF077230703
INDEX 00 34:27:18
INDEX 01 34:29:65
TRACK 09 AUDIO
ISRC GBF077230705
INDEX 00 37:56:40
INDEX 01 38:00:22
TRACK 10 AUDIO
ISRC GBF077230707
INDEX 00 43:43:25
INDEX 01 43:46:17
TRACK 11 AUDIO
ISRC GBF077230709
INDEX 00 50:39:18
INDEX 01 50:42:40
TRACK 12 AUDIO
ISRC GBF077230711
INDEX 00 56:24:68
INDEX 01 56:28:30

As you can see, each of the INDEX 00 values differs by 0:00:01. (Yes, I'm OCD about these things.) I've tried EAC on A and B gap detection modes, with the same results (C stalls). I've tried ImgBurn on 1-2-3, 1-3-2, and 3-1-2 modes with the same results (the others stall). For most of my other CDs, EAC and ImgBurn match perfectly on A and 1-2-3 modes, my default settings respectively. Intuitively it seems like I should trust EAC because it's more advanced, but the last and only other time they didn't match, EAC was horribly off by several minutes on a track while ImgBurn was correct (or at least a reasonable value). From your experiences, which cue sheet would you trust?

This post has been edited by db1989: Oct 6 2011, 10:32
Reason for edit: Please place large items of text within [codebox]one of these[/codebox]
Go to the top of the page
+Quote Post
 
Start new topic
Replies
mjb2006
post Oct 6 2011, 05:24
Post #2





Group: Members
Posts: 816
Joined: 12-May 06
From: Colorado, USA
Member No.: 30694



The CD has a table of contents (TOC) in the lead-in, and it does have some metadata, but it mainly only just points to the sector where each track's index 01 should begin. This is what is used for the initial display of track durations and for jumping from track to track.

The audio data comes after the lead-in, and is multiplexed with error correction info plus several other data streams called sub-channels or subcode channels. One of the subcode channels (Q) contains all kinds of stuff, including the track and index numbers and timecode. This is the info that a CD player uses for its track & time display, and is the only place where pregap info is found.

The ripper doesn't have access to the raw multiplexed data. CD drives demux it in the hardware & firmware, and only provide software access to interpreted streams. The ripper can ask for the audio data only, or the subcode only. I don't know any technical details, but based on comments I've seen by the developers of cdrtools, PerfectRip, and CUETools, it seems that subcode-reading is an inexact science with variable support in CD drives, which is why you can get different results from different drives or different reading methods, or even from read to read on the same drive. I also saw a mention that the subcode data is sometimes offset from the audio data it corresponds to.

If you only changed index 00 locations in the cue sheet, it should not result in different audio files between CDs burned otherwise identically and ripped without error. You should provide more info about your burning and ripping process and how exactly the cue sheets and ripped audio files differ. EAC has a WAV comparison tool (ctrl+W) that can help. Before using it, go to the WAV editor options (Alt+Q), Compare tab, and make sure "Determine differences more exactly" is checked. Also when doing the rips, make sure you don't have Normalize enabled in the EAC options.

This post has been edited by mjb2006: Oct 6 2011, 05:25
Go to the top of the page
+Quote Post
lukpac
post Jan 16 2012, 20:51
Post #3





Group: Members
Posts: 12
Joined: 16-January 12
Member No.: 96485



First post...want to chime in with some information and pose some more questions.

QUOTE (Cynic @ Oct 5 2011, 18:31) *
Unless my understanding of gaps is completely wrong, there is no way to know for sure which is right. Regrettably, it's just one more uncertainty about CD ripping that you'll have to live with.


Of course there's one way to know what is right - burn a CD-R with known gaps and then attempt to detect those gaps. If the gaps don't match what you burned, something is wrong.

QUOTE (mjb2006 @ Oct 5 2011, 22:24) *
The ripper doesn't have access to the raw multiplexed data. CD drives demux it in the hardware & firmware, and only provide software access to interpreted streams. The ripper can ask for the audio data only, or the subcode only. I don't know any technical details, but based on comments I've seen by the developers of cdrtools, PerfectRip, and CUETools, it seems that subcode-reading is an inexact science with variable support in CD drives, which is why you can get different results from different drives or different reading methods, or even from read to read on the same drive. I also saw a mention that the subcode data is sometimes offset from the audio data it corresponds to.


This is going back to ancient times in computer land, but I still have a Mac that runs OS 9, and on it I have a copy of ASTARTE CD-Copy. It has always seemed faster and more reliable than EAC for gap detection. I often come across discs where EAC will hang while trying to detect a gap, but CD-Copy flies right through. I've even gone as far as physically moving drives from my Mac to a PC, and while they work fine under CD-Copy, they all have the same issues in EAC with certain discs. Switching EAC modes and/or switching to Inaccurate will *usually* let me get through a disc, but not always, and usually if I have to do that the gap in question is incorrect. I'm not sure CD-Copy was perfect, but it would sure be interesting to know how it did gap detection.

Thoughts?
Go to the top of the page
+Quote Post

Posts in this topic


Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 2nd October 2014 - 03:46