Just another peculiar AccurateRip result

This could be an example answering a frequently asked question, I am not sure which though:

[IMG]http://www.home.no/porcus/pulse.jpeg[/IMG]

Have a look at Disc 2, Track 5, The Great Gig in the Sky. Possible hypotheses:

- TGGitS is a track that half of us hate and deliberately omit,
or
- There are production differences at track level, which might explain that a single track exhibits weird results.


I'd rather lay my money down on the second of these, but Spoon: Is it possible for you to look up this particular disc and see if it is so that there is a matching entry for this track with AR around sixty?


(Of course you might hypothesize that TGGitS is a track that someone loves so much that (s)he did sixty different rips of his Pulse CD in order to have an accurate rip, hence every other track has that much higher score. That's why I included Disc 1 :D )

What is the accuraterip disc ID? (in idtags or secure rip log).

[QUOTE=Spoon;81685]What is the accuraterip disc ID? (in idtags or secure rip log).[/QUOTE]

DiscIDs and CRCs:

Track 1: CRC32: BA732B5A AccurateRip CRC: 8315CFAF [DiscID: 013-001e606b-013a2eac-aa10d30d-1]

Track 2: CRC32: D26B081B AccurateRip CRC: 47867F94 [DiscID: 013-001e606b-013a2eac-aa10d30d-2]

Track 3: CRC32: 8D8B617D AccurateRip CRC: 5226CA61 [DiscID: 013-001e606b-013a2eac-aa10d30d-3]

Track 4: CRC32: 83B90991 AccurateRip CRC: 4719E9BC [DiscID: 013-001e606b-013a2eac-aa10d30d-4]

Track 5: CRC32: 10AE417D AccurateRip CRC: 6151788D [DiscID: 013-001e606b-013a2eac-aa10d30d-5]

Track 6: CRC32: 846CD5B6 AccurateRip CRC: 5580A89D [DiscID: 013-001e606b-013a2eac-aa10d30d-6]

Track 7: CRC32: 52A1D786 AccurateRip CRC: 2E61B880 [DiscID: 013-001e606b-013a2eac-aa10d30d-7]

Track 8: CRC32: 1F5FE3F8 AccurateRip CRC: 0F03E4F2 [DiscID: 013-001e606b-013a2eac-aa10d30d-8]

Track 9: CRC32: 9B1BCFA9 AccurateRip CRC: 4E088BD3 [DiscID: 013-001e606b-013a2eac-aa10d30d-9]

Track 10: CRC32: 44265A02 AccurateRip CRC: 2E348EA5 [DiscID: 013-001e606b-013a2eac-aa10d30d-10]

Track 11: CRC32: 9088CB85 AccurateRip CRC: D42DE47D [DiscID: 013-001e606b-013a2eac-aa10d30d-11]

Track 12: CRC32: 5F8A20E5 AccurateRip CRC: 31B0304B [DiscID: 013-001e606b-013a2eac-aa10d30d-12]

Track 13: CRC32: 42FE8B0F AccurateRip CRC: D72532E1 [DiscID: 013-001e606b-013a2eac-aa10d30d-13]


(I have checked both mine and a friend's copy, they appear identical.)

I checked, and would put it down to a manufacturing defect (some drives read one way, others another).

[QUOTE=Spoon;81725]I checked, and would put it down to a manufacturing defect (some drives read one way, others another).[/QUOTE]

("another" meaning "the other" in the sense that there are only two prevalent CRCs here?)


My main hypothesis was that this could explain the frequent "CD is brand new but still gives Inaccurate (7) on one and only one track" questions. Anyway, with these high AR figures, it is not a coincidence.



It also supports my speculations in [url]http://forum.dbpoweramp.com/showthread.php?p=76256[/url], and suggests EliC's wishlist entry on a "forced secure" even on accurate matches -- at least if C2 errors are detected, which I understand counts nothing as long as there is an AR match, right?

Some other possibilities encompassed by spoon's comment regarding manufacturing defects:

1. That there are actually two families of discs (perhaps two sets of glass masters) that are digitally identical except for the contents of a single track, probably a single sector.

Or...

2. Similarly, that there is only one family of discs, but the glass master(s) for them contain a series of (weak?) bits that could validly lead to two possible correct interpretations of a single sector with some drives interpreting one way and some drives interpreting another.

Or...

3. That different *methods* of reading (burst vs. secure) the same exact disc may give different AR results due to the occasional drive family firmware bugs ... remember that AR submissions come from EAC, not just dbpoweramp, so the "but dbpoweramp always uses a burst pass first!" rebuttal doesn't necessarily apply to the inconsistent AR match numbers problem.

It would be interesting to do a series of tests in EAC to see if you could get both AR results on that track from the same disc by tweaking the EAC settings.

-brendan

[QUOTE=Porcus;81730]It also supports my speculations in [url]http://forum.dbpoweramp.com/showthread.php?p=76256[/url], and suggests EliC's wishlist entry on a "forced secure" even on accurate matches -- at least if C2 errors are detected, which I understand counts nothing as long as there is an AR match, right?[/QUOTE]

I'm pretty sure that dbpoweramp, with C2 on, will do more than one track read even if an AR match is found if any of the C2 bits returned are 1. Spoon, can you verify this?

-brendan

The issue of these disc defects is at drive level and independant of c2 AFAIK.

[QUOTE=Spoon;81747]The issue of these disc defects is at drive level and independant of c2 AFAIK.[/QUOTE]

Uh ... pardon my ignorance here, but: by "at drive level" I assume that you imply that there are ambiguous bits which drive x will read as 0 and drive y will read as 1. Would still both drives report consistency with all error checks?


Bhoar: what settings do you suggest I tweak in EAC?

[QUOTE=Porcus;81754]Bhoar: what settings do you suggest I tweak in EAC?[/QUOTE]

I was mostly thinking testing a burst pass vs. a secure pass.

-brendan

[QUOTE=Porcus;81719]
Track 5: CRC32: 10AE417D AccurateRip CRC: 6151788D [DiscID: 013-001e606b-013a2eac-aa10d30d-5]
[/QUOTE]

I have the same disc, and have this result:

Track 5: Ripped LBA 70380 to 96780 (5:52) in 0:14. Filename: D:\..\Pink Floyd - 05 - The Great Gig in the Sky.flac
AccurateRip: Accurate (confidence 116) [Pass 1]
CRC32: 10AE417D AccurateRip CRC: 6151788D [DiscID: 013-001e606b-013a2eac-aa10d30d-5]

Same CRC, AccurateRip and DiscID, luckily the confidence is now higher :), but still the other tracks are around 180 compared to this at 116.

This was the settings, from the same log-file:
Ripping with drive 'F: [TSSTcorp - DVD+-RW TS-H653B]', Drive offset: 6, Overread Lead-in/out: No
AccurateRip: Active, Using C2: Yes, Cache: None, FUA Cache Invalidate: No
Pass 1 Drive Speed: Max, Pass 2 Drive Speed: Max
Ultra:: Vary Drive Speed: No, Min Passes: 3, Max Passes: 6, Finish After Clean Passes: 2
Bad Sector Re-rip:: Drive Speed: Max, Maximum Re-reads: 34