Skip to main content

Topic: High AccurateRip Confidence But Verification Fails (Read 2386 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.
  • korndawg
  • [*]
High AccurateRip Confidence But Verification Fails
After doing some research I'm 90% sure I know the answer to this question, but for peace of mind I'm going to ask it anyway. I'm using Exact Audio Copy 1.1 and AccurateRip cannot verify one specific track even though the confidence is consistent with the rest of the tracks from the disc:

Code: [Select]
Track  9

     Filename C:\Music\Def Leppard - Pyromania\09 Def Leppard - Rock Of Ages.wav

     Peak level 100.0 %
     Extraction speed 8.1 X
     Track quality 99.9 %
     Copy CRC B390D08F
     Cannot be verified as accurate (confidence 67)  [447C9337], AccurateRip returned [C9E97861]  (AR v2)
     Copy OK


Later the log summary provides details regarding the AccurateRip discrepancy:

Code: [Select]
  9   | (55/57) Differs in 115 samples @04:20:22,04:21:34,04:21:69,04:22:29-04:22:31,04:23:25,04:33:13


What I think this means is that the data ripped from this specific track matches what the data ripped by 67 other people, but that whoever initially uploaded the data calculated a slightly different AccurateRip checksum, possibly due to a scratch or other disc imperfection. Is this correct? Also, why does the confidence (67) not match the corresponding number from the log summary (55)?

Thanks in advance for your help!

Korndawg

  • korth
  • [*][*][*][*][*]
High AccurateRip Confidence But Verification Fails
Reply #1
Quote
Also, why does the confidence (67) not match the corresponding number from the log summary (55)?

You're looking at results from two different databases. AccurateRip in the above and the CUETools database below.
korth

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
High AccurateRip Confidence But Verification Fails
Reply #2
Your rip of this track differs from the one in the accuraterip database (possibly ripped in error despite what EAC is telling you?).

Your rip of this track differs from the one in the cuetools database by 115 samples.

The accuraterip database is not the same as the cuetools database.

EDIT: korth was faster.

EDIT #2: If it were me, I'd "repair" the track and then look at the differences.
  • Last Edit: 26 October, 2015, 02:16:01 PM by greynol
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • korndawg
  • [*]
High AccurateRip Confidence But Verification Fails
Reply #3
Your rip of this track differs from the one in the accuraterip database (possibly ripped in error despite what EAC is telling you?).

Your rip of this track differs from the one in the cuetools database by 115 samples.

The accuraterip database is not the same as the cuetools database.

EDIT: korth was faster.

EDIT #2: If it were me, I'd "repair" the track and then look at the differences.


If 55 of 57 people from the CUETools database generated the same AccurateRip checksum, isn't it a good bet that the 56 of us produced an accurate rip while the other two rips are outliers? Also, I assume that the "Copy CRC" is some type of file checksum used internally by EAC and not compared to an online database?

Thanks again for your help!

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
High AccurateRip Confidence But Verification Fails
Reply #4
In relation to the track in question, to be clear, you are not part of the "us".

I'll attempt to address outliers when I have more time.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • korndawg
  • [*]
High AccurateRip Confidence But Verification Fails
Reply #5
In relation to the track in question, to be clear, you are not part of the "us".

I'll attempt to address outliers when I have more time.


Understood, that's why I incremented the 55 from the log summary to 56 in order to include myself. 

Regarding the "Copy CRC", is there a utility to verify this value at a later date? For example, if I wanted to verify that a file has not been modified since it was ripped. I'm not aware of a common checksum algorithm (e.g., MD-5 or SHA-1) that generates an eight hexadecimal character string...

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
High AccurateRip Confidence But Verification Fails
Reply #6
That's misleading since you failed to present any data about the tracks that you ripped which were accurate.

Regarding the CRC, if you had included the EAC log header I could tell you what algorithm was used.  While it can be used to verify the integrity of the data, that is not the reason for EAC's presenting it.
  • Last Edit: 26 October, 2015, 04:47:25 PM by greynol
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • mjb2006
  • [*][*][*][*][*]
High AccurateRip Confidence But Verification Fails
Reply #7
Indeed there is a probability that there is only one pressing, accounted for by 67 ARv2 contributors and 55 of 57 CTDB contributors, with the 2 of 57 being bad rips of that one pressing. There is also a probability that there is another pressing, one which is not as commonly ripped, and which differs from the common pressing by a few samples. If that's the case, and if your rip really was error-free, then there won't be any way to know for sure until someone else with the same pressing gets the same result as you, boosting the confidence to 2. You could try the rip with a different drive if you would like to expedite this determination.

  • Porcus
  • [*][*][*][*][*]
High AccurateRip Confidence But Verification Fails
Reply #8
kornchild:
Your wording indicates that you think one of the log entries indicates that you have an accurate rip?

In relation to the track in question, to be clear, you are not part of the "us".


This.

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
High AccurateRip Confidence But Verification Fails
Reply #9
  • Last Edit: 26 October, 2015, 04:46:42 PM by greynol
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • korndawg
  • [*]
High AccurateRip Confidence But Verification Fails
Reply #10
That's misleading since you failed to present any data about the tracks that you ripped which were accurate.

Regarding the CRC, if you had included the EAC log header I could tell you what algorithm was used.  While it can be used to verify the integrity of the data, that is not the reason for EAC's presenting it.

The unabridged log follows. Could you shed some light on the "Copy CRC" algorithm? Also, your reply begs the question of why EAC presents the checksum value in the first place. Thanks!

Code: [Select]
Exact Audio Copy V1.1 from 23. June 2015

EAC extraction logfile from 12. October 2015, 22:40

Def Leppard / Pyromania

Used drive  : ASUS DRW-24B1ST  i  Adapter: 1  ID: 0

Read mode   : Secure
Utilize accurate stream : Yes
Defeat audio cache   : Yes
Make use of C2 pointers : No

Read offset correction : 6
Overread into Lead-In and Lead-Out : No
Fill up missing offset samples with silence : Yes
Delete leading and trailing silent blocks  : No
Null samples used in CRC calculations   : Yes
Used interface   : Native Win32 interface for Win NT & 2000
Gap handling : Not detected, thus appended to previous track

Used output format   : User Defined Encoder
Selected bitrate : 1024 kBit/s
Quality : High
Add ID3 tag : No
Command line compressor : C:\Program Files (x86)\Exact Audio Copy\FLAC\FLAC.EXE
Additional command line options : -8 -V -T "ARTIST=%artist%" -T "TITLE=%title%" -T "ALBUM=%albumtitle%"
 -T "DATE=%year%" -T "GENRE=%genre%" -T "COMMENT=%comment%" %source% -o %dest%


TOC of the extracted CD

Track |  Start  |  Length  | Start sector | End sector
---------------------------------------------------------
1  |  0:00.00 |  4:16.25 | 0 | 19224 
2  |  4:16.25 |  3:25.31 | 19225 | 34630 
3  |  7:41.56 |  3:22.28 | 34631 | 49808 
4  | 11:04.09 |  6:13.42 | 49809 | 77825 
5  | 17:17.51 |  4:43.23 | 77826 | 99073 
6  | 22:00.74 |  4:24.13 | 99074 |  118886 
7  | 26:25.12 |  4:58.58 | 118887 |  141294 
8  | 31:23.70 |  4:03.12 | 141295 |  159531 
9  | 35:27.07 |  4:53.20 | 159532 |  181526 
  10  | 40:20.27 |  4:05.62 | 181527 |  199963 
  11  | 44:26.14 |  3:23.18 | 199964 |  215206 
  12  | 47:49.32 |  5:55.39 | 215207 |  241870 
  13  | 53:44.71 |  5:55.31 | 241871 |  268526 
  14  | 59:40.27 |  4:55.25 | 268527 |  290676 
  15  | 64:35.52 |  6:08.57 | 290677 |  318333 


Track  1

Filename C:\Music\Def Leppard - Pyromania\01 Def Leppard - Rock! Rock! (Till You Drop).wav

Peak level 100.0 %
Extraction speed 5.4 X
Track quality 100.0 %
Copy CRC 47838172
Accurately ripped (confidence 66)  [D34D762D]  (AR v2)
Copy OK

Track  2

Filename C:\Music\Def Leppard - Pyromania\02 Def Leppard - Rock Brigade.wav

Peak level 100.0 %
Extraction speed 6.1 X
Track quality 100.0 %
Copy CRC E395BE90
Accurately ripped (confidence 67)  [AE3F859A]  (AR v2)
Copy OK

Track  3

Filename C:\Music\Def Leppard - Pyromania\03 Def Leppard - High 'N' Dry (Saturday Night).wav

Peak level 100.0 %
Extraction speed 6.3 X
Track quality 100.0 %
Copy CRC 1F112C52
Accurately ripped (confidence 67)  [F6D497FC]  (AR v2)
Copy OK

Track  4

Filename C:\Music\Def Leppard - Pyromania\04 Def Leppard - Another Hit And Run.wav

Peak level 100.0 %
Extraction speed 7.3 X
Track quality 100.0 %
Copy CRC CCD07EA1
Accurately ripped (confidence 67)  [372692C5]  (AR v2)
Copy OK

Track  5

Filename C:\Music\Def Leppard - Pyromania\05 Def Leppard - Billy's Got A Gun.wav

Peak level 100.0 %
Extraction speed 7.7 X
Track quality 100.0 %
Copy CRC 9D25E5FF
Accurately ripped (confidence 67)  [3BD33EDB]  (AR v2)
Copy OK

Track  6

Filename C:\Music\Def Leppard - Pyromania\06 Def Leppard - Mirror, Mirror (Look Into My Eyes).wav

Peak level 100.0 %
Extraction speed 8.1 X
Track quality 100.0 %
Copy CRC 922DFB5A
Accurately ripped (confidence 67)  [80EAD886]  (AR v2)
Copy OK

Track  7

Filename C:\Music\Def Leppard - Pyromania\07 Def Leppard - Foolin'.wav

Peak level 100.0 %
Extraction speed 7.5 X
Track quality 99.9 %
Copy CRC 970AFF5F
Accurately ripped (confidence 67)  [1FFD33F8]  (AR v2)
Copy OK

Track  8

Filename C:\Music\Def Leppard - Pyromania\08 Def Leppard - Photograph.wav

Peak level 100.0 %
Extraction speed 8.6 X
Track quality 100.0 %
Copy CRC 5D5784BF
Accurately ripped (confidence 67)  [BA61ED5C]  (AR v2)
Copy OK

Track  9

Filename C:\Music\Def Leppard - Pyromania\09 Def Leppard - Rock Of Ages.wav

Peak level 100.0 %
Extraction speed 8.1 X
Track quality 99.9 %
Copy CRC B390D08F
Cannot be verified as accurate (confidence 67)  [447C9337], AccurateRip returned [C9E97861]  (AR v2)
Copy OK

Track 10

Filename C:\Music\Def Leppard - Pyromania\10 Def Leppard - Bringin' On The Heartbreak.wav

Peak level 100.0 %
Extraction speed 9.4 X
Track quality 100.0 %
Copy CRC 5332E7CD
Accurately ripped (confidence 67)  [2C1C71BA]  (AR v2)
Copy OK

Track 11

Filename C:\Music\Def Leppard - Pyromania\11 Def Leppard - Switch 625.wav

Peak level 100.0 %
Extraction speed 9.4 X
Track quality 100.0 %
Copy CRC 7717B494
Accurately ripped (confidence 67)  [8090D617]  (AR v2)
Copy OK

Track 12

Filename C:\Music\Def Leppard - Pyromania\12 Def Leppard - Let It Go.wav

Peak level 100.0 %
Extraction speed 10.3 X
Track quality 100.0 %
Copy CRC AFC43C84
Accurately ripped (confidence 67)  [1DCB33F1]  (AR v2)
Copy OK

Track 13

Filename C:\Music\Def Leppard - Pyromania\13 Def Leppard - Wasted.wav

Peak level 100.0 %
Extraction speed 10.7 X
Track quality 100.0 %
Copy CRC A0D6C8EA
Accurately ripped (confidence 68)  [47AD39B7]  (AR v2)
Copy OK

Track 14

Filename C:\Music\Def Leppard - Pyromania\14 Def Leppard - Stagefright.wav

Peak level 100.0 %
Extraction speed 10.8 X
Track quality 100.0 %
Copy CRC 3014CB59
Accurately ripped (confidence 68)  [30BE69DF]  (AR v2)
Copy OK

Track 15

Filename C:\Music\Def Leppard - Pyromania\15 Def Leppard - Travelin' Band.wav

Peak level 100.0 %
Extraction speed 11.3 X
Track quality 100.0 %
Copy CRC 6F0B6620
Accurately ripped (confidence 68)  [A24563C5]  (AR v2)
Copy OK


14 track(s) accurately ripped
 1 track(s) could not be verified as accurate

Some tracks could not be verified as accurate

No errors occurred

End of status report

---- CUETools DB Plugin V2.1.6

[CTDB TOCID: 88nc2JKF6K5ieUfO3ntBG5YANdY-] found
Submit result: 88nc2JKF6K5ieUfO3ntBG5YANdY- has been submitted
Track | CTDB Status
  1  | (57/57) Accurately ripped
  2  | (56/57) Accurately ripped
  3  | (57/57) Accurately ripped
  4  | (57/57) Accurately ripped
  5  | (57/57) Accurately ripped
  6  | (57/57) Accurately ripped
  7  | (57/57) Accurately ripped
  8  | (57/57) Accurately ripped
  9  | (55/57) Differs in 115 samples
   @04:20:22,04:21:34,04:21:69,04:22:29-04:22:31,04:23:25,04:33:13
 10  | (57/57) Accurately ripped
 11  | (57/57) Accurately ripped
 12  | (57/57) Accurately ripped
 13  | (57/57) Accurately ripped
 14  | (57/57) Accurately ripped
 15  | (57/57) Accurately ripped
If you are sure that your rip contains errors, you can use CUETools to repair it.
  • Last Edit: 26 October, 2015, 07:25:50 PM by greynol

  • Nystagmus
  • [*][*]
High AccurateRip Confidence But Verification Fails
Reply #11
You might want to try ripping at 4x instead of 8x, assuming your drive supports it. 
8x is still fast enough to cause some errors due to speed.  Also, 115 samples isn't very much for errors.  They might not be audible.
But I can understand you wanting it to be zero.
  • Last Edit: 26 October, 2015, 05:39:49 PM by Nystagmus
Be a false negative of yourself!

  • korndawg
  • [*]
High AccurateRip Confidence But Verification Fails
Reply #12
kornchild:
Your wording indicates that you think one of the log entries indicates that you have an accurate rip?

In relation to the track in question, to be clear, you are not part of the "us".


This.


Hmm, I was under the impression that the "confidence 67" below indicates that my rip is identical to the rips of 67 other people, even though the AccurateRip checksum of my rip does not match the database:

Code: [Select]
Track  9

     Filename C:\Music\Def Leppard - Pyromania\09 Def Leppard - Rock Of Ages.wav

     Peak level 100.0 %
     Extraction speed 8.1 X
     Track quality 99.9 %
     Copy CRC B390D08F
     Cannot be verified as accurate (confidence 67)  [447C9337], AccurateRip returned [C9E97861]  (AR v2)
     Copy OK


Or have I been smoking too much crack???

  • korth
  • [*][*][*][*][*]
High AccurateRip Confidence But Verification Fails
Reply #13
Cannot be verified as accurate
That would be 0/67 (this track did not match)
Accurately ripped would mean a match

Quote
9  | (55/57) Differs in 115 samples ...

This track did not match but 55 entries differ by 115 samples
Accurately ripped would mean a match
  • Last Edit: 26 October, 2015, 06:27:47 PM by korth
korth

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
High AccurateRip Confidence But Verification Fails
Reply #14
That would be 0/67 (this track did not match)

You can't possibly know the 67 part based on the information provided.

It is also possible (albeit only remotely possible) that it could be 1/X as the database keeps 1-off checksums in limbo until there is a match from a different user ID.  I am not suggesting this is the case, however.  If it turned out that a unique user also submitted 447C9337, then subsequent matches would be shown with a confidence of 2.

In the event that there are different sets of AR checksums for a given track in a given record, the checksum and confidence that is shown to the user is that for the first entry in the record.
  • Last Edit: 26 October, 2015, 06:42:47 PM by greynol
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
High AccurateRip Confidence But Verification Fails
Reply #15
8x is still fast enough to cause some errors due to speed.

I'm not buying this for one second.

Also, FWIW the drive is clearly operating in CAV mode and is doing more like 25x by the time it's hitting this particular track.

Still, it wouldn't hurt to try ripping the track again with the speed throttled, though I wouldn't go as low as 8x, let alone 4x.  With the settings chosen, the actual ripping speed should be less than a third of whatever is selected.

In this case, simply re-ripping again without changing anything could give accurate results.  Ripping in burst mode may give accurate results as well (throttled or non-throttled).
  • Last Edit: 26 October, 2015, 07:11:40 PM by greynol
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • korndawg
  • [*]
High AccurateRip Confidence But Verification Fails
Reply #16
That would be 0/67 (this track did not match)

You can't possibly know the 67 part based on the information provided.

It is also possible (albeit only remotely possible) that it could be 1/X as the database keeps 1-off checksums in limbo until there is a match from a different user ID.  I am not suggesting this is the case, however.  If it turned out that a unique user also submitted 447C9337, then subsequent matches would be shown with a confidence of 2.

In the event that there are different sets of AR checksums for a given track in a given record, the checksum and confidence that is shown to the user is that for the first entry in the record.


Code: [Select]
     Cannot be verified as accurate (confidence 67)  [447C9337], AccurateRip returned [C9E97861]  (AR v2)


So the confidence 67 reflects the C9E97861 rip, not my 447C9337 rip? And my 447C9337 checksum has not been previously submitted by anyone, otherwise the confidence would be at least 2? If so the format is confusing as confidence 67 is listed right next to 447C9337...

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
High AccurateRip Confidence But Verification Fails
Reply #17
The unabridged log follows. Could you shed some light on the "Copy CRC" algorithm? Also, your reply begs the question of why EAC presents the checksum value in the first place. Thanks!

CRC32 for the raw PCM data only.

EAC presents a copy/read checksum so that you can compare with test checksum.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • korth
  • [*][*][*][*][*]
High AccurateRip Confidence But Verification Fails
Reply #18
That would be 0/67 (this track did not match)

You can't possibly know the 67 part based on the information provided.
[...]
In the event that there are different sets of AR checksums for a given track in a given record, the checksum and confidence that is shown to the user is that for the first entry in the record.

Edit: korndawg beat me to it.
So that would be the (confidence 67) found under [C9E97861] (ARv2) as shown in the log. That's what I meant by the 0/67.

Currently it would return (confidence 69) for that checksum
Quote
from my rip
09    [b3e169ce|c9e97861] (058+069/127) Accurately ripped
korth

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
High AccurateRip Confidence But Verification Fails
Reply #19
So that would be the (confidence 67) found under [C9E97861] (ARv2) as shown in the log. That's what I meant by the 0/67.

...and 0 out of any other set of submissions for that track in the publicly-available AR record (v1 or v2).

Then again, there might not be any other sets...

...yet.

  • Last Edit: 26 October, 2015, 07:50:17 PM by greynol
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
High AccurateRip Confidence But Verification Fails
Reply #20
So the confidence 67 reflects the C9E97861 rip, not my 447C9337 rip?

Yes.

And my 447C9337 checksum has not been previously submitted by anyone, otherwise the confidence would be at least 2?

It could have, but it would be in limbo until that same checksum gets submitted from a different user ID.

If so the format is confusing as confidence 67 is listed right next to 447C9337...

Yep.  Some of us came to that same conclusion a long while back.
  • Last Edit: 26 October, 2015, 08:06:54 PM by greynol
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • korndawg
  • [*]
High AccurateRip Confidence But Verification Fails
Reply #21
Thanks for all the answers, especially greynol! Cheers!

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
High AccurateRip Confidence But Verification Fails
Reply #22
You're welcome.

If you try again and happen to get a match, please let us know and how you managed it.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

  • Porcus
  • [*][*][*][*][*]
High AccurateRip Confidence But Verification Fails
Reply #23
Or have I been smoking too much crack???

Maybe just too much Korn?


kornchild korndawg
FTFY

And I might have had too much of Venom. Oh well.


If CUETools offers you to repair, as Greynol suggested, try that.

  • greynol
  • [*][*][*][*][*]
  • Global Moderator
High AccurateRip Confidence But Verification Fails
Reply #24
If CUETools offers you to repair, as Greynol suggested, try that.

With only 115 different samples over the entire album, I don't see why that can't be done.

I'd love to know more about why EAC blessed the track, and welcome additional attempts at ripping using different settings.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?