Tweet
Hi!
A long time ago I wrote a CD ripper with AccurateRip checking, it must have been 15 years now. Recently I picked it up again, and started remembering some things about matching the calculated CRCs with the pressings in the database.
When ripping a particular disc, my tool calculates, and reports the ARv1 CRCs as follows:
When ripping this same disc with EAC (version 0.99, it seems to be the last version to do ARv1):
The last track has the same CRC, but my tool considers it unverified, while EAC says it's accurate.
I then inspected the database entry for this disc (https://www.accuraterip.com/accurate...5-a30dea0d.bin), and parsed it. It contains 21 different pressings for this disc, 19 of which do not match with the calculated CRC on any of the tracks. But two of them look like this:
And it becomes clear, EAC verifies tracks 1-12 using pressing 8, and track 13 using pressing 7.
I then tried CueRipper from CueTools, a current version. It does some strange things, but still mixes the pressings I think:
I know CueRipper calculates both ARv1 and ARv2 CRCs. But the one shown is only the ARv1. Also, it appears to add up the confidences of the ARv1 and ARv2 matches (the ARv2 results from this disc all match the same pressing and have a confidence of 89 (for all but two tracks), 89 + 17 = 106). So it too seems to take the CRCs from both pressing 7 and 8 to verify the rip.
Is this the normal expected way to match calculated CRCs to the database? I would have thought you needed to pick a single pressing to match against. Is this specific to ARv1, or also applicable to ARv2? Why group the CRCs by pressing at all then, and not just have a list of crcs per track? I realize newer versions of EAC calculate ARv2 instead of ARv1 (and ARv1 is deprecated?), but is the matching still done the same way?
Thanks!
A long time ago I wrote a CD ripper with AccurateRip checking, it must have been 15 years now. Recently I picked it up again, and started remembering some things about matching the calculated CRCs with the pressings in the database.
When ripping a particular disc, my tool calculates, and reports the ARv1 CRCs as follows:
Code:
AccurateRip (v1) results: Track 01 accurately ripped (CRC: A2A9078E Confidence: 17) Track 02 accurately ripped (CRC: 82619103 Confidence: 17) Track 03 accurately ripped (CRC: 89BDEE86 Confidence: 17) Track 04 accurately ripped (CRC: 5E31E134 Confidence: 17) Track 05 accurately ripped (CRC: 66082ED8 Confidence: 17) Track 06 accurately ripped (CRC: EBB78E4F Confidence: 17) Track 07 accurately ripped (CRC: 1961FAB1 Confidence: 17) Track 08 accurately ripped (CRC: 49EF9717 Confidence: 17) Track 09 accurately ripped (CRC: C201195D Confidence: 17) Track 10 accurately ripped (CRC: F49C3B72 Confidence: 17) Track 11 accurately ripped (CRC: F4AC07C6 Confidence: 17) Track 12 accurately ripped (CRC: 311741AC Confidence: 17) Track 13 could not be verified as accurate (calculated CRC: 86640AD0)
Code:
Accurately ripped (confidence 17) [A2A9078E] Accurately ripped (confidence 17) [82619103] Accurately ripped (confidence 17) [89BDEE86] Accurately ripped (confidence 17) [5E31E134] Accurately ripped (confidence 17) [66082ED8] Accurately ripped (confidence 17) [EBB78E4F] Accurately ripped (confidence 17) [1961FAB1] Accurately ripped (confidence 17) [49EF9717] Accurately ripped (confidence 17) [C201195D] Accurately ripped (confidence 17) [F49C3B72] Accurately ripped (confidence 17) [F4AC07C6] Accurately ripped (confidence 17) [311741AC] Accurately ripped (confidence 17) [86640AD0] All tracks accurately ripped
I then inspected the database entry for this disc (https://www.accuraterip.com/accurate...5-a30dea0d.bin), and parsed it. It contains 21 different pressings for this disc, 19 of which do not match with the calculated CRC on any of the tracks. But two of them look like this:
Code:
Pressing 7 Track 1 : 20 (8765f22c) Track 2 : 20 (3789c07a) Track 3 : 19 (fc162d13) Track 4 : 21 (fa6ce6a) Track 5 : 21 (dcaab63f) Track 6 : 21 (ce9ba59) Track 7 : 21 (558e3a96) Track 8 : 20 (679ba163) Track 9 : 20 (54a6547c) Track 10 : 20 (56363cc3) Track 11 : 20 (738f8f) Track 12 : 17 (75159c89) Track 13 : 17 (86640ad0) Pressing 8 Track 1 : 17 (a2a9078e) Track 2 : 17 (82619103) Track 3 : 17 (89bdee86) Track 4 : 17 (5e31e134) Track 5 : 17 (66082ed8) Track 6 : 17 (ebb78e4f) Track 7 : 17 (1961fab1) Track 8 : 17 (49ef9717) Track 9 : 17 (c201195d) Track 10 : 17 (f49c3b72) Track 11 : 17 (f4ac07c6) Track 12 : 17 (311741ac) Track 13 : 16 (569a4790)
I then tried CueRipper from CueTools, a current version. It does some strange things, but still mixes the pressings I think:
Code:
AccurateRip summary Track 1 accurately ripped (confidence 106) [A2A9078E] Track 2 accurately ripped (confidence 106) [82619103] Track 3 accurately ripped (confidence 106) [89BDEE86] Track 4 accurately ripped (confidence 106) [5E31E134] Track 5 accurately ripped (confidence 106) [66082ED8] Track 6 accurately ripped (confidence 106) [EBB78E4F] Track 7 accurately ripped (confidence 106) [1961FAB1] Track 8 accurately ripped (confidence 106) [49EF9717] Track 9 accurately ripped (confidence 106) [C201195D] Track 10 accurately ripped (confidence 106) [F49C3B72] Track 11 accurately ripped (confidence 108) [F4AC07C6] Track 12 accurately ripped (confidence 107) [311741AC] Track 13 accurately ripped (confidence 106) [86640AD0] All tracks accurately ripped
I know CueRipper calculates both ARv1 and ARv2 CRCs. But the one shown is only the ARv1. Also, it appears to add up the confidences of the ARv1 and ARv2 matches (the ARv2 results from this disc all match the same pressing and have a confidence of 89 (for all but two tracks), 89 + 17 = 106). So it too seems to take the CRCs from both pressing 7 and 8 to verify the rip.
Is this the normal expected way to match calculated CRCs to the database? I would have thought you needed to pick a single pressing to match against. Is this specific to ARv1, or also applicable to ARv2? Why group the CRCs by pressing at all then, and not just have a list of crcs per track? I realize newer versions of EAC calculate ARv2 instead of ARv1 (and ARv1 is deprecated?), but is the matching still done the same way?
Thanks!
Comment