Skip to main content
Topic: AccurateRip version 1 vs version 2 CRCs inermixed? (Read 477 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

AccurateRip version 1 vs version 2 CRCs inermixed?

Hello all, long-time lurker here (more than 10 years now)! Hopefully this is the right subforum to post this in, if not, please feel free to move it.

For kicks, and to learn a new programming language, I've been working on a Rust re-implementation of the well-known ARFlac.pl script. For my testing, I ripped a well-known album, and downloaded the matching BIN file from the AccurateRip DB. However, after I generated version 1 and 2 CRCs, looking at the DB data, I see that the CRCs I'm calculating are mixed between different pressings (color coded red and blue for V1 and V2 CRCs, respectively):

Pressing 1:

    TrackResult { confidence: 122, crc: 4158045718 },
    TrackResult { confidence: 123, crc: 3175593300 },
    TrackResult { confidence: 125, crc: 1895033188 },
    TrackResult { confidence: 123, crc: 1209064292 },
    TrackResult { confidence: 123, crc: 751048154 },
    TrackResult { confidence: 122, crc: 2692720149 },
    TrackResult { confidence: 122, crc: 3342672821 },
    TrackResult { confidence: 119, crc: 41310113 },
    TrackResult { confidence: 121, crc: 3288026773 },
    TrackResult { confidence: 122, crc: 2772935668 },

Pressing 2:

    TrackResult { confidence: 119, crc: 3733010837 },
    TrackResult { confidence: 120, crc: 3824549311 },
    TrackResult { confidence: 119, crc: 1038071824 },
    TrackResult { confidence: 120, crc: 1723505091 },
    TrackResult { confidence: 122, crc: 841709511 },
    TrackResult { confidence: 121, crc: 3087785059 },
    TrackResult { confidence: 121, crc: 2819070029 },
    TrackResult { confidence: 119, crc: 4152591618 },
    TrackResult { confidence: 119, crc: 2834344192 },
    TrackResult { confidence: 120, crc: 1799856152 },

CRCs I Calculated:

    CRCs { v1: 3733010837, v2: 4158045718 },
    CRCs { v1: 3175593300, v2: 3824549311 },
    CRCs { v1: 1895033188, v2: 1038071824 },
    CRCs { v1: 1209064292, v2: 1723505091 },
    CRCs { v1: 841709511, v2: 751048154 },
    CRCs { v1: 3087785059, v2: 2692720149 },
    CRCs { v1: 2819070029, v2: 3342672821 },
    CRCs { v1: 41310113, v2: 4152591618 },
    CRCs { v1: 3288026773, v2: 2834344192 },
    CRCs { v1: 2772935668, v2: 1799856152 },


There are a few other pressings, but they have much lower confidence numbers and/or CRCs of zero. Based on my understanding from Spoon's excellent code samples and the AccurateRip wiki entry, I would have expected one pressing to be all red, and the other all blue. Is there something I'm not understanding correctly about the way AR data is stored?

Re: AccurateRip version 1 vs version 2 CRCs inermixed?

Reply #1
Your wording makes me ask a potentially stupid question:
Are you sure there are two pressings? Or is it one pressing, that has had both ARv1 data submitted from some rips, and ARv2 data submitted from other rips?
Memento: this is Hydrogenaudio. Do not assume good faith.

Re: AccurateRip version 1 vs version 2 CRCs inermixed?

Reply #2
Your wording makes me ask a potentially stupid question:
Are you sure there are two pressings? Or is it one pressing, that has had both ARv1 data submitted from some rips, and ARv2 data submitted from other rips?

To be totally honest, I'm not sure anymore! My impression was that for a given physical CD, V1 and V2 CRCs would be treated as different pressings entirely.

Re: AccurateRip version 1 vs version 2 CRCs inermixed?

Reply #3
I have ripped CDs in Linux and Windows where tracks that have failed in one OS andsucceded in another have the same audio MD5. It is strange.

Re: AccurateRip version 1 vs version 2 CRCs inermixed?

Reply #4
The order of the results are not important to AccurateRip, so intermingling is possible.

Re: AccurateRip version 1 vs version 2 CRCs inermixed?

Reply #5
The order of the results are not important to AccurateRip, so intermingling is possible.

Thank you, this is good to know! So does this mean that as long as that a given CRC (V1 or V2) for a track number N is found in the Nth position of any of the pressings, that the CRC has found a match, and the corresponding confidence is the confidence of that CRC?

Re: AccurateRip version 1 vs version 2 CRCs inermixed?

Reply #6
Correct

 
SimplePortal 1.0.0 RC1 © 2008-2018