CRC checksum number

[Seppe]

Hello

Doesn't a song remain a song?

Why, or how could the same song (from anywhere in the world) produce a different crc / checksum number?


Thanks

[Spoon]

Re: CRC checksum number

Errors in the song give different CRC values.

[mville]

Re: CRC checksum number

Why, or how could the same song (from anywhere in the world) produce a different crc / checksum number?

Do you mean same song, same CD?

Tracks can often have different digital masters depending on the territory/country of release.

[Seppe]

Re: CRC checksum number

Hi Spoon / mville

I mean the same song, not cd.

Doesn't a song remain a song i.e. the same bits? Because if you alter that - the bits - by error or otherwise, don't you in essence change the song to sounding different - lower/higher base, different pitch or note, different quality of recording etc.

I'm just wondering why confidence levels could not be figures in the thousands or perhaps even hundreds of thousands considering AR is a global database and if a song is released sounding like "Y", then the crc can always only be "X". Not so? If different territories could have different masters, do they sound different?

Thanks

[garym]

Re: CRC checksum number

Hi Spoon / mville

I mean the same song, not cd.

Doesn't a song remain a song i.e. the same bits? Because if you alter that - the bits - by error or otherwise, don't you in essence change the song to sounding different - lower/higher base, different pitch or note, different quality of recording etc.

I'm just wondering why confidence levels could not be figures in the thousands or perhaps even hundreds of thousands considering AR is a global database and if a song is released sounding like "Y", then the crc can always only be "X". Not so? If different territories could have different masters, do they sound different?

Thanks

AR cuts of the reporting info on number of matches at 200, regardless of number of matches above 200.

Also, you may not be fully understanding how acccuraterip works. First, an AR of 50 is not actually any better than an AR of 2. The odds of two other people in the world ripping a different physical copy of the CD on a different computer using a different drive and generating the same bits as your rip, and this NOT being bit perfect are astronomically small. There is a greater change that I'll wake up with wings and a tail tomorrow. Moving from "2" to "50" is just confirming what you already knew.

Also, If AR provides a match, then it is a match. If your CD is from one territory (with a different master) it won't match the database entry from the other territory (with different master and different "bits"). Certainly different masters *can* sound different. That's why they are different masters. But AR is matching your rip to other rips of the SAME underlying music.

[Seppe]

Re: CRC checksum number

The fact that everybody is not mastering from the same source with perhaps differing techniques/processes which then can produce different checksums makes it clearer for me. Would be cool if the music industry could standardise to best codes of practice though. Thanks for all the replies.

[mville]

Re: CRC checksum number

Would be cool if the music industry could standardise to best codes of practice though.

Often, the decision to have a slightly different mix depending on territory is a creative decision, not bad practice.

[Seppe]

Re: CRC checksum number

Often, the decision to have a slightly different mix depending on territory is a creative decision, not bad practice.

Interesting. Didn't know that actually happened. Thanks.

However, I was thinking more along the lines of standardisng production so that manufacturers get the same master (creative changes for territories accounted for). Then perhaps record labels could make available the crc for matching. That way not only could you confirm an accurate rip but also an accurate song (as intended upon release). Not sure if my thinking is correct though.

[garym]

Re: CRC checksum number

Interesting. Didn't know that actually happened. Thanks.

However, I was thinking more along the lines of standardisng production so that everybody gets the same master (creative changes for territories aside). Then perhaps record labels could make available the crc for matching. That way not only could you confirm an accurate rip but also an accurate song (as intended upon release). Not sure if my thinking is correct though.

this *could* be possible, but is unlikely. But the good news is that the ACCURATERIP database is just what you want the record labels to produce. A way of comparing your own CRC to the "correct" CRC. AR database essentially does this for you.

Edit: By the way, "record companies" have no interest in making ripping digital bit perfect copies from CDs easier or more error free. If they had their way, it would be impossible to rip digital files from CDs. Furthermore, even if they are selling you digital downloads, they gain nothing by providing CRCs that folks in the "illegal file trading" area could use to easily confirm that their file copies obtained from the shadows are in fact bit perfect back to the originals.