2

I'm involved in a project where we are ripping thousands of CD-Rs that contain music. These are (usually) one off recordings and not in any music database etc. but we need a way to uniquely identify them when ripping in batches.

The most affective process we've found so far is through generating a CDDB ID for each disc, similar to MusicBrainz (minus the hashing) - https://musicbrainz.org/doc/Disc_ID_Calculation. However, as the article and anyone that's ripped CDs before will know, different CDs can produce the same IDs if the data matches.

Does anyone know of a more robust way of uniquely identifying a CD-R or the contents contained on it? It needs to be something that, when the disc is read in a drive it produces the same result each time for later comparison.

Do burners perhaps record date/time data is written to the disc or even an ID for the burner used?

I've been using cdrtools to look at disc metadata but there seems to be very little unique info I can pull from a CD-R. I'm thinking generating an md5/sh256 checksum for each disc will work, but it takes quite long time. The only other unique ID I'm aware of is the matrix printed on the inner ring of the disc (not practical).

Tom
  • 21

1 Answers1

2

Some of the more "enthusiast" tools for ripping actually do that, and audit your rips as you do. Dbpoweramp has a system called accuraterip and the open source cuetools suite supports that and its own format as well

There's a list of tools that support that - including a command line one on hydrogenaudio.

You don't need a md5 or sha checksum - you're not worried about someone altering the data - and most tools use a simple CRC with some other logic.

There's no way around reading and checksumming the entire drive but it might be useful to use what's essentially a standard, well supported way to do so.

Journeyman Geek
  • 133,878