Ripping Anomalies

Posted by: endlessnessism on 09 January 2014

It's odd that when you're ripping a CD, in dBpoweramp or the HDX or whatever else you're using, the online databases will sometimes produce a result that bears no relationship to the CD you think you're ripping.  I have just had it with a Georgie Fame CD - two out of three databases were convinced it was some live comedy thing and only freedb recognised it for what it was.  I can understand the databases not recognising a CD at all and producing an "unknown disc" result but why would they think it was something completely different?  Do manufacturers re-write discs that they haven't been able to sell, and forget to change the metadata?

Posted on: 10 January 2014 by mrspoon

CDs themselves do not normally contain metadata (CD Text disc do, but they are in the minority)

 

So the online databases have to use the track lengths to match a disc to metadata, as you can imagine with more than 2M unique discs there are collisions where 2 different discs have exactly the same length. CD Singles are the worst.

Posted on: 10 January 2014 by Narooma

That's interesting, I had no idea that that was how it's done. I can see that that would be the best way to recognise older releases but surely it would simple to encode 'modern' CDs with a unique catalogue number along the lines of a bar code?

Posted on: 10 January 2014 by Redmires

I use EAC which often reports several versions of a CD. Often it can just be the changes in capitisation of the tracks, such as "All my loving" or "All My Loving" . I've also noticed that N-Stream sometimes selects the wrong CD when I click on the info button for more detailed notes.

 

On a similar matter, I've been copying my vinyl to HDD and using MP3tag to tag the FLAC files. When I play one in N-Stream, it will find the correct detailed notes of the album. Clever stuff.

 

Posted on: 10 January 2014 by George J

A weird thing.

 

I transferred my remaining vinyl to CD in about 2006, and amazingly iTunes [Gracenote data-base] recognised the recordings of Schnabel and Fournier playing Beethoven's Cello Sonatas like that!

 

To thins day I have never been able to explain that oddity. Of course the German EMI [Electrola] had issued both the LPs and a very rare CD issue of the same recording three decades later ...

 

I must have been extremely lucky to insert the tracking points with supreme accuracy, and also the Rega P3 must have been playing at exactly the right speed!

 

ATB from George

Posted on: 10 January 2014 by Harry

I've had some hilarious howlers but most default to Unknown Artist. I expected this with some of my CDs and the failure rate is about right. It takes little time and trouble to make the corrections and on one level it's quite nice to be in full control of what gets entered into the database, although anything can be modified retrospectively of course.

Posted on: 13 January 2014 by TommayCat

I like to stay on top of the metadata (some of it is really sloppy), especially if the errors are going to make through to the directory and filenames themselves.  It also provides a level of control over the way the album is stored in your library.  I will sometimes rip a CD containing multiple composers using different versions of the metadata to produce two separate 'albums', i.e. Holst Planets and Elgar Cello Concerto on the same CD by the same conductor. I rip the Planets tracks with artist metadata as Holst and Cello tracks with Elgar as artist. Composer metadata will also achieve this, but it is the artist that determines part of the filename.  Not a faithful copy of the CD I grant you, but so much easier to listen to, which surely is the point?