Page 1 of 1

Ripping Strategy

Posted: Sun Nov 30, 2008 5:52 pm
by sbooth
OK , I finally dusted off my OmniGraffle skills and created the flowcharts illustrating how Rip's audio extraction engine works:

Re: Ripping Strategy

Posted: Sun Nov 30, 2008 9:42 pm
by elduderino
Thanks Stephen, this is interesting indeed. I am excited, the strategy seems to make sense to my naive thinking anyway :-) So, if I understand you correctly, you feel that this strategy is superior to a strategy that is based purely on AccurateRip??? Do you use cdparanoia anywhere in the new strategy, or is it entirely based on the new comparison ripper and Accurate Rip? I may be confusing things here, but set me straight if you would.

Thanks again, very excited to get ripping and testing!

Re: Ripping Strategy

Posted: Sun Nov 30, 2008 10:05 pm
by sbooth
There is no cdparanoia in Rip- I think that drives have come a long way since that code was written and that grinding away on individual sectors isn't the best way to go. In my testing (although only on 4 different drives) my strategy seems to be generally successful. So all the ripping code is my own.

Rip isn't 100% based on AccurateRip for two reasons:
  • Not all discs are present in the AR database
  • Some discs may contain errors that would prevent AR verification, but users may still want to extract the audio

Re: Ripping Strategy

Posted: Sun Nov 30, 2008 10:20 pm
by Fuga
No paranoia. Stephen is on record saying he questions the efficacy of paranoia's demands on drives.

Now, as to what the "engine" behind the comparison ripper is, well, I can't recall from the lengthy thread on secure ripping. I think it's pretty much a "Finder copy" relying on C2 error technology. The big thing here is AR, best I can tell.

C2 + AR + matching cheksums = secure rip?

Once things settle and initial Rip bugs are worked out perhaps then Stephen can take time to flesh out the chart for us.

Re: Ripping Strategy

Posted: Sun Nov 30, 2008 10:36 pm
by elduderino
Got it, thanks guys...I look forward to testing.