FLAC In Hell's HellholeThe ripping "rules" / process
If anyone spots an issue with this, then please comment on it. Requires some good knowledge of EAC to understand what I'm talking about. Here's what happens:
CD gaps are detected (securely)
CD TOC is read.
Pregap is analysed (pregap is usually silence or possibly hidden audio before track 1). If pregap is silent then this is noted.
If Pregap contains audio then it will be ripped later and recorded as "00 - Pregap.wav".
No pregaps are inserted at the start of "01 - Track1.wav", they are either discared as silence (and noted for writing a cue later using PREGAP xx:xx:xx) or they are saved as "00 - Pregap.wav" as above.
CD is then tested with a range rip from block 0 to the end.
Blocks that could not be ripped (had errors) are recorded.
If errors were detected user can eject CD to clean it, and re-insert where it will attempt to test blocks that errored (avoids having to re-rip too much).
Once complete with 0 errors then CRC of the entire test is recorded.
Process then begins again ripping all blocks. During this ripping stage errors cannot occur - they should have been picked up and cleaned during testing. Rationale if errors start occuring again then this is really a flakey CD/drive so any rip is suspect. At the end the CRC's are compared and should match.
Rip is complete, now we go into post-rip mode. This will be highly configurable but at this stage the tracks will be split the same as EAC's Append Gaps To Previous Track (Default) method.
Notes:
Unless overread into leadin/leadout is specifically turned off it will ALWAYS try to overread. This is because some drives can/cannot overread sometimes depending on the media. If it fails due to what is likely a overread synch error then it will fall back and rip without the overread on the first/last blocks that are causing the error. If going out of the overread into lead in, or going into the overread of the lead-out there was not digital silence after the leadin/before the leadout then the rip will fail as accuracy of start/end cannot be determined without a drive that supports overreading as maybe some noise was chopped off?
It's anticipated some badly mastered CD's could have issues with this - it would specifically be down to individual case-by-case investigation where future versions of the ripping routine could have some exceptions hardcoded for specific CD's.
Overread can be turned off if this causes problems such as drive hanging/EAC crashing.
CD gaps are detected (securely)
CD TOC is read.
Pregap is analysed (pregap is usually silence or possibly hidden audio before track 1). If pregap is silent then this is noted.
If Pregap contains audio then it will be ripped later and recorded as "00 - Pregap.wav".
No pregaps are inserted at the start of "01 - Track1.wav", they are either discared as silence (and noted for writing a cue later using PREGAP xx:xx:xx) or they are saved as "00 - Pregap.wav" as above.
CD is then tested with a range rip from block 0 to the end.
Blocks that could not be ripped (had errors) are recorded.
If errors were detected user can eject CD to clean it, and re-insert where it will attempt to test blocks that errored (avoids having to re-rip too much).
Once complete with 0 errors then CRC of the entire test is recorded.
Process then begins again ripping all blocks. During this ripping stage errors cannot occur - they should have been picked up and cleaned during testing. Rationale if errors start occuring again then this is really a flakey CD/drive so any rip is suspect. At the end the CRC's are compared and should match.
Rip is complete, now we go into post-rip mode. This will be highly configurable but at this stage the tracks will be split the same as EAC's Append Gaps To Previous Track (Default) method.
Notes:
Unless overread into leadin/leadout is specifically turned off it will ALWAYS try to overread. This is because some drives can/cannot overread sometimes depending on the media. If it fails due to what is likely a overread synch error then it will fall back and rip without the overread on the first/last blocks that are causing the error. If going out of the overread into lead in, or going into the overread of the lead-out there was not digital silence after the leadin/before the leadout then the rip will fail as accuracy of start/end cannot be determined without a drive that supports overreading as maybe some noise was chopped off?
It's anticipated some badly mastered CD's could have issues with this - it would specifically be down to individual case-by-case investigation where future versions of the ripping routine could have some exceptions hardcoded for specific CD's.
Overread can be turned off if this causes problems such as drive hanging/EAC crashing.
posted by FlacInHell at 9:34 pm
3 Comments:
Cool, I look forward to trying this out. :)
I know a specific "hub" that will be using this as soon as it comes out. if you want start a thread over there and give us some updates about it. we'll definately help you bug test and beta test it. a few people in the community may even be able to help code parts of it. ;-)
I am also looking forward to this App. It will be great to give a copy to people who are ripping for the first time. A new generation of quality rips :-)
Post a Comment
<< Home