Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Red Book cd error correction... (Read 12406 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Red Book cd error correction...

I don't really understand how the audio cd format works...

I have read that there is a lot of error correction bytes that will correct most errors caused by scratches and dirty cds...
If the errors can't be fixed, the "most likely data to be in the sector" will be used...

But if just about all errors are corrected like this, how does programs like EAC (that I use) are detecting errors? What are those errors??

Well... How does all this work?...


Thanks a lot

Alex

Red Book cd error correction...

Reply #2 –
Quote from: AlexandreG on
But if just about all errors are corrected like this, how does programs like EAC (that I use) are detecting errors? What are those errors??

The short answer is that EAC reads everything twice, and if they match then it is assumed that there were no errors.

Edit: No UNCORRECTED erors that is.

Red Book cd error correction...

Reply #3 –
Aside from making an attempt to fix problems with synchronization, ripping programs do not and cannot correct errors.  It is all done within the drive.

All data coming from a drive is either error-free, corrected to be error-free or otherwise synthesized if it cannot be corrected to be error-free.  It is not at all uncommon for a drive to synthesize data in a consistent fashion.

Besides ripping everything twice, there are other methods of determining if the data produced by a drive is error-free.  They include making use of C2 pointers and accessing a database of known hash values for the tracks being ripped (AKA AccurateRip).  EAC is capable of doing both of these either instead of, or in addition to reading everything more than once.  There are still other methods that can be used to attempt to see if your drive is synthesizing data as well.  Aside from drives that fail to provide reliable C2 error information, checking data for inconsistencies is probably the least reliable method of detecting errors.

Red Book cd error correction...

Reply #4 –
correct me if I am wrong...

The error correction is made by the drive hardware.

So, is there is an error, the drive itself will try to correct it. In the case of an audio cd (it's not the same for data because you can't interpolate...) is there is no way to correct the error, the drive will "guess" the good value...

I understand that reading everything twice will look for problems because errors won't usually be the same every time... (WHY???????) so the drive will guess slightly different values every time...

Is this right?
 

Red Book cd error correction...

Reply #5 –
Quote from: AlexandreG on
The error correction is made by the drive hardware.

I just said this, even prior to adding the last paragraph in my edit.

Quote from: AlexandreG on
(it's not the same for data because you can't interpolate...)

Who says you can't interpolate?  This is generally what drives do and at least one ripping program as well.

Quote from: AlexandreG on
I understand that reading everything twice will look for problems because errors won't usually be the same every time...

It depends on the drive and the severity of the error.  It's quite easy and quite common for a drive to produce the same erroneous data more than once.

When drives can't (or don't) interpolate, they either hold data at the last known good value until good data comes along or replace the missing data with null samples.  If someone knows another way feel free to comment.

Red Book cd error correction...

Reply #6 –
Well... if you interpolate on non correctable errors with a data cd... it would cause a lot of problems!... with a data cd, the actual data need to be exactly identical to the original... it is yes or no... not "almost and so close that you can't hear it" like with audio cds with uncorrectable errors...

Red Book cd error correction...

Reply #7 –
Quote from: AlexandreG on
Well... if you interpolate on non correctable errors with a data cd... it would cause a lot of problems!... with a data cd, the actual data need to be exactly identical to the original... it is yes or no... not "almost and so close that you can't hear it" like with audio cds with uncorrectable errors...


yes, this is why data CD's have a completely different format on the CD. mainly more robbust error detection and error correction codes. this means that i can't hold wuite as much data, but that data is far less likely to have an unrecoverable error.

Red Book cd error correction...

Reply #8 –
Since CD drives don't use interpolation  while extracting audio data, is it correct to assume that when error correction is performed during ripping, the corrected output will be the same as the output produced when there are no errors at all? Or can CD drives return "corrected" data that is not accurate?

Red Book cd error correction...

Reply #9 –
if the errors are really not bad and are CORRECTED by the drive, it is as good as a perfect one

If there is the need to interpolate, of course, the results could (will?) be wrong...

Red Book cd error correction...

Reply #10 –
Quote from: AlexandreG on
Well... if you interpolate on non correctable errors with a data cd... it would cause a lot of problems!...
That's correct.  Sorry, I misread earlier.

Quote from: dgauze on
Since CD drives don't use interpolation  while extracting audio data
This is not correct.  Most CD drives interpolate when encountering uncorrectable errors while extracting audio data unless the data stream is too far messed up.

Quote from: dgauze on
is it correct to assume that when error correction is performed during ripping, the corrected output will be the same as the output produced when there are no errors at all?
If the errors are corrected, then yes of course; by definition.  The data will be error-free.

Quote from: AlexandreG on
If there is the need to interpolate, of course, the results could (will?) be wrong...
Except for coincidence.
In all likelihood the results will be wrong.  Otherwise we're talking about coincidence.
I suppose it's possible that the data could interpolate and be correct (clipped samples or silence are obvious examples).  I'm not sure how common it is to see three samples in a row form a straight line in music originating from analog sources.

Red Book cd error correction...

Reply #11 –
I think I got it

Last question... If I don't use AccurateRip... there is no way to have a real "prefect" rip (well, at least to know that it is).

It would be possible with C2 errors reports from the drive, but from what I saw, it isn't reliable at 100%...


But I have 2 drives... am I wrong if I say that when I extract audio with both drives, it the crc is the same, the rip should be perfect??

Thanks

Alex

Red Book cd error correction...

Reply #12 –
Quote from: AlexandreG on
But I have 2 drives... am I wrong if I say that when I extract audio with both drives, it the crc is the same, the rip should be perfect??


You reach 'very probable' territory when two different discs show two identical CRCs on two different drives..

Red Book cd error correction...

Reply #13 –
Quote from: rpp3po on
You reach 'very probable' territory when two different discs show two identical CRCs on two different drives..

Now you're getting into AccurateRip territory.

Quote from: AlexandreG on
It would be possible with C2 errors reports from the drive, but from what I saw, it isn't reliable at 100%...

But I have 2 drives... am I wrong if I say that when I extract audio with both drives, it the crc is the same, the rip should be perfect??

If the drives are based on two different chipsets then I'd say you're in good shape.  Swapping modes with the same drive (burst + secure no C2 or secure C2 + secure no C2) and getting matching CRCs is better than C2 with matching CRCs which is better than C2 without matching CRCs, though I am more skeptical with some chipsets in this scenario than others when it comes to usage of C2 pointers.

Red Book cd error correction...

Reply #14 –
Quote from: greynol on
Quote from: AlexandreG on
If there is the need to interpolate, of course, the results could (will?) be wrong...
Except for coincidence.
In all likelihood the results will be wrong.  Otherwise we're talking about coincidence.
I suppose it's possible that the data could interpolate and be correct (clipped samples or silence are obvious examples).  I'm not sure how common it is to see three samples in a row form a straight line in music originating from analog sources.

Yeah. This reminds me of the Cactus Data Shield Copy Protection, where it corrupts bits of data, especially clipped samples.

Both my internal Cyberdrive and LG could never rip my Coldplay X&Y. They gave completely different CRCs each time in Burst Mode. In Secure Mode, the speed would drop to 0.2x so I didn't bother waiting for the results.

Tried with my Sony laptop's internal drive and it couldn't rip consistently either.

Recently, I got a new LG external drive and I thought I'd give it a try. Amazingly, it worked! I was rather surprised when I saw Accurate Rip giving a confidence of 200.
But again, it wasn't very consistent. If I re-ripped the track, sometimes, it would give another CRC.