My Experience Ripping a Damaged / Bad CD & How to possibly recover it (v. guide)

Stereodude

Not really a
Joined
Jan 22, 2002
Messages
10,509
Location
Michigan
I bought a used CD on eBay last weekend that arrived on Thursday along with a bunch of others. I couldn't rip two of the tracks successfully. The 3rd and 2nd from last tracks. I was getting Read and Sync errors with EAC even with my generally most forgiving optical drive drive (a Optiarc (NEC) ND-7200A I put in a USB enclosure). The CD had a few minor scratches, and one that looked a bit more serious. The more serious scratch was exactly concentric (the worse kind). I had the feeling based on lots of ripping experience that the damage to the disc wasn't exactly jiving with the problems I was having ripping it. I've ripped worse looking discs without issue.

However, I screwed around trying to polish it by hand with some Meguiar's PlastX, and made some headway minimizing the larger scratches, and was able to get the optical drive to read further into one the trouble tracks without error but ultimately was not successful. The disc also was starting to get lots of very fine scratches that were visible under the right lighting from the hand polishing and I wasn't making any forward progress. In fact it seemed to be getting worse. I tried ripping it with every optical drive I have installed in a PC (all 8 of them) to make sure it wasn't something specific to the two optical drives in my main desktop, but just couldn't get through it. All the other drives were worse. Some of the drives were old CD burners, some were DVD burners, and some were Blu-ray ROM/burners.

So I got ticked and decided to go with the nuclear option. I got out my Griot's Garage 3" random orbital polisher, the foam pads, and two different Menzerna polishes, 1500 and 4000. In relatively short order brought disc to a perfect glass like mirror finish with not even the slightest hint of any scratches on the bottom or any damage to the CD's reflective layer. It basically looks better than a brand new CD. The kicker, it still doesn't rip without Read and Sync errors. It's better than it was getting further into the problem track before puking with Read and Sync errors but it's basically a pristine disc that won't rip. The best ripping drive I had for damaged/scratched discs couldn't quite get through it (nor could the others). This wasn't the first time I've had what looks to be a completely pristine disc that just wouldn't rip though I think it had only occurred one or two other times out of around 1000 CDs.

I should note there is an odd slightly visible crescent shaped line that is visible in the reflective layer if you look at it just right with a flashlight that goes through the last 2 or 3 tracks on the disc. The shape of the curve appears to match the curvature of the outside of a CD. I'm guessing it's a manufacturing defect of some sort and doesn't correspond to any sort of physical damage to the disc. It's the only explanation I have for why the disc doesn't rip. Anyhow...

I had a disc or two in the past month of my used CD buying spree that ripped okay (according to the low level EAC functionality) with no reads or sync errors and generating a matching test & read CRCs that still report discrepancies with AccurateRip and CTDB for some of the tracks or in more than one of my optical drives. As a result I decided to learn about repairing the rip using the CTDB. EAC mentions it in a somewhat cryptic message. "If you are sure that your rip contains errors, you can use CUETools to repair it." I only learned about the CTDB because EAC added a plugin for it at some point. After ripping an entire disc EAC will give both AccurateRip and CTDB results whether your rip matches others of the same disc. Like this example:
Code:
- CUETools DB Plugin V2.1.6 -

[CTDB TOCID: gReDTXMGStGPtlSWZFW6DQg9Uzg-] found
Submit result: already submitted
Track | CTDB Status
  1   | ( 3/61) Accurately ripped, or (35/61) differs in 687 samples @05:19:56-05:19:57, or (3/61) differs in 235 samples @05:19:56, or (2/61) differs in 687 samples @05:19:56-05:19:57
  2   | ( 3/61) Accurately ripped, or (35/61) differs in 432 samples @04:58:63-04:58:64, or (3/61) differs in 154 samples @04:58:63, or (2/61) differs in 432 samples @04:58:63-04:58:64
  3   | ( 3/61) Accurately ripped, or (35/61) differs in 561 samples @04:28:34-04:28:35, or (3/61) differs in 210 samples @04:28:34, or (2/61) differs in 561 samples @04:28:34-04:28:35
  4   | (57/61) Accurately ripped
  5   | ( 3/61) Accurately ripped, or (35/61) differs in 346 samples @04:17:38-04:17:39, or (3/61) differs in 84 samples @04:17:38, or (2/61) differs in 346 samples @04:17:38-04:17:39
  6   | ( 3/61) Accurately ripped, or (35/61) differs in 482 samples @05:22:19-05:22:20, or (3/61) differs in 158 samples @05:22:19, or (2/61) differs in 482 samples @05:22:19-05:22:20
  7   | (58/61) Accurately ripped
  8   | ( 3/61) Accurately ripped, or (35/61) differs in 837 samples @04:40:26-04:40:27, or (3/61) differs in 280 samples @04:40:26, or (2/61) differs in 837 samples @04:40:26-04:40:27
  9   | (59/61) Accurately ripped
 10   | (59/61) Accurately ripped
 11   | ( 3/61) Accurately ripped, or (35/61) differs in 830 samples @02:57:64-02:57:65, or (3/61) differs in 262 samples @02:57:64, or (2/61) differs in 830 samples @02:57:64-02:57:65
 12   | ( 3/61) Accurately ripped, or (35/61) differs in 394 samples @04:09:48-04:09:49, or (3/61) differs in 138 samples @04:09:48, or (2/61) differs in 394 samples @04:09:48-04:09:49
 13   | ( 3/61) Accurately ripped, or (35/61) differs in 596 samples @04:52:16-04:52:17, or (3/61) differs in 187 samples @04:52:16, or (2/61) differs in 596 samples @04:52:16-04:52:17
 14   | (58/61) Accurately ripped
 15   | (60/61) Accurately ripped
If you are sure that your rip contains errors, you can use CUETools to repair it.

_________________________________

Track  1
     Peak level 100.0 %
     Track quality 99.8 %
     Accurately ripped (confidence 3)  [F272E721]  (AR v2)
     Copy OK
 
Track  2
     Peak level 100.0 %
     Track quality 99.8 %
     Accurately ripped (confidence 3)  [B529B1D0]  (AR v2)
     Copy OK
 
Track  3
     Peak level 100.0 %
     Track quality 99.8 %
     Accurately ripped (confidence 3)  [3EDE17E6]  (AR v2)
     Copy OK
 
Track  4
     Peak level 100.0 %
     Track quality 99.8 %
     Accurately ripped (confidence 25)  [8C78B547]  (AR v2)
     Copy OK
 
Track  5
     Peak level 89.1 %
     Track quality 99.9 %
     Accurately ripped (confidence 3)  [C867703F]  (AR v2)
     Copy OK
 
Track  6
     Peak level 100.0 %
     Track quality 99.9 %
     Accurately ripped (confidence 3)  [CAA5A2EC]  (AR v2)
     Copy OK
 
Track  7
     Peak level 100.0 %
     Track quality 99.8 %
     Accurately ripped (confidence 24)  [6610ACDA]  (AR v2)
     Copy OK
 
Track  8
     Peak level 100.0 %
     Track quality 99.8 %
     Accurately ripped (confidence 3)  [95C82C59]  (AR v2)
     Copy OK
 
Track  9
     Peak level 100.0 %
     Track quality 99.8 %
     Accurately ripped (confidence 24)  [FD89428E]  (AR v2)
     Copy OK
 
Track 10
     Peak level 99.9 %
     Track quality 99.8 %
     Accurately ripped (confidence 25)  [67964FDF]  (AR v2)
     Copy OK
 
Track 11
     Peak level 76.2 %
     Track quality 99.8 %
     Accurately ripped (confidence 3)  [072C4935]  (AR v2)
     Copy OK
 
Track 12
     Peak level 88.1 %
     Track quality 99.8 %
     Accurately ripped (confidence 3)  [B7A8F036]  (AR v2)
     Copy OK
 
Track 13
     Peak level 98.8 %
     Track quality 99.8 %
     Accurately ripped (confidence 3)  [3FA3A722]  (AR v2)
     Copy OK
 
Track 14
     Peak level 100.0 %
     Track quality 99.8 %
     Accurately ripped (confidence 24)  [FC69FC73]  (AR v2)
     Copy OK
 
Track 15
     Peak level 100.0 %
     Track quality 100.0 %
     Accurately ripped (confidence 24)  [D81E4F50]  (AR v2)
     Copy OK
 
 
All tracks accurately ripped
 
No errors occurred
 
End of status report
In the case of the the CTDB there is apparently some small amount of parity information stored for each disc, not just a checksum like the AccurateRip database. You can use CUETools and point it at your folder of FLAC files and have it do a verify for the tracks (just like EAC does for the disc). So it works like this. Open CUETools, point it to the folder with your FLAC files. Be sure to highlight the listing in the folder tree. You also have to give it an output path.


You can see the CTDB results are the same as EAC.

Now for the good part. CUETools can also repair the rip and make it consistent with the other rips by using the Encode option with the radio button and picking repair from the drop down writing new files in another folder. It looks like this:


Click the Go button and it gives you a popup to pick which rips you want to align with.


Click Ok and off it goes. It comes back with:


You can point it at the output directly and do a verify to make sure everything is good if you want. In the case of this example, this is the output log.
Code:
[CUETools log; Date: 4/21/2018 10:26:09 AM; Version: 2.1.6]
[CTDB TOCID: gReDTXMGStGPtlSWZFW6DQg9Uzg-] found.
Track | CTDB Status
  1   | (43/61) Accurately ripped, or (3/61) differs in 922 samples @05:19:56-05:19:57, or (3/61) differs in 687 samples @05:19:56-05:19:57
  2   | (43/61) Accurately ripped, or (3/61) differs in 586 samples @04:58:63-04:58:64, or (3/61) differs in 432 samples @04:58:63-04:58:64
  3   | (44/61) Accurately ripped, or (3/61) differs in 771 samples @04:28:34-04:28:35, or (3/61) differs in 561 samples @04:28:34-04:28:35
  4   | (57/61) Accurately ripped
  5   | (42/61) Accurately ripped, or (3/61) differs in 430 samples @04:17:38-04:17:39, or (3/61) differs in 346 samples @04:17:38-04:17:39
  6   | (40/61) Accurately ripped, or (3/61) differs in 640 samples @05:22:19-05:22:20, or (3/61) differs in 482 samples @05:22:19-05:22:20
  7   | (58/61) Accurately ripped
  8   | (43/61) Accurately ripped, or (3/61) differs in 1117 samples @04:40:26-04:40:27, or (3/61) differs in 837 samples @04:40:26-04:40:27
  9   | (59/61) Accurately ripped
 10   | (59/61) Accurately ripped
 11   | (43/61) Accurately ripped, or (3/61) differs in 1092 samples @02:57:64-02:57:65, or (3/61) differs in 830 samples @02:57:64-02:57:65
 12   | (42/61) Accurately ripped, or (3/61) differs in 532 samples @04:09:48-04:09:49, or (3/61) differs in 394 samples @04:09:48-04:09:49
 13   | (42/61) Accurately ripped, or (3/61) differs in 783 samples @04:52:16-04:52:17, or (3/61) differs in 596 samples @04:52:16-04:52:17
 14   | (58/61) Accurately ripped
 15   | (60/61) Accurately ripped
[AccurateRip ID: 00266f7c-01ae6d2a-ce0fb90f] found.
Track   [  CRC   |   V2   ] Status
 01     [5530368e|21f658b9] (2+3/5) Accurately ripped
 02     [b3069a59|ae72b7ab] (2+3/5) Accurately ripped
 03     [3ab85e12|cec086dd] (2+3/5) Accurately ripped
 04     [12a67b88|8c78b547] (2+4/6) Accurately ripped
 05     [2d059fda|04c5c9ea] (2+3/5) Accurately ripped
 06     [81ebfac3|0f7e3f2d] (2+2/4) Accurately ripped
 07     [43a15e09|6610acda] (2+4/6) Accurately ripped
 08     [38bf235d|0e7aebd0] (2+3/5) Accurately ripped
 09     [7c0a39fd|fd89428e] (2+4/6) Accurately ripped
 10     [1cc5eb17|67964fdf] (2+4/6) Accurately ripped
 11     [7fb2e1e9|68d8988a] (0+3/3) Accurately ripped
 12     [23e6337c|bd9ce96f] (2+3/5) Accurately ripped
 13     [85b48cfa|a559c147] (2+3/5) Accurately ripped
 14     [229ab53c|fc69fc73] (2+4/6) Accurately ripped
 15     [5729d1b5|d81e4f50] (2+3/5) Accurately ripped

Track Peak [ CRC32  ] [W/O NULL] 
 --  100.0 [F17EEFA4] [16A53167]           
 01  100.0 [764956C7] [E2496B56]           
 02   99.9 [8827B473] [5EE0C0EA]           
 03  100.0 [90932E7B] [58B77CD7]           
 04  100.0 [7CB3C067] [6F9171A9]           
 05   89.1 [DC443FC6] [A99FC5E8]           
 06  100.0 [F703851F] [6866FC4E]           
 07   99.9 [874A4501] [CD798B35]           
 08  100.0 [A4EA08C1] [BEFEC7E9]           
 09   99.9 [20AEB7D7] [BC3AA104]           
 10   99.9 [CA1F6A83] [AF62DEF9]           
 11   76.2 [8CC7D7BF] [08254BBA]           
 12   88.1 [69119001] [CBC872CF]           
 13   98.8 [3710851E] [6AEC2308]           
 14   99.9 [46F29777] [4B465926]           
 15   99.9 [32CD9C28] [2BF1009E]
So, back to my problem disc. I decided to just let drive sit and work on ripping it even with the Read and Sync errors. In order for the CUETools repair to work it needs a complete file. So I let EAC sit ripping despite the errors and it finished those two tracks in a few minutes with EAC reporting some read errors and suspicious positions. I fed it into CueTools per the above method and sure enough it was able to repair the rip.
Code:
[CUETools log; Date: 4/21/2018 10:45:40 AM; Version: 2.1.6]
[CTDB TOCID: o5r9F9lT4lO0to8Nqjak6r8L41U-] found.
Track | CTDB Status
  1   | (14/14) Accurately ripped
  2   | (13/14) Accurately ripped
  3   | (14/14) Accurately ripped
  4   | (14/14) Accurately ripped
  5   | (14/14) Accurately ripped
  6   | (14/14) Accurately ripped
  7   | (14/14) Accurately ripped
  8   | (14/14) Accurately ripped
  9   | (14/14) Accurately ripped
 10   | (14/14) Accurately ripped
 11   | (13/14) Accurately ripped
 12   | (13/14) Accurately ripped
 13   | (13/14) Accurately ripped
[AccurateRip ID: 002035a2-013d32e4-b30ff10d] found.
Track   [  CRC   |   V2   ] Status
 01     [c2e5c3fb|52feff13] (19+06/33) Accurately ripped
 02     [127a49b1|bd45aceb] (19+06/33) Accurately ripped
 03     [d49289af|4127835c] (20+06/34) Accurately ripped
 04     [3c72b527|30b9da82] (20+05/33) Accurately ripped
 05     [71c7b80b|62208fa8] (20+06/34) Accurately ripped
 06     [5c371023|e7f72a65] (20+06/34) Accurately ripped
 07     [f593c485|48b0dee1] (20+05/33) Accurately ripped
 08     [152a3478|a321c1b4] (19+06/33) Accurately ripped
 09     [6a08373a|d0b86777] (20+06/34) Accurately ripped
 10     [b37c6c55|11f657ac] (20+06/34) Accurately ripped
 11     [fcc5478a|6b0daf4e] (19+06/33) Accurately ripped
 12     [ab7260cf|e621896b] (19+05/32) Accurately ripped
 13     [e3d155bb|f19d7819] (16+04/27) Accurately ripped
Offsetted by -669:
 01     [434b2b7d] (06/33) Accurately ripped
 02     [65e1aa68] (06/33) Accurately ripped
 03     [7cd942fb] (06/34) Accurately ripped
 04     [3988b5dc] (06/33) Accurately ripped
 05     [af45bea5] (06/34) Accurately ripped
 06     [e3f4ba72] (06/34) Accurately ripped
 07     [ddab22eb] (06/33) Accurately ripped
 08     [2c3a2c90] (06/33) Accurately ripped
 09     [5687afbc] (06/34) Accurately ripped
 10     [e77b09b9] (06/34) Accurately ripped
 11     [651f8017] (06/33) Accurately ripped
 12     [b8c9a9cc] (06/32) Accurately ripped
 13     [ebcabd2e] (05/27) Accurately ripped
Offsetted by -664:
 01     [f44fdb81] (02/33) Accurately ripped
 02     [451d5a79] (02/33) Accurately ripped
 03     [0a652cab] (02/34) Accurately ripped
 04     [f30d208a] (02/33) Accurately ripped
 05     [23be3cdf] (02/34) Accurately ripped
 06     [5cc19ecc] (02/34) Accurately ripped
 07     [2bc3f261] (02/33) Accurately ripped
 08     [f92937b8] (02/33) Accurately ripped
 09     [2b0d6e2a] (02/34) Accurately ripped
 10     [dd868735] (02/34) Accurately ripped
 11     [203ac2a2] (02/33) Accurately ripped
 12     [def44067] (02/32) Accurately ripped
 13     [40aedaa3] (02/27) Accurately ripped

Track Peak [ CRC32  ] [W/O NULL] 
 --   99.0 [0F8BD730] [12E1420F]           
 01   99.0 [3422728F] [7F33C966]           
 02   99.0 [822CA0CA] [608333DE]           
 03   99.0 [E8670A05] [DCBA5DA6]           
 04   99.0 [893B53DF] [79513E33]           
 05   99.0 [61B2B160] [95CE9829]           
 06   99.0 [95991803] [DC575069]           
 07   99.0 [68C486F7] [15C9688F]           
 08   99.0 [EB9758F5] [5CFF297B]           
 09   99.0 [7AD5C9A0] [D4F8DDD5]           
 10   99.0 [62B8EF23] [6B2C1131]           
 11   99.0 [020216BB] [5CA56D61]           
 12   99.0 [17797144] [6CAA55CD]           
 13   99.0 [BE26A5ED] [3B83DCDF]
Success!!!

:bomb: :bomb: :bomb: :bomb: :bomb: :bomb:

CueTools_1.png CueTools_2.png CueTools_4.png CueTools_3.png
 

Stereodude

Not really a
Joined
Jan 22, 2002
Messages
10,509
Location
Michigan
After using CueTools a lot more in the months since starting this thread I have a few comments to add.

1) It is possible that EAC may rip damaged tracks in a way where they can't be recovered. If there are too many samples that vary from the reference it will simply say no match. It's also possible to get a "no match" rip of a track even though the disc is not that badly damaged. From empirical observation it appears that CueTool CTDB uses some sort of parity/redundancy information across the entire disc, not exactly individualized by track which is why it require a complete CD rip to work. If the bad rip of a track has too many or too few samples in a given track it will flag them as no match and can't repair them even if there's only one small problem area.

I have a disc that I've had for many years that has always given me problems when I tried to rip it. I was never able to read it perfectly. When I pointed CueTools to my rip from years ago it flagged several tracks as being no match. I retrieved the CD from the basement, but all my attempts to re-rip those tracks with EAC and use CueTools to repair them were not successful to get rips of those tracks that could be repaired. They would always come back as "no match" even though the drive didn't struggle all that much to read the disc. I almost bought another copy of the CD for ~$4 out of frustration.

I decided to try ripping the disc with CueRipper and was able to get a repairable rip of the disc. It reads the CD into one long continuous FLAC file with a cue sheet to delineate the individual tracks. It reported about 350 samples worth of discrepancies to other rips and I was then able to repair the rip with CueTools (and break it into individual FLAC files). My guess is that this process prevents any sort of sample creep at track boundaries or perhaps CueRipper is written in such a way to make sure it doesn't loose or add samples. It's also possible that ripping to a single continuous file with EAC with a CUE sheet might have worked equally well. I didn't try that.

2) Adjacent tracks can affect problem tracks. Replacing a problem track with a good rip of that track may not yield a complete disc that checks out. The differing track may shift to the preceding or following track. You may have to re-rip the entire disc. I think this is an offset issue. Effectively the disc and some of the tracks will have too many or too few samples. Proper offset for the drive correction should compensate for this if you use different drives, but replacing just a single track hasn't worked perfectly for me.

3) Combining tracks from two distinct rips of two unique physical copies of a given album may not yield a verified good album. Like rip 1 has the first 5 tracks (of 9) good and rip 2 has the last 4 (of 9) good. Taking the 9 good tracks between the two and sticking them together in a folder won't necessarily give you an album with 9 tracks that all check out good. I presume this is from having two distinct pressings, but can't say for sure.
 
S

staxsr007

Guest
I'm not seeing Repair as one of the options in the drop down menu, unless it is somewhere else? Thanks

 

snowhiker

Storage Freak Apprentice
Joined
Jul 5, 2007
Messages
1,507
The rise and fall of compact disc sales. Eventually the only way to obtain music CDs will be in the used market.

chartoftheday_12950_cd_sales_in_the_us_n.jpg
 

LunarMist

I can't believe I'm a
Joined
Feb 1, 2003
Messages
15,205
Location
USA
I suppose we are supposed to listen to the iphones like the Milleniums? Now they all have those drainage tubes from the eardrums.
I don't need audio except on the road or airlines. In a couple of years there will be hardly any cars anyway. We will really be screwed.
 

Stereodude

Not really a
Joined
Jan 22, 2002
Messages
10,509
Location
Michigan
The rise and fall of compact disc sales. Eventually the only way to obtain music CDs will be in the used market.

View attachment 1289
As long as they keep making them, I'll keep buying them. Once they stop making them, I stop buying music unless someone wants to sell me lossless audio files at a reasonable price. And, I currently think the lossy versions of albums are vastly overpriced, so I won't be holding my breath for $5 FLAC albums.
 

feelfreetoblameme

What is this storage?
Joined
Dec 1, 2018
Messages
2
As long as they keep making them, I'll keep buying them. Once they stop making them, I stop buying music unless someone wants to sell me lossless audio files at a reasonable price. And, I currently think the lossy versions of albums are vastly overpriced, so I won't be holding my breath for $5 FLAC albums.
I have purchased FLACs on Bandcamp from independent artists, who surprisingly will spend money on a vinyl and sometimes even cassette release but won't bother with a CD. Hopefully lossless becomes more prevalent. In the meantime, I absolutely love the used CD market, which continues to be flooded as people move on. I've probably tripled my collection in the last few years. But that also means getting more familiar with CueTools, EAC and all the other fun of ripping. Nice sense of accomplishment when a badly scratched 2 dollar CD yields all its audio goodness with the magic of EAC. That also of course means that I almost never play the CD itself since I can load FLACs onto a giant SD card and play those in the car. Convenience and access typically trumps quality when it comes to music, especially for the masses.
 
Top