Any Guides, Stickys, etc on best practices when trying to recover "problem" CD's?

[SerbJ]

I have been using dBpoweramp over the last week or so, to re-rip my modest CD collection. I had been using EZCD prior to that. After getting thru about 700 CDs, I'm now down to a couple dozen "problem" CDs. Some I have been able to recover by just opening up the amount of frames before failure. Others, by examining the details of the failures, determined the insecure timeframe is too small that probably won't be heard. To some extent, feel I'm just wading around the pool trying to learn how to swim on my own. Not always a bad thing, but was wondering if there was a "best" practices or similar document to speed the learning time. I did some searching and found a nugget or two, but not the pot of gold. Any help would be appreciated.

[Spoon]

At the end of the day there will be discs which cannot be recovered, only a new cd will help.

[schmidj]

The used CD market can be your friend. But beyond that, listen to the file of the ripped CD. Does it sound OK to you? Then that is all which should really matter. It is very nice to have a rip that passed accuraterip. You don't have to worry about the quality of the rip or take the time to listen to the file while you are ripping. But unless you are obsessive compulsive, it isn't the end all. The goal should be that your rips should sound acceptable when you are listening. Don't forget that when you listen to vinyl you expected clicks and pops. They were part of the process. Most (not all!) of us wish they weren't there. (in fact now we have software that does a pretty good job of removing them and might well remove the clicks in a bad rip.) But we listened to records and didn't fret too much over the clicks.

Another hint, keep a collection of drives of different makes and models, A CD that fails accuraterip in one drive may well rip perfectly in another. Salvage drives out of old PCs about to go to the electronic waste collection. They are usually free and you may well find one that will read that troublesome CD (but fail on others that your current drive reads well.

[SerbJ]

At the end of the day there will be discs which cannot be recovered, only a new cd will help.

Yea, not expecting miracles, just the best approaches, etc. There was ample online documentation on how the data is stored and arranged on Music CDs. That helped me to understand what a "frame" was.

I see a number of 13-14ms spans of insecure audio. I have it set to interpolate, so it is very unlikely that it would be heard. Especially since most of the music is rock. But, at what length spans that it does become problematic? 300-400ms? I know, probably subjective, but in general.

Is there any way to re-use the data from a prior rip, so the re-rip will work on primarily just the insecure frames, vs re-reading the whole track and trying again? I have gotten some large 300-400 frame errors down to a handful of small gaps. It has been trial and error to get the timeout and number of frame errors correct when opening it to that large of frames, then finding you fall just short.

You can look at some CDs and see numerous scratches. Sometimes you wonder how you got a clean rip, they are so bad! Other times the CD can look very clean, but one track will have unusually high errors. You don't see any pinholes, either. I've read about some type of material break-down on older CDs, maybe that is it? Now, the ones with circular damage from surface contact in the drive, are pretty much impossible.

I was also interested in how effective others had found with waxing? I tried it a year or so ago, but didn't see much improvement, but that was with EZCD, not dBpoweramp. Also tried "polishing", but just destroyed that CD.

If this isn't where this should be discussed, just looking to where that might be.

[SerbJ]

The used CD market can be your friend. But beyond that, listen to the file of the ripped CD. Does it sound OK to you? Then that is all which should really matter. It is very nice to have a rip that passed accuraterip. You don't have to worry about the quality of the rip or take the time to listen to the file while you are ripping. But unless you are obsessive compulsive, it isn't the end all. The goal should be that your rips should sound acceptable when you are listening. Don't forget that when you listen to vinyl you expected clicks and pops. They were part of the process. Most (not all!) of us wish they weren't there. (in fact now we have software that does a pretty good job of removing them and might well remove the clicks in a bad rip.) But we listened to records and didn't fret too much over the clicks.

Another hint, keep a collection of drives of different makes and models, A CD that fails accuraterip in one drive may well rip perfectly in another. Salvage drives out of old PCs about to go to the electronic waste collection. They are usually free and you may well find one that will read that troublesome CD (but fail on others that your current drive reads well.

Thx for the reply. Yea, that would be nice, if they all passed, and for the most part, they did. Only about 4% went into the "needs more work" category. But yes, was looking for general guidelines without having to do listening tests for each track. Most of these, were used CDs, but had passed error detection and CRC check with EZCD.

I do have two other drives, but haven't tried the with dBpoweramp yet. I will give that a shot.

[simbun]

CUETools can often repair rips with minor damage.

http://cue.tools/wiki/CUETools_Database#How_many_errors_can_a_rip_contain_and_still_be_repairable.3F

[SerbJ]

CUETools can often repair rips with minor damage.

http://cue.tools/wiki/CUETools_Database#How_many_errors_can_a_rip_contain_and_still_be_repairable.3F

Thx, that is interesting. That may certainly be something I'll try down the road. For now, I'll keep reading about it. It seems like there will be a big learning curve with that s/w package. I looked at EAC, but quickly came to dBpoweramp because I found the UI too difficult.

[SerbJ]

I have been running with logging set to "detail". This is what I see when set to "Complete". Could I get a brief explanation of what it means?

Track 5: Ripped LBA 99663 to 133380 (7:29) in 3:39. Filename: F:\RE_FLAC\Genesis - Abacab\05. Dodo-Lurker.flac
AccurateRip: Inaccurate (confidence 63) Insecure [Pass 1, Ultra 1 to 1, Re-Rip 11 Frames]
CRC32: A8DC883D AccurateRip CRC: 596AF5A5 (CRCv2) [DiscID: 009-00110029-007ab6e1-730af109-5]
Re-rip Frame: 130263 (00:06:48.000) matched 10 / 12 (c2 dropped 5)
Re-rip Frame: 130292 (00:06:48.386) matched 10 / 12 (c2 dropped 8)
Re-rip Frame: 130307 (00:06:48.586) matched 10 / 11 (c2 dropped 25)
Re-rip Frame: 130321 (00:06:48.773) matched 10 / 11 (c2 dropped 7)
Re-rip Frame: 130336 (00:06:48.973). Insecure (c2 dropped 60)
Re-rip Frame: 130351 (00:06:49.173). Insecure (c2 dropped 60)
Re-rip Frame: 130365 (00:06:49.360). Insecure (c2 dropped 60)
Re-rip Frame: 130380 (00:06:49.560) matched 10 / 12 (c2 dropped 21)
Re-rip Frame: 130394 (00:06:49.746) matched 10 / 11 (c2 dropped 6)
Re-rip Frame: 130409 (00:06:49.946) matched 10 / 11 (c2 dropped 2)
Re-rip Frame: 130423 (00:06:50.133) matched 10 / 11 (c2 dropped 1)

[Spoon]

>Re-rip Frame: 130263 (00:06:48.000) matched 10 / 12 (c2 dropped 5)

Here c2 indicated there were issues on 5 frames, but 10 out of 12 matched and were accepted.

>Re-rip Frame: 130336 (00:06:48.973). Insecure (c2 dropped 60)

The program likely tried 60 times on a single frame, c2 said all 60 had an error, and they did not match.

[SerbJ]

>Re-rip Frame: 130263 (00:06:48.000) matched 10 / 12 (c2 dropped 5)

Here c2 indicated there were issues on 5 frames, but 10 out of 12 matched and were accepted.

>Re-rip Frame: 130336 (00:06:48.973). Insecure (c2 dropped 60)

The program likely tried 60 times on a single frame, c2 said all 60 had an error, and they did not match.

Thx, and I did have that limit set to 60.

[simbun]

Thx, that is interesting. That may certainly be something I'll try down the road. For now, I'll keep reading about it. It seems like there will be a big learning curve with that s/w package. I looked at EAC, but quickly came to dBpoweramp because I found the UI too difficult.

If you just want to use it for verification/repair (which is what I was suggesting it for):

• Download latest: https://github.com/gchudov/cuetools.net/releases/
• Extract the contents
• Delete ./user_profiles_enabled
• Run ./CUETools.exe
• Enable detailed logging: Settings > Advanced > CTDB > Detailed log: True
• Click on the folder icon next to "Input" and select the directory containing the rip in the resulting window
• Change Action to Verify
• Click Go

Paste the log back here and I'll walk you through a repair if it's possible.

[garym]

Thx, that is interesting. That may certainly be something I'll try down the road. For now, I'll keep reading about it. It seems like there will be a big learning curve with that s/w package. I looked at EAC, but quickly came to dBpoweramp because I found the UI too difficult.

I didn't find CueTools to be much of a learning curve. I use it to repair some rips. Pretty amazing really. In Cuetools, select the directory containing your files to be repaired and the new directory where you want it to save the repaired files. Then in Cuetools, select ENCODE option and in the dropdown under that select REPAIR. Also select "tracks" rather than embedded.

[simbun]

I didn't find CueTools to be much of a learning curve.

I too have provided the basic steps but because I included a link to github it's awaiting approval.

Whilst it's not particularly user-friendly it doesn't take long to master.

[SerbJ]

I think I will give CueTools a shot, but will wait until after the holidays. I will need to focus.