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: Need Help With Scratched CD - I Have Multiple Copies of the Same CD (Read 11206 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Need Help With Scratched CD - I Have Multiple Copies of the Same CD

Reply #26 –
Quote from: knutinh on
Intuitively, I would guess that when a single sample is missing, some fc=fs/4 windowed ideal brickwall filter would be a good starting point. But should you just fill inn the missing sample based on those coefficients, or should you also alter neighbor (known good) samples to avoid discontinuities?
There are lots of known techniques for filling-in gaps in audio.

Though this is mostly about "old records", there are parts that are relevant to the problem you pose...

http://www-sigproc.eng.cam.ac.uk/~sjg/springer/index.html


Quote
2. Given one ripping with errors (known or unknown sample indexes) and accurate-rip, is it possible to find the original waveform brute-force?

Well, for a multi-second/minute track with errors of unknown location and duration, I think you can forget the brute-force approach! However, if you know where the error(s) are, this is possible for a single sample. How it scales with more than one sample wrong depends on whether there are some short cuts you can perform on the CRC algorithm in this case. Otherwise it scales with the number of combinations, which is 2^16^n where n is the number of samples to be guessed - pretty impossible by the time you get to one CD frame! Maybe you could try a best guess (using audio restoration techniques) and work "outwards" from there?

I still think I'd prefer Brasso, or buying another copy!

Or, let's be honest, if I've paid for it and it's got scratched before ripping, I have no worries about downloading a replacement.

Cheers,
David.

Need Help With Scratched CD - I Have Multiple Copies of the Same CD

Reply #27 –
Quote from: 2Bdecided on
I still think I'd prefer Brasso, or buying another copy!

Or, let's be honest, if I've paid for it and it's got scratched before ripping, I have no worries about downloading a replacement.

Cheers,
David.

Sure.

There is one topic that have not been covered yet:
Having 3 (or more) channels of rips where insertion/dropping may have occured independently.

If you have max 1 such error per track, then correlation methods is quite straight forward. If you have a number of such errors, then it gets quite compex, I think, to align the N tracks properly, so you can do the median-filtering.

-k

Need Help With Scratched CD - I Have Multiple Copies of the Same CD

Reply #28 –
Quote from: knutinh on
If you have a number of such errors, then it gets quite compex, I think, to align the N tracks properly, so you can do the median-filtering.

This is something that is done quite frequently in, e.g., comparing two versions of source code for software. Highlighting shows which lines have changed, which have been added, and which have been deleted. Applying that algorithm to audio files should be a snap.

Need Help With Scratched CD - I Have Multiple Copies of the Same CD

Reply #29 –
Quote from: pdq on
Quote from: knutinh on
If you have a number of such errors, then it gets quite compex, I think, to align the N tracks properly, so you can do the median-filtering.

This is something that is done quite frequently in, e.g., comparing two versions of source code for software. Highlighting shows which lines have changed, which have been added, and which have been deleted. Applying that algorithm to audio files should be a snap.

Good point. But you often have to do manual confirmation for complex diffs in such systems.

wikipedia: Merge (revision control)
The new Codeville merge algorithm

-k

Need Help With Scratched CD - I Have Multiple Copies of the Same CD

Reply #30 –
I don't think this counts as complex though - it's bit identical with occasional errors and very occasional loss of sync. Where it matches, it matches perfectly - and it's never re-ordered.

Cheers,
David.