MEDIUM ERROR:L-EC UNCORRECTABLE ERROR

Please post here for issues related to DVD discs
Woodstock
Posts: 10330
Joined: Sun Jul 24, 2011 11:21 pm

Re: MEDIUM ERROR:L-EC UNCORRECTABLE ERROR

Post by Woodstock »

skyy99 wrote:But it's real task is more akin to a batch job that processes each independent second of video data. In a batch job it is completely unacceptable to terminate the entire process if one independent piece fails. A batch job must complete and provide a log of what succeeded and what did not succeed. Each second of video is rather independent from the previous in terms of how it's being used (and even from an encoding perspective to some extent).
Dealing with process management, sometimes it is completely unacceptable for a batch job to NOT fail if part of the task cannot be completed. When the files are dependent upon each other being correct, one of them being bad means all of them are useless.

In this case, you have files (titles) that are made up of interleaved data streams. A bad block may affect any or all of them. Faking data for bad areas is the "batch job" equivalent is that it copies most parts of most files, and declaring the result a "success". Which files are bad? Some. But which ones? I don't know, you figure it out.

Or a different example. Failing hard drive. Operating system says it is bad, and won't talk to it. Is all the data lost?

Not necessarily. There are tools out there that will read the entire drive to another drive (or a drive image), substituting zeros for any bad sectors. But, is the result valid, bootable, and useable? Maybe, maybe not. Depends on where the zeros were written.

Can it be recovered? Depends on your available tools, and how much time you want to expend on it. Last time I did such a recovery, it took a WEEK to get 90% of the data off a 1TB drive.

If you want to have a program that accepts damage, MakeMKV isn't it. Maybe what IS it is a combination of a program (like "dd" on Linux) that can be told to replace "unknown data corruption" with "known data corruption" (zeros for bad sectors), then something like MakeMKV that will try to make sense of the damaged-but-now-all-readable image file.
PROs:
1) Fast fail is simple to code.
Having watched MakeMKV work, there is nothing fast about it failing. If you tell it to retry 25 times instead of the normal 5, it will repeatedly request the bad sectors, trying to get the data from the DVD or BD.

True "recover what you can" requires being able to get to the hardware - "Even if what you read fails consistency checks, give it to me anyway." Accessing the drive through the operating system (like MakeMKV does), you get back "FAILURE" and no data.

Players aren't tolerant when given corrupted files. I have yet to see one SKIP an error; they all just stop at the point of the error. You can test this yourself - take a copy of a working MKV or MP4 file, and randomly replace 2K of it with zeros. Try playing it with a player that is tolerant of errors in physical media. And it is rare that 1 bit is bad - it will be a multiple of 2048x8 bits.
skyy99
Posts: 3
Joined: Wed Feb 04, 2015 5:54 pm

Re: MEDIUM ERROR:L-EC UNCORRECTABLE ERROR

Post by skyy99 »

Believe it or not. I'm not trying to argue with you. But I don't think you've really understood the point I'm trying to make (poorly apparently).
Dealing with process management, sometimes it is completely unacceptable for a batch job to NOT fail if part of the task cannot be completed. When the files are dependent upon each other being correct, one of them being bad means all of them are useless.
Completely agree. But that's absolutely not what's going on here. The last bit/byte does not depend on the first. To think it does ignores many examples in reality: hardware players playing past scratches (where FEC does not save them), the ability to repair indexes in various video formats (see examples here for AVI http://www.virtualdub.org/blog/pivot/entry.php?id=25).

Now that independence is established, the fast fail argument does not make sense. Continuing and reporting which independent pieces did fail is more appropriate.
In this case, you have files (titles) that are made up of interleaved data streams. A bad block may affect any or all of them.
Indeed. It may or may not affect any or all of the streams. MakeMKV does not know. It should not assume it affects any of the streams. Again, because it does not know.
Or a different example. Failing hard drive. Operating system says it is bad, and won't talk to it. Is all the data lost?

Not necessarily. There are tools out there that will read the entire drive to another drive (or a drive image), substituting zeros for any bad sectors. But, is the result valid, bootable, and useable? Maybe, maybe not. Depends on where the zeros were written.
I agree. But that's a really poor analogy because (again) a HDD stores data of many different formats. For many different uses. One needs to take the HDD to a specialist to restore data because the HDD doesn't know if you can use 99% of a Word document or a Spreadsheet. But MakeMKV is different in that it has one purpose. It isn't a generic HDD. It has a very specific purpose and deals with a small set of file formats, which it knows well enough to decrypt. Pretending otherwise is unbecoming.
If you want to have a program that accepts damage, MakeMKV isn't it. Maybe what IS it is a combination of a program (like "dd" on Linux) that can be told to replace "unknown data corruption" with "known data corruption" (zeros for bad sectors), then something like MakeMKV that will try to make sense of the damaged-but-now-all-readable image file.
The world is a changing place. Recognizing that MakeMKV can change (or should change) is what makes things better. That's what this entire thread is about. Nobody is saying MakeMKV does something today that it doesn't, today.
Having watched MakeMKV work, there is nothing fast about it failing.
Fast fail has nothing to do with retries. It is a term that just means a process does not continue in the face of a failure (which say a batch job might do if the job tasks were independent).
True "recover what you can" requires being able to get to the hardware - "Even if what you read fails consistency checks, give it to me anyway." Accessing the drive through the operating system (like MakeMKV does), you get back "FAILURE" and no data.
Recover what you can means exactly that. Recover what the OS gives you. That which you cannot recover is... not recovered.
Players aren't tolerant when given corrupted files. I have yet to see one SKIP an error; they all just stop at the point of the error.
Not true. You and most others have experienced the contrary when playing an old scratched up CD in the car and had to seek past the bad part manually. Yes, that's just an audio stream and the codecs are different for video+audio. But the video container formats are not intolerant of many kinds of errors (again, for AVI, see http://www.virtualdub.org/blog/pivot/entry.php?id=25).

Lastly, I'm trying to get you to admit anything here. Just consider that there's room for change in MakeMKV and maybe you'll advocate for simple improvements like this which can improve an already great product.

Thanks,
Matt
kruegra1
Posts: 2
Joined: Sun Feb 19, 2017 7:28 pm

Re: MEDIUM ERROR:L-EC UNCORRECTABLE ERROR

Post by kruegra1 »

Did this get resolved fully? I have the same or similar problem ....

Error 'Scsi error - MEDIUM ERROR:UNRECOVERED READ ERROR' occurred while reading '/VIDEO_TS/VTS_04_1.VOB' at offset '0'

It happens right after title 16 is skipped on two different dvds, 2 different start wars movies. The idea its the dvd with a scratch is apparently incorrect in my case.
ikomrad
Posts: 19
Joined: Fri Feb 10, 2017 5:57 am

Re: MEDIUM ERROR:L-EC UNCORRECTABLE ERROR

Post by ikomrad »

+1 for option to configure to continue ripping after a read failure.
milder
Posts: 15
Joined: Mon Nov 18, 2013 6:27 am

Re: MEDIUM ERROR:L-EC UNCORRECTABLE ERROR

Post by milder »

A remarkable thing: I used a little stuff called "Gel Gloss" on a disc that was unreadable. Worked PERFECTLY! It's a wax cleaner.
dcoke22
Posts: 3093
Joined: Wed Jul 22, 2020 11:25 pm

Re: MEDIUM ERROR:L-EC UNCORRECTABLE ERROR

Post by dcoke22 »

milder wrote:
Sat Mar 20, 2021 7:32 am
A remarkable thing: I used a little stuff called "Gel Gloss" on a disc that was unreadable. Worked PERFECTLY! It's a wax cleaner.
Something from this brand? https://www.gel-gloss.com They make quite a range of products.

I'm curious, did you try cleaning the disc some other way first?
KENSAI
Posts: 62
Joined: Sun Feb 24, 2013 7:50 pm

Re: MEDIUM ERROR:L-EC UNCORRECTABLE ERROR

Post by KENSAI »

I'd sure appreciate the option to rip a disc that might not be a bit for bit perfect copy. I have a brand new boxed set of the Daniel Craig 4K Bond films, two discs ripped fine on my Libredrive, but one is giving me the read errors. No scratches on the disc, it's never been played before. The only option is to, what, just never rip it? It would be great to have SOME kind of option to get the disc copied.
milder
Posts: 15
Joined: Mon Nov 18, 2013 6:27 am

Re: MEDIUM ERROR:L-EC UNCORRECTABLE ERROR

Post by milder »

A bump/update for Gel-Gloss. That stuff is amazing. Having trouble with a 4K UHD of mine...Gel-Gloss to the rescue!
prahal
Posts: 8
Joined: Wed Mar 13, 2024 12:42 am

Re: MEDIUM ERROR:L-EC UNCORRECTABLE ERROR

Post by prahal »

Necromancy, but might be of help to surfers.

Had this on a brand new DVD

Code: Select all

 21:23:05 cyclope kernel: scsi_io_completion_action: 27 callbacks suppressed                                                                                                                                
juin 04 21:23:05 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=3s                                                                                            
juin 04 21:23:05 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 Sense Key : Medium Error [current]     
juin 04 21:23:05 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 Add. Sense: L-EC uncorrectable error                                                                                                                      
juin 04 21:23:05 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 CDB: Read(10) 28 00 00 00 00 00 00 00 08 00 00 00                                                                                                         
juin 04 21:23:05 cyclope kernel: blk_print_req_error: 27 callbacks suppressed                   
juin 04 21:23:05 cyclope kernel: critical medium error, dev sr0, sector 0 op 0x0:(READ) flags 0x80700 phys_seg 4 prio class 2                                                                                      
juin 04 21:23:11 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=5s                                                                                            
juin 04 21:23:11 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 Sense Key : Medium Error [current]              
juin 04 21:23:11 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 Add. Sense: L-EC uncorrectable error                                                                                                                      
juin 04 21:23:11 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 CDB: Read(10) 28 00 00 00 00 10 00 00 02 00 00 00                                                                                                         
juin 04 21:23:11 cyclope kernel: critical medium error, dev sr0, sector 64 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2                                                                                     
juin 04 21:23:15 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=3s                                                                                            
juin 04 21:23:15 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 Sense Key : Medium Error [current]                                                                                                                        
juin 04 21:23:15 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 Add. Sense: L-EC uncorrectable error                                                                                                                      
juin 04 21:23:15 cyclope kernel: sr 5:0:0:0: [sr0] tag#0 CDB: Read(10) 28 00 00 00 00 10 00 00 02 00 00 00                                                                                                         
juin 04 21:23:15 cyclope kernel: critical medium error, dev sr0, sector 64 op 0x0:(READ) flags 0x0 phys_seg 8 prio class 2                                                                                         
juin 04 21:23:15 cyclope kernel: buffer_io_error: 22 callbacks suppressed                                                                                                                                          
juin 04 21:23:15 cyclope kernel: Buffer I/O error on dev sr0, logical block 64, async page read                                                                                                                    
juin 04 21:23:15 cyclope kernel: Buffer I/O error on dev sr0, logical block 65, async page read                                                                                                                    
juin 04 21:23:15 cyclope kernel: Buffer I/O error on dev sr0, logical block 66, async page read                                                                                                                    
juin 04 21:23:15 cyclope kernel: Buffer I/O error on dev sr0, logical block 67, async page read                                                                                                                    
juin 04 21:23:15 cyclope kernel: Buffer I/O error on dev sr0, logical block 68, async page read          
juin 04 21:23:15 cyclope kernel: Buffer I/O error on dev sr0, logical block 69, async page read                                                                                                                    
juin 04 21:23:15 cyclope kernel: Buffer I/O error on dev sr0, logical block 70, async page read                                                                                                                    
juin 04 21:23:15 cyclope kernel: Buffer I/O error on dev sr0, logical block 71, async page read      
and:

Code: Select all

isoinfo -d -i /dev/sr0 
CD-ROM is in ISO 9660 format
System id: 
Volume id: FRIENDS_S10_DISC4
Volume set id: FRIENDS_S10_DISC4
Publisher id: WARNER_HOME_VIDEO
Data preparer id: GLOBAL DIGITAL MEDIA XCHANGE INC.
Application id: 
Copyright File id: 
Abstract File id: 
Bibliographic File id: 
Volume set size is: 1
Volume set sequence number is: 1
Logical block size is: 2048
Volume size is: 1866525
NO Joliet present
isoinfo: Input/output error. Read error on old image
note the "isoinfo: Input/output error. Read error on old image" at the end.

I tried multiple times (though always with the same recent DVD/BD usb writer") no better. The DVD does not mount in linux (only the previous error and incomplete isoinfo.

Cleaned with isopropanol, without much hope as except a few digit marks I made on the disc while getting it to the dvd reader, but there was not a single scratch, brand new. And no more errors, disc mounts....

Code: Select all

isoinfo -d -i /dev/sr0 
CD-ROM is in ISO 9660 format
System id: 
Volume id: FRIENDS_S10_DISC4
Volume set id: FRIENDS_S10_DISC4
Publisher id: WARNER_HOME_VIDEO
Data preparer id: GLOBAL DIGITAL MEDIA XCHANGE INC.
Application id: 
Copyright File id: 
Abstract File id: 
Bibliographic File id: 
Volume set size is: 1
Volume set sequence number is: 1
Logical block size is: 2048
Volume size is: 1866525
NO Joliet present
**BAD RRVERSION (0)
NO Rock Ridge present
Go figure. Could it be there is a thin layer of something on the disc I cleaned with my isopropanol?
Even something that condensated on the disc in storage? (Could have been in my storage as I got the drive out of the plastic but only attempt to read it monthes afterwards?
Post Reply