Is there any program for checking mp3 integrity, errors, VBR header errors and so, which will detect and repair unicode named corrupted files?

So far I’ve tried mp3val, MP3Utility and MP3Test without success

The problem occurs when the scanning program encounters a file name containing accented characters like á or é. Then the scan stops and you get a message like “Cannot open input file”.

To recreate the problem simply name a file like:
Jussara Silveira - Lá Vem a Baina.mp3 or Koité, Habib & Bamada - Wassiye.mp3, then try to test it.

Not a direct answer, but here’s how I deal with such issues.

Once in a while I load all my files in Mp3tag:

http://mp3tag.de/en/

... and using an Action:

http://mp3tag.de/en/help/options_format.html

I shorten all filenames to 64 characters, and
I replace such characters in filenames (only in filenames) with Basic Latin:



This is an easily reversible action, since the proper filenames can be restored from the tags by means of Mp3tag, foobar2000 or any other good tagger.

If you want to try this, I can upload the configuration file for my Unaccent+Shorten Action in Mp3tag (or maybe you can ask in the Mp3tag forums for a better idea).

Those are not even Unicode characters! They're in the 8-bit domain of win-1252.

The tools are seriously broken/out of date.

Costas might be running Windows with a locale that doesn't have those characters in its codepage.

In that case, you need to have dedicated Unicode support to handle that, AFAIK.

BTW, renaming such files is very simple with foobar2000.

Just using $ansi(%filename%) as your target filename should probably do it.

Demetris thanks for the answer. In my case, I have mp3s in many languages (Greek, Japanese, Hindi, Arabic, Hebrew to name a few) so it is very difficult to make a global Unicode to basic latin map.

My workaround so far is to copy my entire collection using the xcopy command with the /n parameter. /n creates copies by using the NTFS short file or directory names (aka 8.3 file names). Then I can run the tests and rename my files back. However, if you have a 500GB collection like me you have other practical troubles (disk space, etc).

I have different problem, and that is with cyrilic names burning onto DVD. Is there some way to do that? I can see these filenames right, I can play them in Foobar, I can copy them over folders, but the second I put them in Nero, I get "file not found" or some similar error.
Any suggestion? And is there some way of translating cyrilic into latin letters? So far I haven't found anything... I don't need phrase translation, jut latinisation.

There is a small app that would do it for mp3 files: mp3translit.
The problem is the app's UI is in Russian (I think it actually needs Windows to be re-booted using Russian codepage to display the menus and buttons).
On the other hand, it seems to support editable symbol conversion tables (Unicode to latin), so one perhaps could edit such a file to include characters from other cyrillic languages. (maybe it works even for non-cyrillic?)

I have to add also that I did use it to rename mp3s and it was alright, but I cannot remember if it can rename any other type of files... Can't remember really...

@kjoonlee: Thanks for the clarification, makes total sense now.

Well, my suggestion is to rename the files to <track#>.<extension> then. 01.mp3, 02.mp3, 03.mp3, and so on... Because when you don't write out the track title in the first place you won't have to worry about converting characters.

Afrer you're done using the non-unicode-aware tools, rename the file back.

But I guess the tools might still get confused by the diretory name right? Hm, what about using the 8.3 names?

hlloyge: I had this problem as well. Only solution so far (thanks to other guys in this forum) is CopyToDVD from VSO.

Have you tried using UDF, instead of ISO 9660? Or if you have to use ISO 9660, have you tried Joliet?