Support and General Use > Audio Playback, Database and Playlists
Rockbox don't recognize an mp3 file
dreamlayers:
The problem goes away if I remove the first first MP3 frame, which contains the Xing VBR and LAME tag. Note that if you remove that frame, you also remove the VBR seek table, and so you may have problems when seeking.
(At the beginning, the file has a rather atypical ID3v2 tag which has an image but none of the usual text info. That is not a problem. If you want, you can remove the tag with Mp3tag.)
Edit:
The file is not VBR, so feel free to remove the frame.
You can simply run the Rockbox vbrfix plugin on the file to make it playable. Highlight the file, hold select, go to "Open With..." and choose "vbrfix".
cjm2576:
Spoke to soon. That did fix it for several of the files but I have a few that still aren't working. This one...
http://podcasts.howstuffworks.com/hsw/podcasts/sysk/2009-08-25-sysk-steal-nuclear-bomb.mp3
VBR fix gets to 49% and reports that it is not a VBR file but Rockbox is still skipping it.
I think at this point I'm also going to try emailing How Stuff Works to see if I can't get any traction on their end for fixing their MP3s.
Thanks again.
Edit:
I tried removing all tags (id3v1, id3v2) using a unix tool (id3v2) and this didn't help.
dreamlayers:
Yes, the MP3 files are broken and How Stuff Works should release fixed versions. The problem is in the Xing/Info header. A flag is set which claims that the header contains an entry with the number of MP3 frames in the file, but that number is zero. So, Rockbox thinks the file has duration zero and refuses to play the file.
I can't really call this a Rockbox bug because the MP3 file is broken. However, being able to play broken files is good, and it might be a good idea to add support for this particular brokenness.
Don't bother with ID3 tag utilities. This is not in an ID3 tag. It is in the first frame of the file. Try getting a copy of vbrfix which can run on your computer, and see if that will fix the files.
cjm2576:
I'll forward this off to them and hope they do something to fix it. In the mean time I can work around it.
Thanks again for all your help.
Lear:
Rockbox makes sure the byte count from the Xing/Info header is reasonable and ignores the byte count if not. Seems like something similar for the frame count would be in order... Just checking for zero would be a start; making it better than that would be nice, but it might be difficult to do in a good way.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version