Help - Search - Members - Calendar
Full Version: MP3 durations are wrong...
Hydrogenaudio Forums > Lossy Audio Compression > MP3 > MP3 - General
Guillaume
I'm having problems with several MP3 files: their durations aren't right! For example, a song is 3:34, but the duration Winamp (or Apollo) recognizes is 3:33. The last second will not be played at all...

This is what I do when making the MP3's:
- EAC: Copy Image & Create CUE Sheet (Uncompressed)
- burn a backup
- Split the image by cue sheet (leave gaps)
- Lame --alt-preset standard

Another thing I hate is when I use Lame's --nogap/--nogapout command, the duration first mp3 file of the album is recognized as 30:41 for example... The album's duration is about 1:05:11. How can this be??

[edit]
Some strange things in EncSpot 2:
- all files' music CRC's are corrupt! (because of the ID3-tagging afterwards??)
- Of some of the files, encoded in Lame 3.92, I cannot see the Lame Header...
[/edit]
Guillaume
I think the music CRC corruption is because of mp3Gaining, isn't is?
Guillaume
sad.gif
Case
QUOTE
Originally posted by Guillaume
I think the music CRC corruption is because of mp3Gaining, isn't is?

This is correct. No need to worry about it.
Guillaume
Another thing: is it correct that mp3's encoded using the --nogap switch do not have a Lame header??

What about that strange duration error? I can understand this with --nogap encoded mp3's, since they lack a VBR header, but a normal --alt-preset standard encoded file?? How come? Can I fix this by a VBR-fix tool?
Moguta
If the MP3 is missing a VBR header, then it will certainly show an incorrect time because the VBR header is what is used to calculate the correct time in VBR MP3s. =P
Guillaume
That I know, duh. But the file does have a VBR header...
Guillaume
A big kick:eek:
Guillaume
I've got another question. Is there any way to find out which gain has been used in MP3 files?
Case
QUOTE
Originally posted by Guillaume
Is there any way to find out which gain has been used in MP3 files?

No. You need to store the logs or save the gain values some other way.
Guillaume
What about the durations? Can I (try to) fix this using a VBR Header fixer? Which tool is best to use (using WinXP)?
CiTay
QUOTE
Originally posted by Guillaume
Which tool is best to use (using WinXP)?


Try the one from here: http://www.willwap.co.uk/
lbschenkel
I'm facing this problem too... In *every* MP3 I create with --aps, LAME 3.92 creates a wrong Xing header:

* the track size is incorrectly displayed in Windows XP and also in Tag&Rename
* the mp3cue plugin for Winamp seeks way before/after the correct position (sometimes even by a margin of 20sec!)
* when I press the left arrow key in Winamp, the song does not always rewind 5 seconds but a lower amount

After I correct the header using vbrfix (www.willwap.co.uk):

* the track size is now displayed correctly in both XP and Tag&Rename
* the mp3cue plugin is *much* more accurate
* the left arrow key in Winamp rewinds 5 seconds

But one side effect of using vbrfix is that EncSpot detects the encoder as GoGo 3.1 (or something similar), and the LAME tag is removed from the file.

I have this problem since LAME 3.88, and it happens in Windows 98/ME/2000/XP. I have not tested in Linux, though. I'm wondering how this (problable) bug can be around so much time without some LAME developer noticing it.
Guillaume
Thanks for the reply. And yes, I'm also facing the problems you described above. Using a VBR header fixer is not something I look forward to, especially because of the loss of the Lame header. But having an MP3 file of about 2 hours (U2 - Live in Boston DVD Rip) and Winamp displaying a track length of only 50 minutes (as well as WMP), is a lot more serious...

I think the Lame developers should really have a look into this and not only fix it in the next stable release, but also make a patch for already encoded files showing this problem. I hope it's possible to create a patch capable of fixing the VBR header and keeping the Lame header intact.
bronco30
I also noticed this problem - I just started ripping my cd's in using EAC and LAME - and am new to the whole lingo here - but from what I understand that VBR encoded MP3's cause the time display to be wrong - but if you fix it with the before mentioned program - it deletes a LAME tag (what does this cause? and does it mess up the file?) I would really like to have the times correct - especially when trying to calculate file time in order to burn a cd. I just started with my CD's but would appreciate some clarification on this issue prior to continuing. Thanks for your help.
Guillaume
A big kick for a good cause... biggrin.gif
NeoRenegade
I have the same problem as originally posted here except with CBR files...

I rip to WAV with EAC — trim the WAV's with WAVTrim (compact and truncate turned on, 1ms left for gaps, search first and last 10s for silence, analog silence under -90dB also removed) — pass the WAV's through RazorLame — then process the MP3's with MP3Gain and tag them with MP3 Tag Studio and by hand in WinAMP.

Somewhere along the line, I'm getting durations for some files that are up to 3 seconds longer than they should be.

The files play just fine on anything, though.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2008 Invision Power Services, Inc.