Help - Search - Members - Calendar
Full Version: Change of lenght in LAME mp3 decoding
Hydrogenaudio Forums > Lossy Audio Compression > MP3 > MP3 - General
RideTheSky
I have an original CD which I want to backup and also to add 2 more songs which are bonus on special edition CD, and to make that backup compilation accurate with special edition. I have those additional songs in LAME 3.97 320kbps without any tags.
The thing is next, when I decode with lame 3.97 with default setting --decode input.mp3 output.wav the lenght of the wav is somewhat different (for 0.150 seconds shorter) then that of the mp3 and that differences result in inaccuracy of backup compilation. It always happens when I decode with lame or dbPoweramp or ALF-LameGUI when i use switches like -t the wav is more shorter, but when I use Easy CD-DA extractor or when I save that wav from SoundForge, or when I put directly into Nero and burn new Audio image the problem is gone and the mp3 and wav in those cases have identical lenght.
Also when i convert LAME wav back to mp3 new mp3 file is again shorter than the original one, and when I use wav from mentioned other programs there is no problem.
So, the question is why that happens with LAME, and is decoding with CD-DA extractor or Nero has identical quality as decoding with lame; Also is there any difference regarding joint-stereo to normal CD stereo convert in these decodings?
Thanks
Dynamic
When you encode with LAME it adds accurate length information in the LAME header to support gapless playback which wasn't possible with standard MP3s. All MP3s are actually longer than the original WAV, and decoders that support LAME accurate length correct for this when the information is present, so it's LAME, foobar2000 and dBpoweramp that are truer to the original CD by having identical track length, but no different in audio quality.

Decoders will all be MP3 compliant in all other respects, so stereo reproduction etc. will be the same for all of them.
RideTheSky
@Dynamic Thank you, I had similar thoughts.
I've tried now to encode back to LAME using the -t switch which prevents that info in LAME header and I've got same to original mp3. Does it then means that using -t flag when decoding, which produces even shorter track lenght file, is more accurate even though that wav is raw PCM and then must be processed before it can be properly used or LAME header info in this wav is absolutely OK and I guess there is no need for using -t flag when decoding.
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.