here they are...
And what did you encode it with ? This file is exactly 3 frames too long, it's either a frontend bug or a bug in LAME itself. Please post source WAV (preferably lossless-compressed) used to encode this.
whoa, this post didn't look like it went up last night, following is the rest of the stuff
Well, I was wondering whether I was missing something but part1 was enough to tell what's most likely wrong (exactly 3 MP3 frames of silence at the end).
ok well here are the wavs, zipped. one important thing: i ripped and ecoded the tracks with EAC, and *trimmed them after encoding* with some mp3 tool, and then when you asked for wavs i went back and trimmed the originals. that being said, all the problems are at the inner ends of the files..
at any rate, the WAVs play gaplessly for me (as do the Jarre mp3 samples you posted in another thread) so i guess the problem is with the encoding??
here is my EAC log:
EAC extraction logfile from 16. May 2004, 18:04 for CD
The Beatles / Abbey Road
Used drive : HL-DT-STCD-RW GCE-8523B Adapter: 0 ID: 1
Read mode : Secure with NO C2, accurate stream, disable cache
Read offset correction : 6
Overread into Lead-In and Lead-Out : Yes
Used output format : R:\Path\lame.exe (User Defined Encoder)
128 kBit/s
Additional command line options : --preset standard %s %d
Use compression offset : 0
Other options :
Fill up missing offset samples with silence : Yes
Delete leading and trailing silent blocks : No
Native Win32 interface for Win NT & 2000
Track 11
Filename T:\r\Abbey Road\11.wav
Peak level 89.4 %
Track quality 100.0 %
Copy CRC B1A33E4F
Copy OK
Track 12
Filename T:\r\Abbey Road\12.wav
Peak level 79.9 %
Track quality 100.0 %
Copy CRC AC46F66B
Copy OK
No errors occured
End of status report
also, what software did you use to find out the file was too long??
Most (if not all) MP3 trimming tools break gapless playback info as far as I'm aware of. Split WAVs, then encode.
[edit] I decoded the file with foobar2000 diskwriter and opened resulting WAV in adobe audition; it had 3*1152 silent samples at the end.
ok this is really strange: I just re-ripped the wavs with EAC (they are separate tracks which lead together, so i'm not 'splitting' any wavs, here), trimmed them in Sound Forge, and then encoded them from a DOS prompt using preset standard and.. PRESTO! gapless playback using mpg123 and 500ms buffer-ahead....
so the problem is with EAC? i get the same 'gap' in playback when i tried another cd with a track transition...
attatched is a screen shot of my external compression settings...
ok, i have isolated the problem, and it appears to be a bug in either lame or EAC--Instead of having the encoder offset correction turned OFF, i had it TURNED ON and SET TO ZERO, (see screenshot) which i assumed was the same thing, but it turns out it was not--having this setting appears to be what was adding the extra 3 samples, since turning it off (and not touching other options) gives me totally gapless playback under mpg123+500ms-buffer-ahead.
i haven't seen this doccumented anywhere on the board, so this might be the reason that some other people i read posts by couldn't get gapless working--but anyway, what exactly is happening here? is this a bug in lame? in eac?
(using lame.exe 3.96, EAC v0.95pb5)
1) Do you do Gap Detection first? <F4>
2) Using Native Win32 calls is not suggested. It can introduce clix.
3) Fill up missing offset samples with silence : Yes - Did you try shutting this off?
i didn't do gap detection, but these 2 tracks have no gap (both are index 01)
i'm using the native windows SCSI interface if that's what you mean, and i never heard it could cause clicks, but the problem here is GAPS--so what does it matter with the windows interface? and the problem is solved with the offset correction turned off
3) no i didn't try shutting that off.. i might, just to experiment, but it seems like it woudl be an accessory to the crime, not the perpetrator...
Why did you enable compression offset at all?
It's not needed when you use lame.exe.
The recommended LAME settings thread (http://www.hydrogenaudio.org/forums/index.php?showtopic=203&) has a guide too.
as i said, i thought setting it to zero was the same as disabling it.. i left it in that state because i ran the 'test' 'just to see' that it really did give zero.. and then i just left it...