well i was listening to some of my recently encoded mp3s and i was really surprised when i came upon what appears to be a pretty egregious flaw. Never having heard ANY kind of obvoius problem in ANY lame version, i reripped it twice, and checked with a regular cd player the original sample.

I included the link to the sample. i'm sorry i'm not a regular in this forum, so i don't know if this is some kind of common problem or what. thanks guys, let me know what you think.

Here's The Sample

INFO
ripped: eac .95 pb 3
encoded: lame 3.90.3 --alt-preset standard
from the forums

sample from: track 3, "don't wait" from devin the dude - devin the dude

note: interestingly enough.. it happened pretty much exactly the same wth all of these presets
CBR 320
VBR AltPreset extreme
VBR AltPreset Medium

Have you tried 3.93.1 ?
Ok normally it should sound the same, but I still hope to hear of one example where 3.93.1 beats 3.90.3

Sorry but I could not ABX the original wav from the --aps mp3

I have been able to reproduce a problem when playing back on my Kenwood car deck. I get a burst of static at the end of some tracks. It seems to be specific to 3.90.3

I had the same problem some time back with Mitioks compiles.

I have never had the problem with the same music using 3.90.2 So, I had stuck with this rev of course.

I do not get this burst of noise from the PC using those same 3.90.3 files.

Perhaps it is a 'feature' on the Kenwood KDC-MP8017.

well, i don't know what abx means but the original sample wav is right next to the mp3 in the rar.

did i post this to the wrong place? i was wondering if anyone who works on lame would be able to use this to help improve it- because it is strange since it is just a simple bass riff.

btw just tested it with 3.90.2- same problem - didn't try 3.93 because isn't 3.90 what everyone is supposed to be using anyway?

I hear nothing wrong on my PC.

ff123

what player are you using?

what the hell is going on here? i notice now that i don't hear it in WMP 8, but i hear it with the crossfading, directsound, and waveout outputs for winamp 2.91, AND my brand new 3rd generation IPod. What gives? Anyone? Now I feel like i'm just crazy... and since when does WMP do anything better than Winamp??
And i can understand it being some kind of decoder problem, but now why would my ipod have the EXACT same problem? thanks guys

ps. by the way the bass right after the string is pulled makes a kind of "thunder" sound. a kind of garbled, almost digital amplification and distortion of the bass. and no, my EQ is not on.

EDIT: I reinstalled winamp, and now it is not a problem anymore (?!?!). but i still hear it on my IPOD. still, any guesses at to what the issue with winamp was?

I hear nothing wrong either.

Player: foobar2000

I tried it first using foobar2000, and then using Cool Edit Pro 2.1. Nothing wrong that I can hear in either case.

I also don't see anything funny going on either in time domain or in spectral view (i.e., no clipping).

EncSpot shows 256 kbit/s frames, but nothing above that. Nothing out of the ordinary.

ff123

Winamp applies eq and dsp to mp3's but not to wav's, or at least that is what i've heard. Most likely, you had either eq or dsp enabled in Winamp without being aware of it.

ff123

I hear a noise in both samples. It sounds like what could be a finger slipping off a string (a ticking or chirping sound), and it occurs in two places. The first occurs at 1.760 seconds and the second at 7.422 seconds (the accuracy is due to CEP).

Later.

Since MS threw probably hundreds of thousands of dollars at it to try and replace winamp in the hearts of users.

I wouldn't doubt that WiMP has a very capable MP3 error correction routine.

Please, people: Sure, Microsoft does make software that is buggy sometimes, but that certainly doesn't mean their software is in any way sub-par. Hell, MacOS crashes too. As do WordPerfect, Winamp and Mozilla.

That's very interesting... The Ipod shouldn't have a broken installation Perhaps try to encode that sample to AAC and listen to it on the IPod - then you know if it's after decoding or while/before.

Regards, fileman.

So the reliability of software doesn't have a bearing on the quality of the software? Er... yeah. Try telling that to anyone who makes a living off of any particular piece or pieces of software...

There's no reason to assume that Microsoft's support of MP3 is high quality/reliable/whatever without having verified so. In fact, if anything, it's more likely that it will be broken more than other implementations given how it already doesn't support many things properly as it stands now (mp3 vbr bitrate display being an example).

Oh, and FWIW, other software crashing doesn't make Microsoft products any better. It simply means that the other software is unreliable as well.

I believe this is the issue, concerning an unoptimised decoder.

Use the --strictly-enforce-ISO switch and see if this fixes it.

let my quot that site

UPDATE 10/2/00: Winamp Bug Fixed!

probally not that :-)

no luck. i'm going to try and force this issue upon Apple, because it no doubt is a problem with the mp3 decoding in the ipod/

WiMP's MP3 decoding is, and always has been, broken.

I've tried to play MP3's from Lame 3.89, 3.90 and 3.93 on WiMP 6.4 (aka WiMP 2), 7.0, 7.1 and 9.0. There are always VERY obvious decoding errors, taking the form of blorps and chirps.

This is on both a PII-300 system with Win98SE and a SB16 soundcard, and a PIII-733 system with Win98SE and a SB128 soundcard.