So, after a few days of running 3.95 through its paces, what's everyone's conclusion on it? I fiddled around with it myself a bit, and it encoded the beastie boy's funky boss to a larger file size than 3.90.3, which was disappointing.

patience, my son

it took months of testing by many people (probably years of actual man hours) just to conclude that 3.90.3 was a good build. i doubt they'll know any time soon, even if everyone involved in 3.90.3 tested it.

just go by what sounds good to you. or rely on a preset.

Jury is still out.
You cannot possibly want to have even something remotely approaching a conclusion with just one week's worth of non-coordinated testing.
We'll just have to wait some more

whatever the verdict...

3.95 was rushed too quickly (and sloppily? sloppy?) which doesn't bode well for testing and the whole "recommended" thing


later

dude, it's been over a year since the last stable release!

dude. the timtable of a release doesn't necessaryly indicate it's quality. 'stable' 3.95 aswell as 3.95.1 had bugs, which are whiped out atm in 3.96a (hopefully).

@ xmixahlx: I woudn't say sloppily though...

...i would say sloppily

both lame 3.95 and 3.95.1 still have bugs (decoding)
3.95.2 isn't released yet to fix them

so they rushed to get 3.95 into the 128 mp3 test, and still haven't released a proper build.

sloppy


later

who in hell uses LAME for decoding, anyhow?

xmixahlx, as it seems.

otherwise I couldn't understand his frustration.
3.95.1 works great for me. it was under development for more than a year and I think that a lot af testing went into it developer side wise.

where was xmixahlx when 3.94 beta was released for testing? where are his bug reports. It's easy to blame the others... ts, ts

no...

it's easy to say that 2 rushed "stable" releases in 2 days that still have a large bug (decoding) is sloppy

bug reports are on this board...

no frustration, just an acknowledgement


if you can't admit that this is sloppy, it is called "denial" not a gung-ho spontaneous HA poster


later

Many things are released with bugs... Take for exemple some Nero AAC encoders, some foobar2000 releases... at least one mppenc.exe version (problem at --quality 10), Fraunhofer fastenc, iTunes, etc... Even the first LAME 3.90 was buggy (with --fast presets), and was corrected with 3.90.1/3.91... And I'm not talking about commercial products, always buggy! Roberto's test are generally buggy (offset problems, vorbis 4,25/4.25 - 48khz issues...), HA board is sometimes buggy...

All softwares (except maybe notepad.exe) are therefore sloppy... There's no need to blame lame developers. They work hard for free, and need encouragement...

Hi

Someone WILL use LAME to decode, but it's not you or me.

But in fact, even famous WinAmp DO have decoding bug until 2.9x version, so I stick to MAD decoder.

And now Foobar2000, because WinAmp 3 is too disappointing and Foobar2000's excellent expandability.

I've personally ditched 3.90.3 --alt-preset standard in favour of 3.95.1 -preset standard. Why? well, a few years of development must mean something (usually) .. and i'm not a person that lives by the "if it ain't broken, don't fix it" slogan, i like new things, i like to test new things, i like progress.

I have nothing negative to say about 3.95.1 atleast. Bitrates are lower in the -ps preset than in -aps, and i cant notice any quality differences, thus - better (smaller filesizes, equal quality to my ears).

Wouldn't It be Wise, to release Lame 3.96 that is 100% fixed? (I mean the bugs that were found from 3.95, decoding etc)

I think then it would stop all confusion....

Cheerz

xmixahlx, stop being so negative. anything of the complexity of LAME is going to have "large bugs". the only way to avoid them is to walk away from the computer.. and even then you might come across large bugs . btw don't take this seriously, you probably know way more about this than me. I just don't think it means they're sloppy, more that it's not their paid job, and it's a bit harsh having a go at them. like someone said, missing a bug in decoding probably means they weren't looking at decoding.. call it sloppy if you like, but for most people LAME is still the best MP3 encoder and we'd like some more dev please

3.95 seems good but if you want to be more certain you'll have to wait longer. I go with bidz on this one

For us (Lame dev) a decoding bug is not considered to be a major bug, considering that the goal of Lame is ENcoding.
We do not have the needed time to had a wonderfull rock-solid decoder, and choosed to use our time on the encoding side.

If anyone wants to take the task of correcting the many decoder issues, he would be welcome.


btw, what does "sloppy" means?

slop·py
adj. slop·pi·er, slop·pi·est

1. Marked by a lack of neatness or order; untidy: a sloppy room.
2. Marked by a lack of care or precision; slipshod: sloppy use of language.
3. Informal. Oversentimental; gushy.
4. Of, resembling, or covered with slop; muddy or slushy: sloppy ground.
5. Watery and unappetizing: a sloppy stew.
6. Spotted or splashed with liquid.

sloppi·ly adv.
sloppi·ness n.

Edit: in the context of Smiff's post, it should be somewhere in between 1 and 2.

You shouldn't

I don't know the smallest thing about LAME's code, so if what I'm requesting is absurd, please forgive me.

Do you thing exact-length decoding could be implemented in some new version of LAME? You know, to produce gapless decoding, a la foobar2000.

Could be, but someone has to do it.

when i use "sloppy" it isn't duragatory

it is criticism not blame

it may be negative, tho... and that would be my opinion

point releases are common, but there still isn't a proper build yet (3.95.2)


later

@xmixahlx

Why don't you do something about it, rather than just complain and whine about it?. Anyone can modify the LAME sources, and fix things, supply patches, or whatever.. so it's really up to you if you aren't satisfied or happy with the current releases.

good call bidz,

hey lets make a big deal about it. while we are, you know, slightly annoyed and/or concerned about how they handled lame between january 8 and january 12. while new version releases were daily (cvs/stable)

what i should do is try to make my own lame!
yes!

and this way, i can only offer criticism to myself! wonderful!


get real

btw, i love the colorful words so far: blame, complain, whine, negative, "not happy"

well you certainly are doing this now.
I think bidz got a point there.
the lame project is given to us freely without charge, by developers who spend a part of their personal life to make it better, as best as they can. we all know that there has been internal 'problems' concerning the lame team, so hey, don't execpt any wonders concerning the rate of releases or maybe even their qualities. I'm not saying we should worship all lame developers and sacrifice a white goese at midnight or anything, but a little bit of respect may not be too wrong. after all, everybody is free to say what he wants, and to use the software he wants. if somebody is not happy with lame or it's related topics, either don't use it, or try to make it better in a contructive way. if you don't got the programming skills and the knowledge to contribute to lame (which is very understandable), one might find other ways of helping, eg listeningtest etc (if required and wanted by the developers, and maybe even without).

let's get back to a discussion here, before everbody starts to go nuts about something that isn't worth it.

that's generally not a bad thing at all.

yes. everybody.

just my 0,02 €

edit: though I think everything has been said to make that matter clear, someone correct me if I'm wrong.

lame 3.96a1 is a bugfixed version of 3.95.1, without problems at decoding