Hi All,

I have just inspected my last 2 album purchases from allofmp3.com. I used Mr Questionman. It indicates that they are using fast encoding! Is this very bad. I select preset standard when ordering.

In current versions of LAME so-called fast encoding is considered to be at least as good as the older code. Rather than fast it should probably be referred to as the new algorithm, which also happens to be faster.

If those mp3s are in VBR (alt-preset standard fast) and encoded in LAME 3.97 or newer then theirs nothing to worry about, since the new vbr-new (fast) setting is better then the old one.

By the way, I heard allofmp3.com was forced to shut down, so it's changing it's business. Ah well, we'll have to wait for the chinese version of allofmp3.com.

http://news.google.com/news?hl=en&ned=...tnG=Search+News

to quote one site:

Despite the statements by the U.S. Trade Representative Office, AllofMP3.com has distributed a press statement, saying that the site will not be shut down by Russian authorities. A spokesman for the site was quoted by Arstechnica.com website as saying: “My understanding is that the Russian government has agreed to certain changes in the copyright area, and those changes to its current law are to be determined. AllofMP3 will alter its business model to fit within the new parameters.”

IIRC, the quality of the files available from allofmp3 has been questioned here before.

http://www.hydrogenaudio.org/forums/index....amp;hl=allofmp3

Ripping from CDs burned from lossy sources. I pretty much stopped buying from them after that thread.

It's ran by russian mafia anyways, and it wouldn't be legal in any modern civilized country.

Russian mafia, sure...

What else then? I thought the whole country was, in the strict lead of Mr. Putin

You think a commercial outfit is likely to be using Lame? I doubt they care much about the quality just the incoming cash...

Mind you there are a handful people who upload to p2p sites saying they use EAC/Lame.

LAME MP3 encoding is pretty much the scene music release standard, buddy. It's more than a handful - nearly all releases are LAME-encoded.

Emusic (which is second only to the iTunes store in the U.S. for legal downloads) uses LAME to encode their MP3s. Emusic provides fairly low cost service for small labels, they have quite a bit of interesting stuff.

I have tried downloading the same album from both Emusic and iTunes Store, and usually found that the AACs from iTunes Store sound significantly better - but I cannot tell why as I do not know how they obtain and encode the material.

At least Emusic is cheap and allows me to experiment easily.

Though most of the idiots persist in using stupid settings like CBR, -m s stereo, and other cargo cult tweaks. It must be a bit depressing to be a lame developer and know that every dope in a scene "group" thinks they're smarter than you are.

From recollection current scene rules require that releases be VBR. Anything that is CBR would be old, internal or a non scene release.

That may be dependent on which album you downloaded from Emusic. Some older material (pre-LAME 3.9*) there was encoded using FHG.

Ahh... I believe some influential members of music scenes are members of HA too... raise your hand, please

(keeps my hands down)

LAME has been used for commercial use.

List of legal online music stores that use LAME:

Emusic
Bleep
AllofMp3

List of commercial software that use LAME:

Winamp
Sony's DRM Rootkit

Nearly all people who uploads mp3s on p2p sites are f**king n00bs who usally rip at 128 - 192 CBR in pure Stero and think Joint Stero destroys stero and thinks VBR is evil and usally rips in WMP (Fhg) or iTunes (crappy Fhg) and somtimes Winamp (LAME).

They're almost exclusively --alt-preset-standard (or whatever) nowadays, but they're almost all stuck on 3.90.3. A few scene groups use newer releases, but yeah.

Excellent... I assumed they would just use something, er, more basic.

>>>>Nearly all people who uploads mp3s on p2p sites are f**king n00bs who usally rip at 128 - 192 CBR in pure stereo and think joint stereo destroys stereo and think VBR is evil and usally rip in WMP (Fhg) or iTunes (crappy Fhg) and somtimes Winamp (LAME).<<<<

This is what I feared... I imagine that anyone dabbling in p2p would be fairly safe with 256 or 320k rips not to mention FLACs, though?

From what it looks like these days, allofmp3.com won't encode any more songs in the near future so ...

The .NFO files on some releases actually explain why 3.90.3 is still the standard for some groups although it's outdated. Something about compatibility... I don't really buy it, though.

They probably just don't know any better. 3.90.3 has been the standard "high quality" version for so long that I guess they just didn't bother seeing if there's a newer release that's been proven to be better.

Nah, the .NFO files I've seen even mention LAME 3.96. There's apparently a specific reason for it, but I'm pretty sure any properly informed person could easily find a solution other than using an outdated encoder.

Err, is there anything wrong with my LAME settings?
-m s -V 0 -q2

Simple stereo is is not superior to joint stereo and will only reduce the overall file quality, and -q2 is rather pointless. It would also be advantageous to use --vbr-new instead of the old default vbr routine.

isn't --vbr-new used by default now? or am i missing something here?

In 3.98 alpha I think it has become default, but not in the current stable release 3.97.

I'm still using 3.96.1

Any tips on what to do at this point? IE, what version of LAME to use, what settings (I don't want any setting other than -V 0/preset extreme, I just want to know the best way to get it done)? I've been primarily using Apple Lossless for almost a year now, I only make LAME MP3s for sharing with other computers in the house and in a way, added compatibility.

Read the topics at the top of the MP3 General forum

Alrighty, so let me make sure I'm up to snuff on LAME:

Using -m s is paranoid, stupid, and might actually be detrimental to sound quality.
LAME 3.97 is recommended over 3.96.1.
-V 0 --vbr-new/--preset fast extreme is the way to go for max, non "320 kbps CBR", sound quality.

i see. thanks for clearing that up. right now, i'm using 3.98 since...mp3 is lossy. i would use 4.0 but there has been very little development.

reason why 3.90.3 is preferred over 3.96+ is because the filecheck system that sites use dont recognize 3.96 properly. the script used in glftpd (the ftpserver that most sites use) isnt updated to ensure that when someone does encode vbr they dont try to fluff up the filesize.

Thanks for the explanation. The following is a sample NFO file that I just ran across in an archive I decompressed:



Looks like you're right .

I wired a few Franklins to this service over the last 8 month. Trying to stay on top of WTO news and it's ensuing $hitstorm so my credits will be used on time.

Encoding was either MPC, Braindead (Beta 1.14) or Vorbis, 256 var (Lancer SSE3MT - aoTuV r1). Can't say for sure if source had already been converted, files sound good (to me). Albums run 2 bucks, tags are perfect and distribution is fast.

I'd hate loosing this option - "mafia" or not - alternative domestically to buy legal content on the cheaps is a fading pipe dream, there are simply too many satisfied 99c suckers amongst us.

Torrent and ptp fetched content is a waste of bandwidth. I spot the occasional gem on usenet, but who's got the time? As of tonight, Russia is still encoding.

lame 3.90.3 is still used of huge codec centrals, because it is

- still the most tested one in mp3 codecs with amazing results

- still most compatible on file check progs for ftp sites / hdd scans

- integrated in all codec central's progs (that's the main reason why they still not switch to newer lame versions: they're simply afraid of unknown bugs)

- when people order their albums, they mostly select 3.90.3 encoded tracks to download (I don't know why, maybe when they see 3.90.3 encoded, they say it's great quality or whatever)

- almost all bugs of it are well-known of them


I asked some server admins who store audio data. I received most often these answers.

Also "private music societies" store all cds compressed to mp3 or other formats, where all files have the same codec (to provide highest compatibility). Small societies already have updated to 3.97, because they didn't detect any errors. But the big ones have a lot to do for updating... testing, testing, testing, updating the other files and so on. And some of them simply don't do it right (like transcoding from lossy to lossy and so on ) and then they say "uh, we don't achieve advantages with updating haha"...

That's the reason why I don't use public download servers for mp3 sources... You always don't know where the music really comes from, and that's why I don't like it.

as a conclusion I'd say it's not that bad that these rip societies don't update to newer lame versions. For those who download the albums, they often even don't know what mp3 means exactly. In other words, they won't can tell the difference between cd and lossy encoded either.
Who said in the times where 3.90.3 was standard that it was audible if it was a cd or a 3.90.3 encoded file? If the 3.90.3 encoded one sounds amazing enough, why changing codec then?

I don't want to say that 3.97 and newer is worse than 3.90.3, on the contrary, I think 3.97 and newer is a lot of better than 3.90.3 (speed, qual/size ratio, problem samples improved...), but if 3.90.3 also can do the job as a cd replacement encoder.... then it will be so forever (ok, let us see what the future will bring ) just like 3.97.

But I can't understand why some private rippers (who rip their own cds) still are using 3.90.3.... That's not explainable except of placebo.......
3.97 / soon 3.98 are recommended.... the points on wiki and HA recommendations are self-explainable.