LAME *Is* An MP3 Encoder

Frac writes: "LAME, which stands recursively for "LAME ain't an MP3 Encoder", might have to change its name. They have finally replaced the last bits of functionality of the ISO dist10 encoder code. Previously, Fraunhofer closed down all encoders based on the ISO mp3 encoder code (free or not), and LAME used GNU patching as a loophole to continue development. What this news means is that we now have a fully open-sourced (LGPL) mp3 encoder that Fraunhofer can't take away. Congratulations to the developers of LAME! "

Comments

Re:What's wrong with BladeEnc?

[DeeKayWon]

I used to use Bladeenc at 256kbps to make all my MP3s. But when I switched to a Fraunhofer-based package, I realized that high frequencies were suffering in the Blade encoding, and sharp sounds (like the sound of two drumsticks colliding) were severely blunted. I haven't tried LAME yet, but according to r3mix.net, it's better than the Fraunhofer. But then again, they didn't give much for comment on the specifics of the actual perceived sound quality between Fraunhofer and LAME. Human ears, not frequency diagrams, are the only real benchmark.

Re:Take an example

[eGabriel]

People use mp3 because that is what is being offered. By offering something else, I can influence users, in however small measure, to
change their minds and use free software.

Vorbis is unencumbered. MP3 is not. Even if
LAME is completely unencumbered at this time,
I will back away from it if there is a free alternative.

It is what I think users should demand, but if they won't, it is what I will provide.

I am not saying that in the future Ogg Vorbis will be free from some of the same problems we have with MP3. I have considered that point at length.

What matters is that our actions, should we move, clearly state that developers and users will not tolerate restrictive patents, will not tolerate
legal meddling... we just want to hear the music.

So it's a sidestep; it's a sidestep I feel good about.

This is much less about whatever is happening with the record industry and more about patents. We shouldn't need to consider either.

At present, it is questionable whether it is legal for me to even offer my own original music for download, as I used BLADEENC, not a licensed
copy of Fraunhoefer, to encode it.

That aside, what if an ISP suddenly decides that ALL mp3's are probably pirated (and how could they tell?) They will do what they have to in order
to protect themselves, and require that you remove
your legitimate files.

With a new file format, we dodge the mainstream for a time. Hopefully education will have time to catch up.

Anyway, I don't wish to stand near the Hindenberg when it spirals down. If you want to stand up for an algorithm that was released in poor faith and is synonymous with "piracy", I don't really think you are wrong for doing so.

Like I said, let's reflect a bit.

In which countries?

[Pseudonymus Bosch]

Fraunhofer is a German institution, but I thought that Europe have a more restrictive position on patents.

Can somebody tell which countries are the Fraunhofer patents issued for?
__

Re:Take an example

[Lord Kano]

With a new file format, we dodge the mainstream for a time. Hopefully education will have time to catch up.

It will only be billed by the RIAA et all as "The NEW MP3". Unless we stop their ignorance now, we'll see a domino effect. MP3, OV, WAV, AU, AIFF, and anything else that we use will be seen the same way.

A line must be drawn in the sand. Invoking the righteous indignation of Captain Jean Luc Picard in Star Trek First Contact This far, no further. We've tolerated your lies long enough. No more! Not now! Never again!

If you run from an angry dog, he'll chase you. If you turn your back on a mugger, he'll stab you. You can't avoid this fight, and delaying it will only make it harder to win.

LK

Some Facts

[Cinquero]

The MP3 patent is here:

http://www.patents.ibm.com/details?pn=US05579430 __

The licensing related information is here:

http://mp3licensing.com

One note: not only MP3 _encoders_ are 'protected'... MP3 data (streams, music...) is 'protected' too! Just have a look at the conditions for licensing.

Re:Some Facts

[rcw-work]

You can claim that anything is protected, the first amendment gives you that right (in the US anyway).

Whether the claims of protection are actually true or not is another discussion, one in which the courts have the final word.

LIME?

[Piic]

Well, without a lick of salt and following shot, it just doesn't have the same flair.

Quality of LAME exceeds Fraunhofer (tested by c�t)

[FutileRedemption]

ok, thats just half of the truth.

What ct found in their test lately (using the ISO patch LAME version) was that in all disciplines LAME is as good or even slightly better than Fraunhofer.

Except for speech at 32kbit/s, where Fraunhofer was better (I suppose we can live with that).

Bottom line: Forget about all other encoders, and use LAME instead. Your (and potentially other peoples) ears will thank you for this!

I hope this is an unnecessary reply.....

[TuRRIcaNEd]

...as I suspect this is a troll, however, Just this once I will take on your argument.

I am going to say this once, and only once:

WE ALL PAY FOR ALMOST EVERY MICROSOFT PRODUCT ON THE WINDOWS LICENSE THAT IS INCLUDED IN THE COST OF EVERY PC.

I realise that the caps are unnecessary, but here goes. You wanna pirate MS software? Presto! The price of all MS software goes up. Integrating IE into Windows is a canny move because they can hike the license cost even more, and snare those who don't know into believing they are getting something for free. You paid for IE and WMP when you bought your machine. Either that or you pirated the software, and every one of us who is forced to run Windows, for whatever reason (And, unhappy as it is, most of us are), is paying for you to use IE and WMP. It's because of practises like these that they are in the position where Gates can sit behind his wall of money and claim he's doing the computing community a service. They're also cheapskates, as they have only paid for the 56k-limited version of Fraunhofer to ship with Windows, leaving us to pay for full versions with whatever software we have to buy, while claiming 'full MP3 functionality' under Windows. If it's hobbled, they should say it's hobbled.

The reason projects like LAME, GNU and the like are necessary is because shareholder greed is dictating to the computer industry at the moment. What started as a communal effort to make existence better and easier by a bunch of techies is now an industry behemoth making a very small number of people *extremely* rich. By simply taking from them, by pirating, or whatever, you only give them an excuse to make more money out of the honest. By replacing what they do with a free (as in either beer OR speech) solution of equal or better quality, you obviate the need for their product, and their existence. By obviating the need for their existence, you return computers to their original purpose; to help people perform tasks that they could not perform as quickly, or at all, before.

RIAA?

[Signal 11]

Well, I'm sure Metallica will be calling you up soon to protest how you could use this to potentially interfere with their copyright and cause them to lose sales.. and then they'll sue you for damages that haven't yet occurred...

*cough*

Re:RIAA?

[Pinball Wizard]

Hmm...someone is obviously abusing their moderation powers at Signal 11's expense.

Take a look at his user info. Most of his latest comments are either 0 or -1.

In particular, I am reposting this comment.

I'd like to remind people that the majority of open source projects DO fail. Remember, even the best baseball player of all time couldn't even hit the ball half the time. Go easy on these people, they're putting real effort into things.. and even if it doesn't turn out they deserve our praise. For every Apache or Linux there's hundreds of lower quality, less-developed programs out there. That doesn't mean sourceforge, or open source, or the authors.. suck. What that means is that you ought to be a smart consumer and not use something that doesn't fit your criteria. Nobody here said you have to use software that sucks - you are a consumer.. act like one: purchase/download the products that best meet your needs.

A perfectly decent comment, certainly not deserving 2 troll and 2 overrated moderation points.

Whoever is beating up on Signal 11, please stop. You are not going to get away with this much longer, you know.

Its the best encoder as well.

[kperrier]

Not only is it now an encoder, its one of (if not the) best ones out there. Look at http://www.r3mix.net/ for a pretty through analysis of encoders. DISCLAMER: I am not affiliated with http://www.r3mix.net/ in any way, shape or form. Kent

Re:Its the best encoder as well.

[Greg W.]

Ars Technica has a review of MP3 encoders -- they compare Bladeenc, LAME, Xing and Fraunhofer.

For those too lazy to read it, Fraunhofer comes out the winner, followed closely by LAME. Bladeenc and Xing get ripped apart (no pun intended).

Re:Not quite...

[TheCarp]

*Bzzt* you lose.

They have a headlock on anyone writting an encoder
in a country that is backwards enough to allow
Software patents. In countries that do not have
a patent system, or whose patent system is smart
enough to not allow software patents, then they
have nothing.

Why use MP3 format?

[Adam Bertil]

I don't want to go in the battle what encoder is the best...(Fraunhoffer is the best then Bladeenc then LAME if you use 128KBit encoding :-) ) But there are now alternatives that are patent free... Vorbis is very good(not as good as Fraunhoffer on some files YET) its better than Bladeenc adn LAME. Vorbis can be found here http://www.xiph.org/ogg/vorbis/index.html Those who knows how CVS works just download the source and compile. It only seems to work with redirect/pipes, so use /vorbis/examples/encoder_example. And redirect a wav file to it cat music.wav | ./encoder_example > music.ogg Now use the XMMS plugin and listen, great quality some files bigger are bigger than MP3 but there are no patents involved. Don't know if it works for windows users but you could give it a try...

Re:Standard justification

[TheCarp]

Your argument makes an assumption, that not
everyone believes is correct.

You assume that the means of "Using a loophole
in the law" or "breaking the law" is necissarily,
in and of itself, wrong.

I would personally argue, that the law can be
either right or wrong. The law is an absolute
measure of nothing more than the law. It is
not a measure of morality. It is only immoral
to break the law, when the action is immoral.

As for equating free software with the russian
revolution.... yea some of the ideas are similar,
but the russian revolution suffered from corrupt
leadership that was more interested in grabbing
power for themselves than actually helping the
people.

In the free software movement, noone is running
around trying to solidify their own powerbase,
at the expense of the masses.

As for communism....why point at the russian
revolution...what abou tthe farm (www.thefarm.org)
which is a community right here in the US that has
been around since the 70s and still is...they have
been experimenting with progressive forms of
communism quite sucessfully for years.

Avoid the legal issues about MP3...

[kcarnold]

A lot of the arguments in this discussion seem to center around the legality of MP3 encoders like LAME. Some people contend that all MP3 encoders are illegal. To avoid all this big mess, I propose that the open-* people among us adopt a different and superior, IMHO, format, which would be Vorbis. We discussed this format earlier on Slashdot, but I think it deserves being brought up again.

For a little background, Ogg Vorbis is a completely open format with no patent issues or other messiness. It was developed by Xiphophorus. Theoretically, Vorbis has a higher quality than MP3 at almost all bitrates. For more information, see the FAQ on the page I linked to above.

I have modified the example encoder in the CVS tree to make it much more user-friendly, added detection for the WAV header, put in a status display, etc. My modified encoder is called Vorbize. XMMS and WinAMP plugins are available.

I encourage everyone who believes in open formats to use Vorbis. It's Just Better (TM).

Remember, just because Everybody Else uses [Windows|Mac] doesn't mean we shouldn't use [Linux|BSD|whatever]. Apologies to Mac users.

Vorbis 1.0 file format is frozen/ready for use

[xiphmont]

The Vorbis 1.0 file spec is now frozen; the libs and headers in CVS are compliant. A formal Vorbis 1.0 release will happen as soon as a better encoder/decoder util is in CVS (and a bit more tuning happens that will *not* affect the format). Streams encoded as of now should be immortal. Once Kenneth rebuilds vorbize with/against the updated libs, those streams are the Real Thing.

BTW, the changed codebooks do not affect the bitstream format; the codebooks are packed and included in the bitstream itself. The changes we made to the format recently were more mundane than that.

Monty
http://www.xiph.org/

the ISO sources weren't the issue

[daw]

Fraunhoffer shut down previous mp3 encoders based to patents which ostensibly apply to ANY mp3 encoder, regardless of implementation. Just the fact that the ISO sources are expunged doesn't mean LAME is in the clear.

Re:the ISO sources weren't the issue

[_dim]

From the MPEG, Patents, and Audio Coding FAQ (recommended reading):

• Q15. If I don't use their source, can I make my own MP3 decoder without paying FhG?
• A15. Legally, FhG may or may not have rights regarding patented technology that is necessary to make an MP3 decoder. If they do, it is within their right to enforce it and prevent you from making any MP3 decoder, whether or not you had help from them to do it.
• Q16. If I don't use their source, can I make my own MP3 encoder without paying FhG?
• A16. If you infringe on their techniques, it is within their rights to seek recourse, whether or not you had help from them, or whether or not you intentionally or knowingly infriged.
• Q17. If I don't use any of their techniques, can I make my own MP3 encoder without paying FhG?
• A17. Yes.

So it seems you will have to just invent your own techniques/algorithms for encoding, but what this precisely means is not entirely clear.

In the case of LAME, because there now is a totally independent implementation with maybe totally independent techniques/algorithms, it could really be free. At least I hope so. :-)

Oh, and don't forget that not all countries allow patents on algorithms, like Sweden (where BladeEnc comes from).
--

Re:the ISO sources weren't the issue

[MasteroftheVoxel]

Yes the compresion algorithm is patented, so you can't just do your own implementation of the algorithm.

But, you _can_ make your own compression algorithm that still conforms to the mp3 standard. As was said earlier, there are _many_ ways to encode mp3s, and they patented the best one. By pulling open code from various sources (it seems) LAME is able to come pretty close to Fraunhofer's ideal solution.

BLADE, by the way, sucks.

Umm, no...

[Sir_Winston]

They could patent anything they want, but only patenting their specific code or methodology would stand up in Court. MP3 is a publicly-usable standard, just like MPEG-1, PNG, JPEG, etc. It's not like GIF, which is patented under increasingly stupid terms. But the interesting thing about MPEG-3 (MP3) is that, unlike with MPEG-1, the standard refers only to the way the finished file is constructed and how it can be decompressed, rather than covering the encoding process. Thus, there are many ways to encode a file to the MP3 standard, not just one standard implementation, but unlike AVI you don't need the codec installed for the specific encoder--a standards-compliant MP3 decoder will decode an MP3 encoded with any of the codecs. The problem is that the Fraunhoffer reference implementations were being widely used by people, without licensing fees, in many products, and Fraunhofer wanted money for its implementation. That codec arguably produces better-sounding results than others, like BLADE, preserving more of the important wave characteristics which make a song sound true to the original .wav, while cutting more of the unnecessary junk which can't be perceived by the human ear and thus just takes up space.

But, I repeat, MP3 format can be used by anyone, and it's only a certain way of encoding to that file format which was covered by Fraunhoffer. Any MP3 encoder which doesn't use Fraunhoffer's implementation will therefore be perfectly legal. In any event, I suspect that there'll be some way to plug any codec you want into LAME, and the Fraunhoffer codec is easily obtained from the Net.

Re:Umm, no...

[daw]

Sir,

How art thou full of shit? Let me count the ways.

They could patent anything they want, but only patenting their specific code or methodology would stand up in Court.

I don't know whether their patents would stand up in court. Neither -- OBVIOUSLY -- do you. The point is that the question of whether LAME will get harrassed by Fraunhoffer in the future is not settled by them expunging the ISO sources, since Fraunhoffer/Thomson claim to have patents covering any mp3 encoder. Whatever their merits, their claims are not (as you seem to think) grounded in their implementation. See

http://www.mp3.com/news/095.html

for an example.

As for "their specific methodology," well of course LAME uses their specific methodology -- it was created by slowly rewriting the ISO sources, bit by bit. It works the same.

MP3 is a publicly-usable standard, just like MPEG-1, PNG, JPEG, etc. It's not like GIF, which is patented under increasingly stupid terms.

It's a public standard. As for publicly usable, it's covered by many patents, just like -- well -- GIF. Here's one

http://www.patents.ibm.com/details?pn=US05579430 __

This covers:


A digital encoding process ... in which scanned values of the acoustical signal are transformed ... into a sequence of second scanned values, which reproduce the spectral composition of the acoustical signal


i.e., essentially any audio compression in frequency domain. Since the mp3 format just consists of a bunch of quantized frequency information it's hard to imagine what you claim -- something which produces mp3 files that isn't covered by this patent.

But the interesting thing about MPEG-3 (MP3)

mp3 is short for MPEG-1, layer III. There is no MPEG 3.

Thus, there are many ways to encode a file to the MP3 standard, not just one standard implementation, but unlike AVI you don't need the codec installed for the specific encoder--a standards-compliant MP3 decoder will decode an MP3 encoded with any of the codecs.

What's your point here? AVI is unlike mp3 in that it (like Quicktime) is just a wrapper format for multimedia files. But the particular compression algorithms it supports (including -- gasp! -- MPEG-1 layer 3) are available in many implementations and any standards-compliant decoder for some format will decode any AVI encoded to that compression format. There are also lots of implementations of GIF compressors. How does any of this support your confused ideas?

The problem is that the Fraunhoffer reference implementations were being widely used by people, without licensing fees, in many products, and Fraunhofer wanted money for its implementation.

No, that wasn't the problem at all. The problem Fraunhoffer saw had nothing to do with any particular implentation, since they claim to have patents covering any implementation. See the mp3.com URL I mentioned earlier.

That codec arguably produces better-sounding results than others, like BLADE

Christ. I'll try to use small words, okay? BLADE is a lightly modified version of the ISO sources, which is to say -- surprise! -- the Fraunhoffer reference implementation. Hard to see how it could sound better. Perhaps you're thinking of the Fraunhoffer Radium codec, their private, closed-source implementation that they license out to software vendors.

In any event, I suspect that there'll be some way to plug any codec you want into LAME, and the Fraunhoffer codec is easily obtained from the Net.

Have you ever used LAME? Do you know what it is? It's just a codec. If you plugged another codec into LAME there'd be nothing left. That's like saying I suspect there'll be some way to plug any word processor you want into Word Perfect, so you can always use MS Word.

Think, boy, think.

Re:Umm, no...

[Sunracer]

> But the interesting thing about MPEG-3 (MP3) is
> that, unlike with MPEG-1, the standard refers only
> to the way the finished file is constructed and
> how it can be decompressed, rather than covering
> the encoding process.

MP3 == MPEG-1 Layer 3, not MPEG-3.

it's nice

[operagost]

Only prob is that FreeBSD and OS/2 don't support the state changes required for SSE optimization. MMX works though, and I believe (not sure) 3DNow!
It's fast as hell, though, with excellent quality.

Re:Just use the apostrophe, like RMS

[Anonymous Coward]

What's the status of Blade, by the way?

it's done, it has been done for two years. it stars wesley snipes and kris kristofferson, you can get the dvd at circuit city. pretty good but very violent so I hope your into that sort of thing?

Re:Not to rain on LAME's parade but...

[Frac]

Also, the last time I checked, LAME was GPL and not LGPL...

You haven't checked for a LONG time.

LAME 3.52beta November 8 1999

By permission of copyright holders of all GPL code in LAME, all GPL code is now released under a modified version of the LGPL (see the README file)

Go get your free Palm V (25 referrals needed only!)

Ogg/vorbis doesn't have the same problems

[SurfsUp]

There's nothing wrong with MP3.

Yes there is - the encoding/decoding algorithms are patented in such a general way that producing a codec that doesn't violate the patent is nigh on impossible. (If you can do it, please let us know!) Your only hope would be that the patent won't stand up in court, and if you want to test that, be my guest. Let us know how it turns out.

What makes you think that Ogg Vorbis wouldn't suffer from the same public relations goof ups that MP3 has?

The fact that I've been to the site, read the FAQ, gotten the code out of cvs, compiled it, read it, run it, read the license and joined the mailing list. Good enough for you? I also looked into my crystal ball and saw many streaming ogg plugins there.

The fact that the ogg encoder is both competitive with MP3 and explicity/entirely (L)GPL makes it a no-brainer for adoption by streaming broadcast servers.
--

it's the algorithm

[stevens]

Just the fact that the ISO sources are expunged doesn't mean LAME is in the clear.

Exactly. Apparently the compression algorithm is patented, not just a sample implementation of it. Use the implementation, whether it is your code or theirs, and you might be violating the patent.

I'd like to have more info on this. The LAME site doesn't seem to contain any speculation about whether or not the software is encumbered.

Steve

Interesting link...

[fReNeTiK]

Incidentally, ArsTechnica did some intersting comparison tests between various MP3-Encoders recently.

Re:Do as I say but not as I do

[scumdamn]

Read above. MP3 is a standard. The only thing that's patented is the way that it's encoded. LAME uses a different method of encoding MP3 files.

Re:MP3 encoding?

[Anonymous Coward]

Fraunhofer was able to close down encoders based on the ISO mp3 encoder code. Therefore LAME wasn't allowed to offer an MP3 encoder for download, which was based on this code.

But they couldn't stop the distribution of the original ISO code. So LAME decided that instead of giving you a usable program to download, they'd tell you to download the ISO source, and then use their patches to modify it.

Re:Believe it or not, some of use DO still use x86

[scumdamn]

And Microsoft thinks that everybody is using Windows.
There will always be companies that don't release software for competing architectures. Intel is the least common denominator. If we waited to talk about a program until it was released to all the politically correct platforms (LinuxPPC, Alpha, and Sparc) and all the politically correct OS's (OS/2, *BSD, Solaris) we would be sitting around here with our thumbs up our butts.

Not quite...

[mindstrm]

Sir.. the actual method of encoding is patented (ie: breaking the sounds down however they do it, and encoding it just so..).
The ISO sources were simply an example.

If it's an encoder that produces data that is decoded by mp3 decoders, as per the fraunhoffer patent, then it falls under their patent, unless, of course, there is a 100% completely different mathematical way to arrive at the same results..

Re:Not quite...

[molog]

I have to admit that I do not know patent law that well. I thought that you could patent an algorithm, but not just doing something. RSA handles encryption but if applies to the process of encrypting data then all encryption would be covered by the RSA patent. Thus, their algorithm is patented, but if someone finds a different way to encode the sound to MP3 then that's that I thought. Again, I don't know the laws specifically. If I am wrong can you point me where to go to find out?
Molog

So Linus, what are we doing tonight?

Re:Parts of standards _are_always_ patented!

[Score Whore]

Because it's hard to make a good encoder. Just like MPEG-1/MPEG-2 are standards, the ways of encoding them are an art form. Fraunhoffer has a big lead time advantage on their competitors. Even with all the hard work that has gone into LAME, they wouldn't be anywhere near where they are if they hadn't had the Fraunhoffer/ISO codecs to work from and test against.

Not a stupid question...

[Svartalf]

The ISO implementation has been available to the public as a referant for implementors of systems using MPEG-1 Layer III (MP3) codecs for sound. It was still freely downloadable when Fraunhoffer cracked down on all the encoders. LAME consisted of a dramatic enhancement of the psycho-acoustic model used to encode MP3's in the form of a patch set for the ISO code. Since it's a differing model, being slapped on top of the currently available codec, I suspect Fraunhoffer has less room to harass them- it's not a program or even their model (which is what is patented- the other portions, like the MDCT, etc. are in the public domain.). It all depends on what the model LAME is using is based upon, etc.

Very cool... but let's reflect a moment

[eGabriel]

MP3 is a very good audio encoding format. I am
happy with the sounds and the compression.

On the other hand, it has become associated with
every sort of horrible legal problem an algorithm
and file format could encounter.

Now that Vorbize exists, I think we ought to
disassociate ourselves from this cursed gem called
MP3. It is clear no one involved with its original design wishes us to have it, and soon
silly laws will likely make it a crime to even
say "mp3" aloud in mixed company.

I have a good deal of my music available online,
(www.devo.com/fabco), and am changing to Ogg Vorbis as soon as the spec appears to be stable.

I know it will make it more difficult for listeners at first, but I hope that goodwill toward free software will prevail.

Re:Parts of standards _are_always_ patented!

[Mr_Dyqik]

Otherwise, why bother opening up the standard if you can't recoup licensing fees? Because those standards can be used to ensure interoperability between your product and the content developed for other adhering products. Thus you can produce a media player of some sort and be sure that there is some media to play on it, without having to go out and get people to convert to your non-standard format. People won't buy a media player of some sort if there's nothing to use it on. Although if you can patent parts of it _and_ be sure of marketing success then it may be an idea to patent the methods, and wait for people to jump on the bandwagon.

Re:LAME + Liveice

[otis wildflower]

people like me seem to have spent the time finding audio samples which don't encode well and complaining.

Then you're doing your job..

:)
Your Working Boy,

What's wrong with BladeEnc?

[istartedi]

There was a lot there about how good LAME is, but I didn't see BladeEnc in the comparison.

They said you need 256kbits for CD quality. From BladeEnc site: [BladeEnc] Supports the following bitrates: 32, 40, 48, 56, 64, 80, 96, 112, 128, 160, 192, 224, 256 and 320 kBit/s.

Re:What's wrong with BladeEnc?

[barleyguy]

I've personally used Blade Encoder, as well as older version of the Franhofer encoder. Blade had significantly lower quality at the same bit rate, though going to a higher bit rate can even things out.

I'd be willing to bet that if you ripped the same song at a low bit rate (64-128 or so), and compared the quality between Blade Encoder and LAME, LAME would have noticably better quality. At 256 and higher, any properly working encoder should give you CD quality.

Um, this might not be legal yet...

[Millennium]

Remember, Scum^H^H^H^HFraunhoefer didn't patent the ISO code. They patented the algorithms they use in their encoder. That means any code that uses that algorithm violates their patents.

Now, despite what FhG says, it probably is possible to create an MP3 encoder without using those algorithms. But only if the LAME team managed that is LAME totally legal. And by the way, no one has managed to do it yet.

Yeah, it's a stupid patent; you're not supposed to be able to patent mathematical equations. But it's the way it is, and we've got to work within the system while we fight to change it.

Take an example

[Lord Kano]

In the past few years firearms have been used to kill schoolchildren in several high profile incidents.

Ammonium Nitrate was used to kill over 160 people in one horrible act.

During the period of time when the US was involved in the Vietnam conflict, more people were killed in domestic car accidents than were killed in combat in SouthEast Asia.

Do you see the NRA, Sturm Ruger & Co., K-Mart, Wal Mart, Ryder, Chrysler, General Motors, Ford, or any of their satisfied customers/clients/members backing away from those products?

Why should we? There's nothing wrong with MP3. What makes you think that Ogg Vorbis wouldn't suffer from the same public relations goof ups that MP3 has?

People use MP3s to rip off music because that's what's easy.

People use .zip files for wares because that's what's easy.

People will use whatever vehicle is necessary to do the things that they want to do. Without MP3, there will be something else.

If the record companies weren't price gouging, MP3 wouldn't even be an issue. BTW, the FTC came to an agreement with record companies today which *might* bring down the prices of CDs.

LK

Re:Take an example

[Lord Kano]

You are correct, I should have been more specific.

Do as I say but not as I do

[jeroenb]

I think it's pretty pathetic that when Microsoft uses a loophole in an agreement to start exploiting someone else's technology to their own ends people use it as an example of MS' evil, but when some GNU project does the same thing they're applauded for their efforts...

Sure, I agree that software patents suck and that everybody should be allowed to write his own encoder if he wants to, but that's beside the point here.

Re:Do as I say but not as I do

[dattaway]

I think it's pretty pathetic that when Microsoft uses a loophole in an agreement to start exploiting someone else's technology to their own ends people use it as an example of MS' evil, but when some GNU project does the same thing they're applauded for their efforts...

Close, but no cigar. Microsoft also offers the product for free under a restrictive license. This puts the competing company at a severe economic disadvantage. When the competitor folds, Microsoft is then free to jack up the prices. By then, the consumer base is hooked on that product. You can bet upgrades will force said consumer to use a new version in order for existing documents to work. Its all about monopoly practices.

When the source is free under a non restrictive license, you will not be forced into submission.

Kudos!

[iCEBaLM]

Kudos to the LAME guys! I use LAME myself and am very thankful that you guys kept the project going and have made it so far as to replace all of the ISO code!

Now, new names:

LAME An MP3 Encoder
LAME Amazing MP3 Encoder
LAME Acronym, Must Elaborate
LAME Autonomous MP3 Encoder
LAME Another MP3 Encoder
LAME Anybodies MP3 Encoder

-- iCEBaLM

MP3 encoding?

[Anonymous Coward]

I am certainly no expert on this stuff, so I was hoping someone here could explain this. Does Fraunhofer have a patent on MP3 encoding, so that mp3 encoders had to be licensed? Is that what prevented an open source encoder previously to this?

The article says Fraunhofer closed down all encoders based on the ISO mp3 encoder code (free or not), and LAME used GNU patching as a loophole to continue development. What do they mean by "GNU patching"? Loophole in the GNU license? What was the loophole?

Sorry if I sound stupid. I'm just trying to understand better.

GNU patch

[yerricde]

GNU patch is a utility, part of a GNU system, that uses .diff files produced by diff to make modifications to a file.

Believe it or not, some of use DO still use x86..

[agravaine]

Please do not assume that everyone reading uses x86 hardware

He didn't *assume* it, he flat out told you that it uses x86 assembler and figured you'd be smart enough to figure out if it's useful to you.

Are you saying he shouldn't have posted something that might be of interest to the >= 90% of Linux users out there who have an x86 machine because it might have hurt the Alpha and Sparc users' feelings?

This is a trend that has been growing and growing on slashdot, and it's really starting to get on my nerves. Anytime somebody does something cool that runs under Linux86 (tm) someone has to rain on their parade, and complain about lack of [Suse|Slackware|Alpha|Sparc|Arm|ucLinux|PalmPilot| TRS-80] support, rather than say - "hey cool! they started supporting the most common flavor of linux - it should be much easier for them to add support for my platform now!"

My favorite (and I do NOT mean to imply that the poster I am currently replying to is this clueless) was a post I saw in a discussion about IBM's new JVM, bitching that IBM was just supporting x86. After all, the poster argued, how hard is it to just recompile the JVM on a sparc? [In case you don't get the irony of this naive poster's question: IBM's JVM does just-in-time compilation of Java Bytecode to native x86 instructions, so getting it to run under SPARC is NOT just a simple recompile, it would have required a serious extra development effort to integrate their JVM with a different compiler backend that emitted optimized SPARC instructions. - and IBM doesn't even sell SPARC-based products; in fact, they compete against them!]

DONT use GOGO instead

[szyzyg]

It may be fast - but it's based on a really old version of LAME, and for some unknown reason they desicded to change half the command line options....

I'm not impressed ;-)

LAME + Liveice

[szyzyg]

LAME is so good it's now the default encoder with liveice. In fact - this reminds me that I need to go and update the web pages ;-)

Mark, Gabriel and Takehiro are clever people - they seem to have made the bulk of the changes. OTOH people like me seem to have spent the time finding audio samples which don't encode well and complaining.

It's also pretty good at low bitrates - try these icecast servers -
http://abv-icy1.myplay.com:8000
http://abv-icy1.myplay.com:8010
http://abv-icy1.myplay.com:8020

they run LAME 24 hours a day.

Ripping off?

[Millennium]

I can't believe that you are actually *boasting* about the fact that Fraunhofer can't protect their intellectual works.

What do you mean? Fraunhoefer's got its intellecual works perfectly intact. That would, incidentally, be the code it produced. And the LAME team basically said "Screw you, Fraunhoefer, we can make MP3's without using your precious 'intellectual works'" and went and wrote their own. What's the matter with that? That's perfectly legal and moral. We didn't want to use their stuff, so we made our own stuff. When has that ever been bad?

Re:Ripping off?

[Millennium]

The purpose of patents is to reward inventors.

I'll start with your last statement, because it's the easiest to definitively disprove.

Patents were seen as very important by the founders of this nation. So important, in fact, that the patent office is defined in the Constitution, before the office of President is even mentioned. And why did they do this? Well, they spell it out right there. Article I, Section 8, Paragraph 8: "To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries;"

The emphasis is mine, but look at it again. The purpose of patents is not to reward inventors; that is only a means to the real end. Patents are actually made to benefit the people by encouraging the growth of the arts and sciences. It works like this: an inventor agrees to give up all rights to an invention and release it all to the public. In return for this, the government will first allow that inventor exclusive rights to that invention for some time (currently up to seventeen years, which was fine in less technologically-advanced times but is now far too long; technology makes a seventeen-year time period way too easy to abuse).

MP3 was an invention about math and science, not about code.

Hold on; you're confused. Fraunhoefer did NOT patent MP3. It patented several algorithms which are typically used in creating MP3's, but FhG holds no patents on MP3 itself. It would like you to think it held "virtual patents" on MP3 by means of "irreplacable" algorithms, but this is not necessarily the case. It is quite possible that the LAME team has managed to completely replace all of the patented algorithms with new ones, not stepping on a single patent in the process. This is their right, just as it was FhG's right to invent the "old" algorithms. It's a little thing called competition.

People had to figure out how human audio perception works...

This was not FhG's doing, I should point out. They built upon earlier research for that part of the bargain. Now, to continue...

and figure out how to mathematically transform audio signals in a way that reduced their information content without altering how they would be perceived by humans.

That last bit is arguable; I know plenty of people who claim they can hear the difference (I can't, and I don't see how they could, but I'll believe them). It's not that relevant, though. The fact remains that Fraunhoefer never patented MP3; no one did, no one can. If someone can make an MP3 encoder that doesn't tread on a single FhG patent, then more power to them; that patent has been abused in some pretty disgusting ways and a legitimate way around it would be nothing short of poetic justice.

Re: Some Facts revisited

[Malogato]

The patent you referenced doesn't appear to actually protect the storage format (I just read it and all the claims, and I don't see it that way). Although with all the very specific claims (it's a very well-designed patent compared to many I've seen), it does look difficult to create an MP3 file out of an analog source file without using ANY of the mechanisms outlined in all the claims on this patent. Whether or not this patent will be enforced remains to be seen - but I can't imagine how one might go about encoding an MP3 without using tabluar lookups and the other mechanisms that appear to be protected...

Law vs. morals

[David Jao]

Microsoft and Fraunhofer uses loopholes in law to play dirty tricks that are perfectly legal but very harmful to the public. In other words, they are legally right, but (in the views of many) morally wrong.

Free software groups and projects like LAME sometimes violate the letter of the law, but their actions are very beneficial to the public (I know that I for one have certainly benefited from LAME's efforts). These groups may be legally wrong, but as far as I am concerned they are morally right. They are doing the right thing, and while we always try to fight to change bad laws, in the meantime the law be damned.

Remember that at one point segregation and "separate but equal" was written into the law. I'm not saying that patent law is as harmful as legal segregation. My point is simply that the law is not always morally right. In an ideal world, the law would always be morally right, and the public could get away with just obeying laws unquestioningly without thinking for themselves whether the law deserves to be obeyed. Unfortunately, we don't live in an ideal world. I hate to break it to you, but in this world people do have to think for themselves regarding the merit of our laws.

A final point which everyone seems to have ignored is that LAME distribution could (and probably will, if necessary) be made totally legal simply by moving the distribution site to a place that does not honor the Fraunhofer patents.

Not to rain on LAME's parade but...

[Tord]

...the ISO code had NOTHING to do with the patents!

It's a common misconception that different MP3 encoders are affected by the patents since they are based on the ISO code. That is NOT correct.

Patents deals with technology and procedures, copyright deals with the source code!

The reference code is provided by ISO to help people understand MP3 encoding/decoding and they have nothing to do with the patents (except that they accepted patented or patent pending technology into the MPEG Layer 3 standard, which they should have a big kick in their butt for).

Fraunhofer IIS and Thomson Consumer Electronics owns a lot of patents on technology used in MP3 encoding/decoding and they will demand a license from any encoder that uses their patented technology no matter if it's based on the ISO reference source or NOT.

Personally I don't know why the LAME team always have taken the approach of not distributing the entire source but only a patch, but I guess they decided to play it safe. Distributing the ISO reference source doesn't breach any patents, but it might possibly be a copyright infringement against ISO.

Take a look at their page, does that say anything about the patents not being enforceable against compiled versions of LAME anymore? No it doesn't, it just says "All ISO code removed!" among the new features, which of course is a nice milestone to reach (no possible copyright infringement, having complete masterhood of the code, having replaced all old bug-filled code with new clean code etc.), but doesn't affect the patent situation.

Technically it should be possible to create a completely patent free MP3 encoder by carefully reviewing all the patents (17 patents in total, unless their lawyers have "forgotten" to send me some) and then making sure that whatever implementation you go for doesn't use any of those specified processes, which is bloody hard since these patents were designed to intercept any attempt like that. Then if you succeed you would probably still have to go to court since they would sue you anyway, hoping that your implementation is close enough to get you stopped.

Also, very few people knows this, but there is a ticking bomb hidden in all this. Fraunhofer and Thomson don't have all the mp3 related patents, they are just the ones who have decided to demand a license for the use of their technology and pulled their patents into a common pool that you can license. More companies are claimed to have patents on mp3 technology (they are listed in the ISO documentations), but they are currently not enforcing them. What if they suddenly start to demand licenses for the use of their technology? Then it doesn't help that you have Fraunhofer/Thomson's permission, you also need another license to go on...

Also, I think that Slashdot should have checked this a bit more throughly before posting it (like checking with one of the LAME developers), the fact that they say "No more patching! Full souce code distribution since all ISO has been replaced!" and nothing about patents should have raised warning signs alone...

Also, the last time I checked, LAME was GPL and not LGPL...

Tord Jansson
BladeEnc Creator

L.A.M.E.

[Stavr0]

Litigation At Metallica Excessive
---

What's patented about MP[EG1audiolayer]3

[yerricde]

IANAL

The only thing that's patented is the way that it's encoded.

All MPEG1 audio layer 3 encoders encode the same way (fourier and MDCT to frequency domain, quantization according to some psychoacoustic model, huffman coding of the output). That's patented. Someone may patent a new psychoacoustic model, but any encoder that outputs MP3 streams infringes Fraunhofer/Thomson's patent unless it is licensed.

<rant> Boycott RCA and GE; Thomson owns RCA, GE, and the MP3 patent. Boycott Sonny and Cher; Sonny Bonehead retroactively increased copyright terms to unconstitutional levels. Don't boycott Metallica; just stay within their license by copying live CDs instead of studio CDs. </rant>

RIAA Warning posted on solaco.org

[small_dick]

Before using any lame or sulaco.org product, be careful to review the RIAA mp3 warning. It could save you from a MAJOR lawsuit.