Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: For the sake of argument: is FhG mp3enc superior to LAME? (Read 10392 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

For the sake of argument: is FhG mp3enc superior to LAME?

I'm asking this, because that discussion (with one party arguing for FhG, the other for LAME) sprung up not too long ago on IRC.

I was quite surprised, because since Fraunhofer stopped marketing their mp3enc and fastenc encoders quite a while ago, I've been under the impression, that LAME has become superior to it during the mid to late 2000's. I was arguing citing various ABX tests, performed by groups of people, but that was dismissed as anecdotal, which is not entirely incorrect.

As LAME saw (and still sees) development, I remember seeing improvements over time from version to version. Also, as I understand it, LAME performs better with slightly lower target bitrates than FhG. Granted, MP3 doesn't perform well at all in general with lower target bitrates, no matter the encoding variant, whether it's constant bitrate, variable, etc. However, my understanding is, that the quality comparison very much depends on the target bitrate that's looked at, etc.

I have to admit, that I don't have much experience with the FhG encoders, though. I don't remember ever using them for any stretch of time. Hence, I couldn't argue for LAME, simply because I don't have enough 1st encounter experience with it.
I remember mp3enc being bundled with Winamp. Since I'm not a Windows user, I can't really confirm for how long that has been the case, and if it's still the case, even.

I specifically *don't* mean MP3pro, btw.

At the same time, they mentioned, that since the last patent held by Technicolor has expired, FhG has released the patent into public domain, is that true? I just know, that since the patent expired, FhG doesn't collect licensing fees for it, etc. but as it being released into public domain actively by FhG seems confusing to me. Also, I couldn't find any information about that claim anywhere.

I can kinda see, how it is next to impossible to argue for one or the other, but I'd at least like some commentary on those claims.


Re: For the sake of argument: is FhG mp3enc superior to LAME?

Reply #2
Link is 404-ing

Re: For the sake of argument: is FhG mp3enc superior to LAME?

Reply #3
Lame is still being updated though. I don't think fhg is.
According to http://lame.cvs.sourceforge.net/viewvc/lame/lame/doc/html/history.html
LAME 3.99 was released 5+ years ago.
Unfortunately since then there were only bugfixes and other non-quality related changes.

LAME 3.100 changelog: "PSY model tunings, to improve audio quality of problem sample 'lead voice'" doesn't count.
It was an attempt to improve tonal sample but instead it increases bitrate for all kind of material without filtering actually tonal samples. So it doesn't work as it should.

P.S. Helix is another interesting MP3 encoder. It has an excellent quality and it's extremely fast.
https://hydrogenaud.io/index.php/topic,113324.msg933009.html




Re: For the sake of argument: is FhG mp3enc superior to LAME?

Reply #4
They are both very good:

http://listening-tests.hydrogenaud.io/sebastian/mp3-128-1/results.htm

Lame is still being updated though. I don't think fhg is.

Edit: fixed extra character on the end of the link
Any newer tests than 2008? Perhaps comparing the latest LAME with FhG?

http://listening-tests.hydrogenaud.io is unfortunately an Apache standard page, doesn't link to any available listening test results.

P.S. Helix is another interesting MP3 encoder. It has an excellent quality and it's extremely fast.
https://hydrogenaud.io/index.php/topic,113324.msg933009.html
Whelp, even if just for testing, Helix would be a nice encoder... if I could find the sources.
it seems the page (helixcommunity.org) is a friggin nightmare to navigate. It doesn't mention the MP3 encoder anywhere on that page, so I'm not even sure if it's hosted there.

Having said that, I find it kinda staggering, that LAME isn't managed using git (whether on github or a private git server).

Re: For the sake of argument: is FhG mp3enc superior to LAME?

Reply #5
They are both very good:

http://listening-tests.hydrogenaud.io/sebastian/mp3-128-1/results.htm

Lame is still being updated though. I don't think fhg is.

Edit: fixed extra character on the end of the link
Any newer tests than 2008? Perhaps comparing the latest LAME with FhG?

I doubt it.  Fhg hasn't been updated in 10 or 15 years at this point.  Its not relevent and hasn't been for a very long time.

Having said that, I find it kinda staggering, that LAME isn't managed using git (whether on github or a private git server).

Anyone still using CVS is never going to upgrade to git. 


Re: For the sake of argument: is FhG mp3enc superior to LAME?

Reply #7
polemon, all I can find for Helix encoder  it's windows binaries http://www.rarewares.org/files/mp3/helix_mp3enc_CVS.zip
But if You are not windows user ....
All I could find, are the files on Rarewares, too. Doesn't seem like they sources are available anywhere.

Having said that, Helix was packaged in 2005, almost 12 years ago. I doubt it has seen a lot of development in the mean time.

Re: For the sake of argument: is FhG mp3enc superior to LAME?

Reply #8
I guess I got the sources (don't remember where I got them from).
I have loaded them up to my webspace where they will reside for a limited time.
Everybody interested can download them from helix_mp3enc_CVS.zip.
Oops, seems like it's the file IgorC linked to. It contains source code.
lame3995o -Q1.7 --lowpass 17

Re: For the sake of argument: is FhG mp3enc superior to LAME?

Reply #9
Yeah, I feel like an idiot right now, the sources are contained inside the package, along with the compiled program...
I'll see if I can make it compile under Linux, now that I do have the sources.

Shouldn't be that complicated, as it doesn't seem to rely on any graphical stuff, etc.

Re: For the sake of argument: is FhG mp3enc superior to LAME?

Reply #10
I'm asking this, because that discussion (with one party arguing for FhG, the other for LAME) sprung up not too long ago on IRC.

I was quite surprised, because since Fraunhofer stopped marketing their mp3enc and fastenc encoders quite a while ago, I've been under the impression, that LAME has become superior to it during the mid to late 2000's. I was arguing citing various ABX tests, performed by groups of people, but that was dismissed as anecdotal, which is not entirely incorrect.

As LAME saw (and still sees) development, I remember seeing improvements over time from version to version. Also, as I understand it, LAME performs better with slightly lower target bitrates than FhG. Granted, MP3 doesn't perform well at all in general with lower target bitrates, no matter the encoding variant, whether it's constant bitrate, variable, etc. However, my understanding is, that the quality comparison very much depends on the target bitrate that's looked at, etc.

I have to admit, that I don't have much experience with the FhG encoders, though. I don't remember ever using them for any stretch of time. Hence, I couldn't argue for LAME, simply because I don't have enough 1st encounter experience with it.
I remember mp3enc being bundled with Winamp. Since I'm not a Windows user, I can't really confirm for how long that has been the case, and if it's still the case, even.

I specifically *don't* mean MP3pro, btw.

At the same time, they mentioned, that since the last patent held by Technicolor has expired, FhG has released the patent into public domain, is that true? I just know, that since the patent expired, FhG doesn't collect licensing fees for it, etc. but as it being released into public domain actively by FhG seems confusing to me. Also, I couldn't find any information about that claim anywhere.

I can kinda see, how it is next to impossible to argue for one or the other, but I'd at least like some commentary on those claims.

From my memory, Mp3Enc is the same creature as the acm shipped with WMP10. Fhgmp3 is really only competitive at bitrates of 112 and 128. Lame is better at medium to medium high and Blade is even better but only at 256 and 320.

Re: For the sake of argument: is FhG mp3enc superior to LAME?

Reply #11
'MP3Pro' from Thomson is equiv to Fhg's Fast Encoder (normal mp3 mode).

Re: For the sake of argument: is FhG mp3enc superior to LAME?

Reply #12
Quote
From my memory, Mp3Enc is the same creature as the acm shipped with WMP10. Fhgmp3 is really only competitive at bitrates of 112 and 128. Lame is better at medium to medium high and Blade is even better but only at 256 and 320.
I doubt Blade sounds better than LAME @ 256Kbps. If I remember correctly there was a listening test where Blade didn't do so well against LAME @ 224Kbps