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: Latest LAME? (Read 16690 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Latest LAME?

Just wanted to check that 3.100a2 is still the latest version (production or test) of LAME available, and that halb27's variant currently stands at 3.100a2-m.  I'm encoding some new MP3s after over a year's hiatus, and as far as I can tell - through Google searches, Sourceforge and browsing this forum - nothing has changed with LAME.  Thanks in advance.

BTW I'd also be interested in thoughts on stickying halb27's variant on this board.  I'm not aware of anywhere else that it's available.  And 3.100a2 itself is not available on Sourceforge or other common locales for LAME.

Latest LAME?

Reply #2 –
Just curious. How's development going on the latest version? Haven't heard anything for quite a while. Are we anywhere close to getting a beta or a final?

Latest LAME?

Reply #4 –
Quote from: lock67ca on
Just curious. How's development going on the latest version? Haven't heard anything for quite a while. Are we anywhere close to getting a beta or a final?

Look at http://lame.cvs.sourceforge.net/viewvc/lam...me/?sortby=date

Here's the recent changes of libmp3lame (the encoding library):

Quote
Thu Jul 19 20:57:52 2012 UTC (2 years, 4 months ago) by robert
Improved detection of MPEG audio data in RIFF WAVE files.
see tracker item 3545112: Invalid sampling detection


Sat Oct 6 11:44:56 2012 UTC (2 years, 2 months ago) by robert
+ new command line option "--gain x",  more convenient to adjust replay gain, just use whatever decibel value LAME reported before.
+ out-dir option mentioned in long help text


Tue Oct 9 21:37:13 2012 UTC (2 years, 2 months ago) by robert
trying to improve strong tonal samples (lead-voice.wav)


Tue Dec 17 11:24:56 2013 UTC (11 months, 3 weeks ago) by robert
applying submitted patches
#64: fixes undefined references when compiling with mingw32 (send by Patrick)
#59: changes static_assert into a more portable form (send by Doug Cook)
#55: fixes issues with wave files >2GB (send by lvqcl)
plus some minor changes to reduce compiler warnings with GCC v4.7.1


Tue Jan 28 14:48:45 2014 UTC (10 months, 2 weeks ago) by robert
some minor changes:
- do not disable replaygain calc and find peak sample, when not writing LAME info header (patch tracker item #66)
- remove deprecated linker option from Makefile.MSVC (bug tracker item #442)


So maybe we'll see a beta in year 2020? I don't expect it to happen sooner. But maybe I'm too pessimistic.

Latest LAME?

Reply #5 –
I updated my builds also. Now i use GCC 4.9.2 which seems to speed up things a little more.
All builds are 'generic' ones for SSE2 or AVX capable processors.
The generic builds do not really benefit from AVX. This would make more sense on specific
compiles tuned for one CPU architecture.

Here's the bundle with Gogo 3.13a (just for fun...), Lame 3.99.5 and the latest cvs Lame 3.100a2.

Latest LAME?

Reply #6 –
I also wonder about Lame 3.100 development,. As I consider 3.100a2 to be excellent I'd welcome a beta version a lot.
lame3995o -Q1.7 --lowpass 17

Latest LAME?

Reply #7 –
Hmm...based on the lack of further response, I'm guessing that 3.100a2, halb27's m variant, is the latest version available.  Considering anything above -V3 on 3.99.5 or newer already sounds transparent to me, I probably shouldn't be too worried about new versions anyway.  Still, it makes sense to always use the latest and greatest available

Halb27, I'd personally recommend that we sticky your final 3.99.5 and 3.100a2 variants in this thread, or otherwise make them available for download in a prominent location on this site.  It would be easy for them, and the accompanying documentation, to get lost in this forum.

Latest LAME?

Reply #8 –
Quote from: halb27 on
I also wonder about Lame 3.100 development,. As I consider 3.100a2 to be excellent I'd welcome a beta version a lot.

Sorry, I'd not found spare time in the past year. 100a2 has some issues at lower settings, I wanted to fix before a beta.

Latest LAME?

Reply #9 –
Quote from: robert on
Quote from: halb27 on
I also wonder about Lame 3.100 development,. As I consider 3.100a2 to be excellent I'd welcome a beta version a lot.

Sorry, I'd not found spare time in the past year. 100a2 has some issues at lower settings, I wanted to fix before a beta.


No need to apologize!  We wouldn't have LAME without the hard work you and others have put in on this.

Latest LAME?

Reply #10 –
Quote from: robert on
Sorry, I'd not found spare time in the past year. 100a2 has some issues at lower settings, I wanted to fix before a beta.

Thanks. Take your time to do what's necessary.
lame3995o -Q1.7 --lowpass 17

Latest LAME?

Reply #11 –
Quote from: BFG on
Halb27, I'd personally recommend that we sticky your final 3.99.5 and 3.100a2 variants in this thread, or otherwise make them available for download in a prominent location on this site.  It would be easy for them, and the accompanying documentation, to get lost in this forum.

Glad to see that you like to use my variant,  but - sorry - I don't want to mix up my variant with official Lame.
What ideas do you have about making my variant available here at a prominent location? Perhaps I can help in this respect.
lame3995o -Q1.7 --lowpass 17

Latest LAME?

Reply #12 –
We're quite happy to host them at Rarewares if that suits and you would like to provide a suitable description, and the binaries.

Latest LAME?

Reply #13 –
Quote from: john33 on
We're quite happy to host them at Rarewares if that suits and you would like to provide a suitable description, and the binaries.

You're welcome to host them at Rarewares: Lame3995m and Lame3100m.
lame3995o -Q1.7 --lowpass 17

Latest LAME?

Reply #14 –
These are now available at Rarewares, lame-bundles page.

Latest LAME?

Reply #15 –
Thank you very much.
this is a long time I didn't use MP3 encoding (I remember that years ago, only with LAME at relative high biterate, i could make compressed encoding of my lossless content without scarifying too much on perceived quality for the car deck unit).
Today I use mobile device with android and opus codec at average 130 Kbps for a decent quality on the go. When I will have some time to loose I will check what is the best compromise for me with latest built of LAME despite the fact that compatibility with MP3 decoding now is not an issue.

Latest LAME?

Reply #16 –
Quote from: robert on
100a2 has some issues at lower settings, I wanted to fix before a beta.

Awesome!

16 years of development and still counting.

Looking to results of the last public test it seems like LAME struggles on tonal and quiet passages. Though an overall performance and consistency of quality are high.
I remember Helix MP3 encoder did a bit better than LAME on tonality but worse than it on transients (at least in my test).  Probably alpha 3.100 should do better on tonality now due to improvements for  'lead voice' sample.

Also today as a lot of people encode at high bitrate (-V2, -V0 ..... LINK) it could make sense to spend most of time tuning those bitrates.

Looking forward a new fresh LAME and I'm sure guys here (me too) will help to test it.

Latest LAME?

Reply #17 –
Can I make CBR Mp3 with this version of lame?? what happens with lame_enc ? I could not find any lame_enc in the lame  bundle..

Latest LAME?

Reply #18 –
These particular version are the command line executable only. They do not include the dll as it is very unlikely the frontend program that you would use it with would support the additional option.

Latest LAME?

Reply #19 –
Quote from: IgorC on
Looking to results of the last public test it seems like LAME struggles on tonal and quiet passages. Though an overall performance and consistency of quality are high.
I remember Helix MP3 encoder did a bit better than LAME on tonality but worse than it on transients (at least in my test).  Probably alpha 3.100 should do better on tonality now due to improvements for  'lead voice' sample.

Also today as a lot of people encode at high bitrate (-V2, -V0 ..... LINK) it could make sense to spend most of time tuning those bitrates.


For what it's worth, -V0 -Y (with lowpass disabled) typically sounds better to me than -V0 by itself in quiet passages.  At least, it does when using halb27's 3.100a2 variant.  I have no idea why, and considering that most of the time -V3 or higher sounds transparent to me, it could be my imagination.  But it might be worth testing formally.




Latest LAME?

Reply #20 –
Quote from: BFG on
Quote from: IgorC on
Looking to results of the last public test it seems like LAME struggles on tonal and quiet passages. Though an overall performance and consistency of quality are high.
I remember Helix MP3 encoder did a bit better than LAME on tonality but worse than it on transients (at least in my test).  Probably alpha 3.100 should do better on tonality now due to improvements for  'lead voice' sample.

Also today as a lot of people encode at high bitrate (-V2, -V0 ..... LINK) it could make sense to spend most of time tuning those bitrates.


For what it's worth, -V0 -Y (with lowpass disabled) typically sounds better to me than -V0 by itself in quiet passages.  At least, it does when using halb27's 3.100a2 variant.  I have no idea why, and considering that most of the time -V3 or higher sounds transparent to me, it could be my imagination.  But it might be worth testing formally.


agree, new tonality so far so good, lame 3.100a2 tune on 320kbps bitrate
-b 320 -Y -q0
-V0 -Y

http://www.mediafire.com/download/kxxti5bz...15_win32_mod.7z

but some freq not perfect, bug maybe
nothing special, GOGO-no-coda ~256kbps VBR + mp3packer