I've notice something new on rarewares.
Just FYI,
I'm doing some encoding now, only for testing purposes of course
i can't find this on the rarewares. strange?
i can't find this on the rarewares. strange?
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=277937")
[a href="http://www.rarewares.org/mp3.html]http://www.rarewares.org/mp3.html[/url]
Second entry from the bottom.
Although I believe these things shouldn't even be announced here at the forums. Stupidity often ensues when it happens.
If it was to be announced, I guess John33 would have done it, and announced it at RW's index page...
sorry about that/ i never wouldn't looked towards the bottom, would've thought that it would be at the top with the other LAME versions.
why is it that testing for 3.97 and 4.0 are going on simultaneously?
i can't find this on the rarewares. strange?
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=277937")
[a href="http://www.rarewares.org/mp3.html]http://www.rarewares.org/mp3.html[/url]
Second entry from the bottom.
Although I believe these things shouldn't even be announced here at the forums. Stupidity often ensues when it happens.
If it was to be announced, I guess John33 would have done it, and announced it at RW's index page...
[a href="index.php?act=findpost&pid=277943"][{POST_SNAPBACK}][/a]
Okay, then lets have the following in one of the earliest comments in this thread:
[span style='font-size:14pt;line-height:100%']This version (4.0 alpha) is probably of worse quality than the latest stable version! [/span]
This compile is for developers and people who want to do listening tests only. When a version becomes ready for mainstream testing, its called "beta". The above version is especially probable to contain ugly quality-bugs because it is based on a completely new (untuned) encoder.
Some may recall that I posted alpha 12 a while back, and I've been updating this as Takehiro updated the CVS. As it's now gone to alpha 13, and therefore the old link is broken, I thought it better to hide it at the bottom of the mp3 page at Rarewares so that it was more readily available.
As Roberto has indicated, it wasn't intended to be announced, as such, it was just there for those who took the trouble to look.
As stated many times before, this is NOT yet ready for user testing. Takehiro will ask for that when he's ready. It is just there for people to see what progress he's making and the direction it's going in.
I tested this out last night and was suprised how quick it encoded an mp3. It was at least twice as fast as 3.90.3, possible more. Is it written with SSE CPU optimization code built it or it utilising a new compression engine?
The initial file output was much smaller but in comparison had a much lower bitrate for preset standard than 3.96.1. I'd like to know more info about this version so i know what to look out for. I'd like to start some hardcore testing
I tested 4.0 against 3.90.3 & 3.96.1 on only 3 problem samples but in all cases 4.0 sounded better, might be worth doing some testing
I tested this out last night and was suprised how quick it encoded an mp3. It was at least twice as fast as 3.90.3, possible more. Is it written with SSE CPU optimization code built it or it utilising a new compression engine?
The initial file output was much smaller but in comparison had a much lower bitrate for preset standard than 3.96.1. I'd like to know more info about this version so i know what to look out for. I'd like to start some hardcore testing
[a href="index.php?act=findpost&pid=278195"][{POST_SNAPBACK}][/a]
LAME 3.97 is already twice as fast as 3.90 in my computer.
Might be useful in doing some testing, but I think we should concentrate on 3.97 until Takehiro gives us the "go!" sign.
Might be useful in doing some testing, but I think we should concentrate on 3.97 until Takehiro gives us the "go!" sign.
[a href="index.php?act=findpost&pid=278201"][{POST_SNAPBACK}][/a]
Right. I remember Takehiro was mildly annoyed last time people jumped the gun and started madly speculating about lame 4.
I hope people without any clue about LAME development won't go around announcing these pre-testing versions anymore.
Might be useful in doing some testing, but I think we should concentrate on 3.97 until Takehiro gives us the "go!" sign.
[a href="index.php?act=findpost&pid=278201"][{POST_SNAPBACK}][/a]
Right. I remember Takehiro was mildly annoyed last time people jumped the gun and started madly speculating about lame 4.
I hope people without any clue about LAME development won't go around announcing these pre-testing versions anymore.
[a href="index.php?act=findpost&pid=278217"][{POST_SNAPBACK}][/a]
there should be a password protection for the archive...something like "FOR TESTING ONLY" would be good
Thing is there seems to be no documentation or build notes/change logs surrounding this build. Which makes it quite hard to test and what to look our for.
For a start i dont even know who built it. And i was suprised to see it leap frog alpha 7 build and thread.
Thing is there seems to be no documentation or build notes/change logs surrounding this build. Which makes it quite hard to test and what to look our for. [a href="index.php?act=findpost&pid=278333"][{POST_SNAPBACK}][/a]
Quite simply, this version shouldn't be tested. At all. The developers aren't asking for it.
So, you shouldn't be looking out for anything.
Move on, people... leave LAME 4 alone until one of the devs asks for testing.
Move on, people... leave LAME 4 alone until one of the devs asks for testing.
[a href="index.php?act=findpost&pid=278336"][{POST_SNAPBACK}][/a]
Definatly seems like the best suggestion. Is this the result of an open source project where the CVS can be accessed by anyone? Who are the respected active developers of this board? I see Gabriel mentioned a few times and his web site is very interesting
Move on, people... leave LAME 4 alone until one of the devs asks for testing.
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=278336")
Definatly seems like the best suggestion. Is this the result of an open source project where the CVS can be accessed by anyone? Who are the respected active developers of this board? I see Gabriel mentioned a few times and his web site is very interesting
[a href="index.php?act=findpost&pid=278340"][{POST_SNAPBACK}][/a]
[a href="http://lame.sourceforge.net/developers.html]http://lame.sourceforge.net/developers.html[/url]
I can't believe the fuss this has generated!!! So, it's gone again!!
Move Along, citizens. There's Nothing to See Here!
Move Along, citizens. There's Nothing to See Here!
[a href="index.php?act=findpost&pid=278370"][{POST_SNAPBACK}][/a]
Too late
I can't believe the fuss this has generated!!! So, it's gone again!!
[a href="index.php?act=findpost&pid=278356"][{POST_SNAPBACK}][/a]
I can't believe the fuss this has generated!!! So, it's gone again!!
[a href="index.php?act=findpost&pid=278356"][{POST_SNAPBACK}][/a]
Looks like it's back up again with a suitable "FOR TESTING ONLY" warning.
I know that I'm always interested to take a quick look (and listen) to these alpha compiles and get a feel for how the lame4 development is proceeding. I'm sure that many others similarly like to have a curious look and do so without making any fuss about it.
So thanks to John33 for reinstating it.
Before there are any other comments, this is there for your own appreciation only. Neither Takehiro, nor any of the other lame-devs, is directly interested in any testing you may do at this time.
There is no Change Log, or any other relevant documentation. LAME 4.0 is virtually a total rewrite, so little if any of the work done on 3.9x will go forward.
So, have some fun with it, but don't even consider using it for any other purpose.
LAME 4.0 is virtually a total rewrite, so little if any of the work done on 3.9x will go forward.
"Rewrite" is right. The 4.0 alpha is blazingly fast even compared with the 3.97 alphas, at least on my pc. Nice.
Thanks john33 for putting it up again. these alphas are indeed a lot of fun to play with.
At first its amazingly fast but then the fun ends.
LAME 4.0 is virtually a total rewrite, so little if any of the work done on 3.9x will go forward.
If thats the case, then i wonder: Looking at the promising test-results of the upcoming 3.97, wouldn't it make sense - after 3.97 got finished - to say "okay, thats it - the 3.9x series reached its limits - now lets focus all efforts on lame4" ?
I mean, if 3.97 turns out as good as it seems, then we would have a tuned, stable and good 3.97 - a perfect point to abandon the 3.9x series (except of possible bugfixes) and focus all energy on lame4 because all work on the 3.9x series cannot be "ported" to the future of lame(4).
I'm kinda curious on whats the strategy of the lame-devs on this matter.
- Lyx
LAME 4.0 is virtually a total rewrite, so little if any of the work done on 3.9x will go forward.
If thats the case, then i wonder: Looking at the promising test-results of the upcoming 3.97, wouldn't it make sense - after 3.97 got finished - to say "okay, thats it - the 3.9x series reached its limits - now lets focus all efforts on lame4" ?
I mean, if 3.97 turns out as good as it seems, then we would have a tuned, stable and good 3.97 - a perfect point to abandon the 3.9x series (except of possible bugfixes) and focus all energy on lame4 because all work on the 3.9x series cannot be "ported" to the future of lame(4).
I'm kinda curious on whats the strategy of the lame-devs on this matter.
- Lyx
[{POST_SNAPBACK}][/a]
(http://index.php?act=findpost&pid=282717")
Lyx - see this [a href="http://www.hydrogenaudio.org/forums/index.php?showtopic=29575&view=findpost&p=263440]thread[/url]. I asked the very same thing.
i tried the lame 4 alpha 14, and it's really fast , amazing, now the encding it's x2 fast.
If with lame 3.97 i spend 30'' with 4 alpha 13 it's 17'' ..woowww...
When a fast encoding for ogg vorbis? I have the optimized sse2, but in comparison with this lame "looks slow" (and it's).
I decided to use Lame4.0a14 to re-rip and encode Glenn Gould Goldberg Variations 1981 recording. Not entirely happy with the 3.97b2 encode using -V 2 --vbr-new on my ipod. The 4.0 encode using -V 0 sounds terrific. I've encoded a few more classical piano CDs and the results are excellent. Congrats to the developers!
I decided to use Lame4.0a14 to re-rip and encode Glenn Gould Goldberg Variations 1981 recording. Not entirely happy with the 3.97b2 encode using -V 2 --vbr-new on my ipod. The 4.0 encode using -V 0 sounds terrific. I've encoded a few more classical piano CDs and the results are excellent. Congrats to the developers!
Shouldn't you try -V0 on 3.97b2 before jumping to 4.0a14?
I decided to use Lame4.0a14 to re-rip and encode Glenn Gould Goldberg Variations 1981 recording. Not entirely happy with the 3.97b2 encode using -V 2 --vbr-new on my ipod. The 4.0 encode using -V 0 sounds terrific. I've encoded a few more classical piano CDs and the results are excellent. Congrats to the developers!
Shouldn't you try -V0 on 3.97b2 before jumping to 4.0a14?
Sorry, should have said this. Yes, I did and was happy with the result. But thought that seeing I had the .wav image of the CD (one of my favourite classical discs) ripped anyway, I should give 4.0 a run. I've since encoded Gerard Willems' Beethoven Sonatas with similar good results, again replacing 3.97b2 V2 and V0 versions. I'll be trying a couple of Wynton Marsalis trumpet discs tonight to see how 4.0 handles it.
I should also explain that I've been a lurker for the last few months learning about EAC/Lame from the great comments/tutorials/etc on this board. I've since re-ripped/encoded over 550 CDs to replace older 128bps MusicMatch MP3's to go onto my new 60G ipod. This included setting up a variety of AAC vs MP3 tests using iTunes. But I've decided to stick with MP3 via Lame. I've also experimented with FLAC to see if it makes sense to create an archive of my "cannot replace" CDs.
So thanks to all the posters who have shared their experiences with us lurkers.
Does anyone have en idea of when lame 4.0 MIGHT be available in a stable release? I suppose this wouldn't be in just a few months, but I haven't followed any other lame development (and recently noticed i encoded my entire cd-collection with 3.90)
Not entirely happy with the 3.97b2 encode using -V 2 --vbr-new on my ipod. The 4.0 encode using -V 0 sounds terrific. I've encoded a few more classical piano CDs and the results are excellent.
I think you are likely to get flamed here if you don't provide some ABX-results that show that 4.0 sounds better (to you) than 3.97 b2. You are violating TOS #8.
Not entirely happy with the 3.97b2 encode using -V 2 --vbr-new on my ipod. The 4.0 encode using -V 0 sounds terrific. I've encoded a few more classical piano CDs and the results are excellent.
I think you are likely to get flamed here if you don't provide some ABX-results that show that 4.0 sounds better (to you) than 3.97 b2. You are violating TOS #8.
Well, I'm not sure that jumping up two notches on the quality settings and claiming that it sounds better constitutes a TOS #8 violation. It's common sense! Even when we are talking about a beta vs an alpha of the same codec.
But it has already been said: quality comments on LAME 4.0 alphas are not welcome until public tests have begun.
And besides, this guy just said that he is not "entirely happy" with V2 on the recommended LAME compile. I guess that says a lot (and that does constitute a TOS #8 violation)
It would be nice to have more details. You said that 3.97 --V2 isn't that good. That's interesting: developers may be interested to know what the problem is and to correct it (-> submitting a sample illustrating the issue may be useful).
You said that 4.00b14 -V0 is terrific; what about 3.97b2 -V0 too?
Not entirely happy with the 3.97b2 encode using -V 2 --vbr-new on my ipod. The 4.0 encode using -V 0 sounds terrific. I've encoded a few more classical piano CDs and the results are excellent.
I think you are likely to get flamed here if you don't provide some ABX-results that show that 4.0 sounds better (to you) than 3.97 b2. You are violating TOS #8.
Man, I want to avoid those TOS #8 police. They sound mean!
Without wanting to get flamed off the board on my first post (and I'm happy to crawl back to my lurking hole again), I was searching for any threads about Lame4.0, found this one, and decided to post a thanks to the developers given my positive experience. I wouldn't want anyone to read my post as any sort of scientific comment on one rev vs another.
'nuff said, back to ripping...
It's not a matter of being
mean, but of proving statements
you made:
Not entirely happy with the 3.97b2 encode using -V 2 --vbr-new on my ipod.
What exactly weren't you happy about? Is that ABXable?
It's not a matter of being mean, but of proving statements you made:
Not entirely happy with the 3.97b2 encode using -V 2 --vbr-new on my ipod.
What exactly weren't you happy about? Is that ABXable?
Will everyone please lighten up. The man was merely expressing his opinion. He's not preaching to convert anyone to some cult religion. It's his opinion, and he doesn't have to prove it. When discussing 'perceptual' encoders, there is no absolute truth. Microsoft has published ABX listening tests which show Windows Media Encoder to superior to everything. Call me skeptical. "Can't we all just get along...."
Will everyone please lighten up. The man was merely expressing his opinion. He's not preaching to convert anyone to some cult religion. It's his opinion, and he doesn't have to prove it. When discussing 'perceptual' encoders, there is no absolute truth. Microsoft has published ABX listening tests which show Windows Media Encoder to superior to everything. Call me skeptical. "Can't we all just get along...."
You're on the wrong forum, buddy. ABX tests are infallible when implemented correctly. I'm not going to argue it though here... if you don't like the rules, there are other places on the internet to hang out. Around here, he DOES have to prove his opinion.
Will everyone please lighten up. The man was merely expressing his opinion. He's not preaching to convert anyone to some cult religion. It's his opinion, and he doesn't have to prove it. When discussing 'perceptual' encoders, there is no absolute truth. Microsoft has published ABX listening tests which show Windows Media Encoder to superior to everything. Call me skeptical. "Can't we all just get along...."
You're on the wrong forum, buddy. ABX tests are infallible....
Infallible?? Well, if you say so. I stand corrected.
Will everyone please lighten up. The man was merely expressing his opinion. He's not preaching to convert anyone to some cult religion. It's his opinion, and he doesn't have to prove it. When discussing 'perceptual' encoders, there is no absolute truth. Microsoft has published ABX listening tests which show Windows Media Encoder to superior to everything. Call me skeptical. "Can't we all just get along...."
You're on the wrong forum, buddy. ABX tests are infallible....
Infallible?? Well, if you say so. I stand corrected.
They are, and they are reproducible. And they are the most scientific way of proving such a statement.
One is entitled to his opinion, but once he decides to share it, it must be backed up by evidence.