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: 3.94 alpha, Presets, Takehiro's Stable Branch, Etc... (Read 25847 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

3.94 alpha, Presets, Takehiro's Stable Branch, Etc...

Reply #75 –
What is the current status of the 3.94 branch? Any update?
There are mainly 2 points: "fixing" of current presets and introduction of new presets.


side note: source code of mpc encoder is already available

3.94 alpha, Presets, Takehiro's Stable Branch, Etc...

Reply #76 –
Gabriel, I thought Frank Klemm had the source code on his website but it was encrypted = non-really-available...

If it is really available is anyone going to replace lame's problematic psychoacoustic algorithm with mpc's excllent and highly tweaked psychoacoustic algorithm?----I think this should be a top priority for lame development.

What do you guys think?

3.94 alpha, Presets, Takehiro's Stable Branch, Etc...

Reply #77 –
Quote
Gabriel, I thought Frank Klemm had the source code on his website but it was encrypted = non-really-available...

If it is really available is anyone going to replace lame's problematic psychoacoustic algorithm with mpc's excllent and highly tweaked psychoacoustic algorithm?----I think this should be a top priority for lame development.

What do you guys think?

Firstabe the 1.93 sv8 alpha source is not encrypted.

Second, I'm not at all so sure that replacing the psychoacoustics and getting it to work with Lame so that it gives good results is so simple like you make it sound. I suspect it would need quite a lot of tweaking.

Third, imo it's a bit questional just to blatantly rip MPC's psychoacoustics and use it in Lame. The least what should be done if the psychoacoustics is blatantly ripped, is to ask permission from Frank and Andree first..
That would be at least more polite than just blatantly ripping and using MPC's psychoac like Lame's own..
Juha Laaksonheimo

3.94 alpha, Presets, Takehiro's Stable Branch, Etc...

Reply #78 –
Dear JohnV,

First, I might be wrong about the extent to which it is easy to swap the psycho-acoustic algorithms... to be honest I don't really know too much about the specifics in this case.  But I have talked with Andre and I found out that his psycho-acoustic algorithm was originally taken from the mpeg 1 layer 2 and then he tweaked it-- 60% of it was changed after all of his tweaks.  Since mp2 and mp3 are both layers of mpeg perhaps swaping these algorithms will be easier.  The only obstacle I see is if the algorithm makes uses of certain parameters not available in mp3....

The reason I think it would be great to make the swap is that I remember Takehiro and others commented that one reason why vbr in lame is not so good is because the psychoacoustic algorithm is not so great.  I think the algorithm in mpc must be much better.

Lastly, of course I am not advocating just to "blatantly rip MPC's psychoacoustics and use it in Lame." Of course, we should ask permission, I just wanted to see if anyone thought this was a good idea.  I am not hijacking the stuff as we speak...  Also, in a way by releasing the source code (open source) they seem to be saying that it is available for use (at least under certain conditions, no?)

3.94 alpha, Presets, Takehiro's Stable Branch, Etc...

Reply #79 –
psymodel used in transform codec is very different from that of subband codec. I think it should not be easy to Musepack psymodel into LAME.

I do not stop if someone try it, but at least I never try to do it.
May the source be with you! // Takehiro TOMINAGA

3.94 alpha, Presets, Takehiro's Stable Branch, Etc...

Reply #80 –
Quote
What is the current status of the 3.94 branch? Any update?
There are mainly 2 points: "fixing" of current presets and introduction of new presets.

Currently nothing is updated. I want to merge some Dibrom's improvement about RaveBass.wav, but there's still no source code.

The largest problem in the branch is an internal buffer handling error which R.A.F.'s was reported (http://www.hydrogenaudio.org/forums/index.php?act=ST&f=16&t=4217). IMHO. I got the .rar file last week, but because I have no extraction environment (for both of .rar and .ape), I haven't checked it yet.
May the source be with you! // Takehiro TOMINAGA

3.94 alpha, Presets, Takehiro's Stable Branch, Etc...

Reply #81 –
Quote
The largest problem in the branch is an internal buffer handling error which R.A.F.'s was reported (http://www.hydrogenaudio.org/forums/index.php?act=ST&f=16&t=4217). IMHO. I got the .rar file last week, but because I have no extraction environment (for both of .rar and .ape), I haven't checked it yet.

Could this error please be confirmed by someone else as well.
I have only a dial up connection at the moment, so downloading the files is a problem for me.

3.94 alpha, Presets, Takehiro's Stable Branch, Etc...

Reply #82 –
Quote
Quote
What is the current status of the 3.94 branch? Any update?
There are mainly 2 points: "fixing" of current presets and introduction of new presets.

Currently nothing is updated. I want to merge some Dibrom's improvement about RaveBass.wav, but there's still no source code.


Sorry for the confusion, but the last code of yours I tried fixed all the problems in this sample that mine did.  All I had changed before to help with this sample were the attack thresholds.  As I had said, this didn't work very well because it caused problems on other samples like Death2.

The main reasons I haven't released the code:

1.  The new presets were basically no good. (= no improvements to release)
2.  Your improvements made any of the more hackish fixes (which didn't work well) implemented beforehand unnecessary.
3.  I don't want people to use this non-optimal source (people will download, compile, and use it if I put it up, regardless of what I say about it), because as I said it's really not good.  I want to get something halfway decent going and then release that.

3.94 alpha, Presets, Takehiro's Stable Branch, Etc...

Reply #83 –
From a CVS sweep this morning, I notice that lame has mysteriously gone from version 3.93 alpha 3 to 3.94 alpha 1!!

Dibrom, this is likely to cause confusion I think, since I assume from a look at recent CVS changes that this has nothing to do with your efforts at all.

Edit: No mystery, my CVS sweep yesterday was too early, and this morning's was too late!! However, there is still the possible 3.94 version confusion issue!
 

3.94 alpha, Presets, Takehiro's Stable Branch, Etc...

Reply #84 –
This has been mentioned before, but it comes from time to time. Why not defaulting the presets ?
Here it comes again, from http://forum.cdfreaks.com/showthread.php?s...&threadid=57943 :

Quote
true, there are hacks to lame that make it better... and if those hacks are good... we'll see them in the next version...


And actually, I can understand the point of view. Lame is Lame, and if it is to be compared to others encoders, all encoders would have to be setup with their default settings. It wouldn't be fair if one encoder is set on JS, the other in ABR, another with scale 0.9, one with lowpass 16, and Lame with Alt-presets...

So why not release Lame 4.00, with the alt-presets defaulted ? The gap between Lame default and Lame --alt-presets would justify switching from 3 to 4.

Sorry if I am rambling an old idea. You know, I'm not active in MP3 discussions, my subject is rather audio CD, but as a new moderator in CD Freaks audio, I'll have to deal with Mp3.