Just wondering what the word was on 3.96? is better to use thatn lame 3.92 or are there still buggs?

go ahead and use it. I do. Give us a shout though if you find any critical problems. AFAIK there are none.

3.96.1 is much faster, new Vx VBR presets, low-mid bitrate quality is better, high bitrate quality should be no worse.

thanks guys

But AFAIK LAME 3.94+ don't uses Dibrom's switches & tunnigs... Where's the truth, this is why I'm still using recommended 3.90.3...

But that doesn't mean that 3.96.1 isn't heavily tuned. And although 3.96.1 has been tested to death like 3.90.3, it still has been tested (do a search).

I still use light bulbs made in 1881 because they use Thomas Edison's carbon filaments

I like this one a lot...

I know the answer is probably somewhere else... But is 3.96 DEFINITELY better on CBR (I'm thinking 160), or DEFINITELY worse, or is it in the same boat as VBR?

I use LAME 3.96.1, mainly with preset standard, without problems at all. (till now....)

Although I have seen many posts on this subject I have yet to see a definitive conclusion. My personal conclusion is that they are extremely comparable, with both performing very well (better than the competitors, and other versions of LAME). One performs marginally better than the other in some instances, and the other in other instances. 3.96.1 produces slightly smaller files (presumable mainly because it uses any bitrate at VBR, whereas 3.90.3 won't go below 128Kbps).

I guess this is much like the "which lossless codec is best?" question.

However, it does concern me that many people are sticking to 3.90.3, solely as there "hasn't been adequate testing" - or, more simply, as this site does not recommend it.

There will never be adequate testing if users are not prepared to try it. It is just as well the same approach was not taken with Lame 2.0.

I wonder what the current LAME developers have to do to turn people's heads.

NB: I used to play the drums in a punk band (for "punk" read "loud and unaccomplished"), so any chance of me performing an ABX test are out the window. However, I do like to believe that I am using the best tools to do a job.

Nope, 3.96 and later (or was it 3.95.1?) don't go below 128kbps. It was just 3.95 that tried to lower it to 96. The newer versions don't bloat high-frequency information as much, is all.

not only that ... but 3.96.1 is way faster than 3.90.x

Forgive me for two newbie questions.

I use EAC+LAME and I'd like to know:

1) is there any difference between alt-preset-standard and preset-standard?

2) if using 3.96.1, is there any difference between alt-preset-standard and Vx (or which V preset is the recommended equivalent of alt-preset-standard)?

Thanks for your patience.

Alessandro

1) No. --alt-preset standard was changed to --preset standard.

2) No. A table showing that relationship.

Thanks for the fast reply!

Just to make sure I understand correctly, you are saying that if I put in EAC command line options --alt-preset standard or --preset standard or -V2 (or should it be -V 2?) I obtain the same result. Is that right?

Thanks again.

Alessandro

If you are using 3.96.x, yes (eg with 3.90.x. the recommended forum version, this will not be the case). You can prove it for yourself, encode the file with each of the options then decode back to wav (all can be done within EAC). Then use EACs compare wav feature and it should show that all three files are the same.

Yeah, that's right!
If you have in EAC LAME 3.96.1 the switches --aps / --ps / -V2 should produce exactly the same MP3 files.
If you use LAME 3.90.3 & the switch -V2 it should be quite different from the --aps MP3 file.
AFAIK the switch --ps 3.90.3 don't know, but I don't know what would it do...
Maybe a simple 128 CBR MP3 or nothing = error post.