I've been lurking on this forum for a really long time. I've leared a fair amount about lossy audio codecs and audio rippers. But I have trouble sorting it all out or explaining what I know to people. I think this is because it's more of a faith that I believe that EAC is the best ripper and that I believe LAME is the best MP3 encoder and that I believe 3.90.3 is the best version of LAME to use. In my experiences, I have found the best, most consistent results with EAC and LAME, but I don't really know why that is. Basically it's like saying I'm Catholic without knowing anything about what is in the Bible.

I've tried using this forum's search function, but I was unable to find anything to satisfatorily answer my questions.

Basically I want to know:

1) Why is EAC the best audio ripper? I realize it has something to do with error correction and/or reporting, but why is EAC's error correction/reporting better than say C2 or any other audio grabber's error correction/reporting?

2) What makes LAME better than Xing, Blade, or the Fraunhofer encoders? I certainly believe that LAME is better from my experiences with it, so I guess what I'm really asking is what makes the Xing and Blade encoders so bad (and to a lesser extent the Fraunhofer encoder)?

3) Why is it exactly that joint stereo is better than full stereo? Is js only better in LAME, or is it better MP3-wide?

4) I gather that the 3.90.2 version was the most tested and hence the most trusted. 3.90.3 is basically 3.90.2 with the added -Z switch (right?). Why is it that no few people bothered testing 3.92 or 3.93.1 extensively, if at all? Is it because this is basically Dibrom's forum and Dibrom stopped working on LAME after 3.90.2? Has anyone that has done any testing between 3.90.2 and 3.92 or 3.90.3 and 3.92 (or any other versions like 3.93.1 and 3.95.1) made a determination as to which is better on the whole?

If anyone could answer these questions in detail it would be much appreciated.

http://doc.hydrogenaudio.org/wikis/hydrogenaudio/EAC


http://doc.hydrogenaudio.org/wikis/hydrogenaudio/MP3

You should spend some time reading these. Most everyone is worn out giving detailed answers to questions you can find here in the wiki/knowledge base.

peace

If these two statements were true, then you wouldnt be asking this question.

Well, they both are. I've found things that sort of answer my question, but don't to the extent I'm looking for.


I've read all that stuff before. Unfortunately I find it far too general and I'm not looking for answers to why MP3 is good or bad, but why LAME is good and the other encoders aren't.

Thanks for the effort though.

Be more specific in your questions or you'll get equally general answers.

Well, three of my questions aren't even touched by the information given so far, general or technical.

EAC is good because it features Secure ripping which enables it to rip the music on your CD with less glitches/pops/clicks.

LAME is good as a MP3 encoder because for the same bitrate it produces a waveform that sounds more like the original, i.e. it is more difficult to distinguish the resulting waveform from the original compared to another MP3 encoder.

Ok, I was able to find the forum FAQ that points me to all the right places. Actually, I had come across it before, but completely forgot about it. Sorry for this thread.

EAC and LAME r0X0rs!!!

Because it simply is better. The program is coded better.
C2 is not an audio grabber feature, but a device feature. If your CD-ROM device can report C2 errors 100% accurate, it is as good as the Secure Mode EAC uses.



LAME has a better psychoacousting model, LAME uses short blocks and long blocks...



Theoretically, it is "world-wide". Ogg Vorbis uses Joint Stereo, MusePack uses Joint Stereo...
You will find a lot of posts explaining why Joint Stereo is better the Stereo (not that it sounds better, but it saves bytes).

Blade is bad because its author misunderstood how to code a lossy codec correctly. Lossy codecs are about psychoacoustics (What the brain assumes the ear hears), and lossy codecs can only be properly tuned by subjecting their encodes to innumerable objective listening tests.
Blade was tuned by comparing spectral graphs, which only works (and to a certain extent) in very high bitrate mp3 encoding. That's why Blade somehow seems to work best (When Blade was still well used!) at CBR bitrates above 256.

Only Lame (and fraunhofer?) uses joint-stereo properly (no quality loss but lots of bits saved). So Lame sounds better than others codecs that don't (properly) use joint-stereo at the same bitrate.

That is not totally true. Blade still has the same psychoacoustics as the ISO demonstration code.
So the author did not misunderstood lossy codecs, he choosed to focus on speed optimizations and only this.

My error then, many apologies.. I had thought I read about this somewhere before.