I’m currently preparing my library. I am ripping Usher’s Confessions Cd in itunes mp3 192 VBR highest quality vs Apple lossless vs Lame mp3 3.96 using Exact audio copy. AAC is not included since I want forward compatibility with anything I Might get in the future.
Now obviously Apple Lossless will be the best of the bunch. But I was looking for any objective (or subjective!) opinions on these codecs and how they compare. Itunes mp3 has conveinence on it’s side. But LAME (supposedly) has quality.
I’m currently conducting my own tests, but i’d like to see what others think.
I can’t answer your question, except to say that it depends on which is more important, file size or quality or compatibility.
I must comment that whomever decided to name the format “Lame” (even if it stands for something meaningful) should be soundly horsewhipped and banished from those types of decisions.
Well I just ripped Usher’s Confessions CD in all three formats. I can’t tell a damn difference between the three. Thus Itunes wins for convienence. (file sizes are slightly larger then Lame) I’d still like to know which is technically superior though.
Another question. Whats the difference between Normal Stereo and Joint-Stereo? I did a little research but i’m not sure. From what I gathered Normal Stereo will produce two channels of One left and one right. While joint-stereo will produce a single monotrack and the 2nd channel contains information on the differences between the left and right.
Normal stereo will encode a 160 bitrate average Mp3 at 80 bitrate average for each channel.
Whereas Joint stereo can have a >80 bitrate monotrack(and thus highly fidelity) and since the differences channel is less information dense it be encoded at a lower average bitrate while maintaining quality.
Is this right?
If so, why does anyone use normal stereo at all?
or perhaps is it that joint stereo algorhythms aren’t perfect and theres quality loss over normal stereo?
What happens when one encodes in VBR? Shouldn’t Joint stereo simply be able to scale up the differences track to anything it wants? Saving space over normal stereo and maintaing quality?
Variable BitRate encoding apportions bits according to the complexity of the music. During less complex and/or silent portions of the music, fewer bits are required to convey the information so a smaller bitrate is used. Conversely, complex portions are accorded more bits. Over the course of a piece of music, bitrates will fluctuate pretty much continuously. Overall, the end result of VBR encoding is a better sounding file at a smaller filesize. DVDs, for example, use both audio and video VBR encoding to fit huge amounts of information onto a single disc.
As for ipod vs LAME (LAME Ain’t an Mp3 Encoder), I’ve never used ipod encoding but the general consensus on the web is that LAME is the best currently available MP3 codec. I use the LAME “–alt-preset standard” setting for music on my flash-based player. It typically yields an average bitrate of 160kbps which I find more than adequate for portable usage.
If you want to do further reading, I recommend Hydrogenaudio for more information about digital audio than you can ever hope to digest in a single lifetime.
Exactly, especially considering the “LAME” in the acronym LAME ain’t an MP3 Encoder has no meaning at all. It could have just as easily been something DAME aint an MP3 Encoder or FAME ain’t an MP3 Encoder or GAME ain’t an MP3 Encoder or MAME ain’t an MP3 Encoder or NAME ain’t an MP3 Encoder or SAME ain’t an MP3 Encoder or TAME ain’t an MP3 Encoder or VAME ain’t an MP3 Encoder. I hate it when a letter in an acronym stands for the acronym. It’s dumb.
“Hey what does the “L” in “LAME” stand for?”
“LAME.”
“Okay then what does the “L” in THAT LAME stand for?”
“LAME.”
“Alright then. What does THAT “L” stand for?”
“LAME.”
It never ends.
Alright, sorry 'bout the hijack.
Still the best encoder though, even though it says clearly that it’s not. If you have a sound editing program like soundforge you can look at the sound file that is created with the different encoders. Then you can see just how superior LAME is.
Files with huge differences between the left and right channels will, presumably, sound worse with joint stereo. Most MP3 files aren’t like that, so I’m not sure why joint stereo isn’t the default either.
Here’s a good rundown of the differences between joint stereo and regular stereo.
MP3 encoders can use separate left and right channels for some parts of the song and “mid/side” stereo for other parts. When you select joint stereo, the encoder is supposed to analyze the music and use M/S for the parts where it’ll be an improvement, and go back to L/R for the parts where it won’t (e.g. when the left and right channels are out of phase). There’s also a third kind of stereo encoding, “intensity stereo”, which throws away some stereo information and should only be used at low bitrates.
Looks like the only reason not to select joint stereo is if you have a bad encoder.
Thanks for that first link; excellent read; I will never stray away from Joint Stereo again! On the other hand, as I encode everything with EAC/LAME, that problem should never arise, because (a) LAME defaults to Joint Stereo (but analyzes the signal to see whether it’s worth it or not) and (b) I encode in VBR 192 kbit, so the bandwidth saving issue is really a non-issue.