My question about Joint stereo, it'll be quick Options

[agrizokh]

Hello everyone.

I am encoding my flac colection into mp3 with lame, and I have one matter still a bit unclear to me. I am going to encode with parameters -S -b 320 --noreplaygain, and I have a question about stereo mode. As i read in the wiki LAME only enforces joint stereo by default when the bitrate is lower than 128, otherwise it'll be full stereo? So, basically the first question - if i encode with parameters above, will i get full stereo?

And question two. I've read lots of speculation on stereo vs. joint stereo, and what I understood from the endless debates. Joint stereo is 99% as good as normal stereo in terms of quality, but it makes the file size considerably smaller. Joint stereo artifacts are almost unheard.

But there are some people who claim that joint stereo is actually better than normal stereo in terms of sound quality for some reasons. Well why do they say that? Can it be true?

Thanks in advance!

[pdq]

Joint stereo is the default at ALL bitrates.

Joint stereo results in either smaller files or better quality or both. Forced stereo is inferior.

The reasons for the above have been discussed many times. Do a search.

[Zoom]

There is an article on the Hydrogenaudio Wiki about joint stereo: http://wiki.hydrogenaudio.org/index.php?title=Joint_stereo

What LAME usually does is decide on a frame by frame basis which method to use. As you go from a lower quality VBR setting like V6 up to V0, LAME will use more LR frames rather than the MS frames. I would say to just use -b320 --noreplaygain. LAME has been tuned to perform the best on most music using presets.

[sld]

For an encoder that can handle Joint-Stereo maturely, the worst scenario is that Joint-Stereo offers no benefit over Simple Stereo. The best scenario is that the space savings offered by JS allows the encoder to increase the bitrate for frames that are harder to encode. There is definitely an improvement, although it may not be detected by the listener.

[agrizokh]

Joint stereo results in either smaller files or better quality or both. Forced stereo is inferior.

And the reason for that would be that in for normal stereo it can only encode in 160kbps per channel, and in joint stereo this restriction is lifted, so when the two channels are very similair, it can minimize the stereo effect and enhance the bitrate, something like that, right?

The reasons for the above have been discussed many times. Do a search.

I've read that huge bunch of topics and it's more discussion than dogmas for noobs like me, and you never know which opinion is correct after all. So I just wanted to clear it out for myself a bit, hope it's not bugging you too much. Maybe some topic like that should be put with clear descriptions which is BETTER (that's most important for the average man) with tech details backing it up.

Thanks for the replies, after this topic decided to go joint stereo. Gonna do some tests on some stereo-heavy dub sooner or later to absolutely clear my hesitations... =)

[odyssey]

It doesn't reduce stereo effect. It optimizes the encoding algorithm to take advantage of each channels similarity (to simplify it). When two channels are similar it can use the bits it would otherwise need to encode two channels, instead to improve audio quality.

[pdq]

And the reason for that would be that in for normal stereo it can only encode in 160kbps per channel, and in joint stereo this restriction is lifted, so when the two channels are very similair, it can minimize the stereo effect and enhance the bitrate, something like that, right?

Just to clarify, in what you refer to as "normal" stereo there are not 160 kbps for each channel, rather 320 kbps total between both channels. The encoder is still allowed to allocate more bits to one or the other channel as needed.

The mode in which each channel gets exactly 160 kbps does exists, but I'm not going to tell you the switch for it in case some fool reads this and decides to use it.

[agrizokh]

oddyssey, pdq, thanks! got it finally!