Is there any difference between stereo & "joint stereo" on MP3's?

I noticed this yesterday when I was converting a WAV file to 320kbps. Is there a difference in sound or stereo separation?

No bad pot jokes please.

Thanks!

 

Joint Stereo is usually what's called "mid/side stereo". In normal stereo, the L and R channels are encoded separately. However, in joint stereo, the L and R channels are used to form an L+R and L-R channel to be encoded. The reasoning behind this is that if most of the audio information is in the centre of the stereo image, the L-R channel will be bare and thus very easy to encode.

Good MP3 encoders will automatically choose between mid/side stereo, intensity stereo and normal stereo depending on the track, whichever will yield the best results at the chosen bitrate. Joint stereo is better for stereo tracks where the audio is mainly focused in the centre of the stereo image, like a lot of chart music. Normal stereo will better suit something with heavy channel separation like the first few Beatles albums.

 

Yes.
Joint Stereo achieves a smaller file with no artifacts by only encoding the center and the difference between the left and right channels, instead of encoding the center twice.
It works great unless you plan on listening to the files through a system like Dolby PLIIx, which WILL result in artifacting.
Here's the difference between normal stereo and joint stereo:

 

I've used LAME for years and years. Anyone know if the standard -V 2 preset encodes in Joint Stereo or Stereo?

 

So this is basically more about making a smaller sized file rather than sound quality...at least that's how I'm reading it.

It's strange, because I converted the WAV file in iTunes under "custom" and it was in "joint stereo", so I left it there. When I checked the file after converting, it just said "stereo". I was a bit puzzled because the VBR files I get at Emusic are labelled "joint stereo" and I'm looking at both in iTunes. As long as the quality is the same, I really don't care though.

 

But t isn't !

 

I have always wondered what this meant, great thread and answer!

 

No. Joint stereo is only a method of storing information more efficiently. It has nothing to do with actual stereo sound.

I think forum member Brian W explained it best in a simple way:

 

Odd. The file size differences look to way too much.

 

Probably a mono track which has just dither differences in each channel.

 

Thanks for the info.

Here's a curveball, did you ever see an MP3 that is "dual stereo"? Anybody know what that is about?

 

Stereo encoded separately on each channel. Makes a larger file.

 

So all these labels are just basically variations on file sizes and have nothing to do with differences in sound quality or stereo seperation. Correct?

Is there a "label' you prefer for 320kbps files?

 

It does have to do with quality, Joint Stereo should almost always be used. It allows more bits to be available during encoding which is used to improve quality:

http://www.hydrogenaudio.org/forums/index.php?act=ST&f=16&t=4678

 

I think you might be mistaken about joint stereo. Read here: http://en.wikipedia.org/wiki/Joint_stereo

Quote: "The term joint stereo has become prominent as the Internet has allowed for the transfer of relatively low bit rate, acceptable-quality audio with modest Internet access speeds." This doesn't sound like high-quality to me.

 

And why do you suspect that joint stereo helped provide "low bit rate, acceptable-quality audio"? Because it allows the same quality as normal stereo with a lower bit rate! If you had two files with the same quality, one at 7 MiB and another at 4 MiB, which would you choose?

As said above, joint stereo isn't the best options for all tracks but is usually more efficient than normal/discrete stereo. Again, as said above, the encoder should automatically choose which is best for the current track depending on stereo separation.

 

The best-sounding one, but that's just me.

 

There's been a lot of misinfo about joint stereo on the web over the past decade. Lots of wrong info. It's no wonder people are confused.

Here's an old web page about joint stereo myths (it's captured here on the wayback machine).

The wikipedia article on joint stereo also touches a bit on basic info.

Much of the confusion is probably due to there being two types of joint stereo used in MP3. There is intensity stereo and mid/side stereo. Both got lumped in with being called joint stereo with MP3. The myths article and the wikipedia article explain what intensity stereo is. It's lossy. Encoding to intensity stereo loses information. Mid/side stereo is mathmatically lossless. You can encode left/right stereo to mid/side stereo and back to left/right stereo without losing any information. Lossless codecs can take advantage of mid/side stereo to improve compression rates for music that has a high correlation between the left and right channels. Since converting to mid/side is a lossless transformation it can be used in lossless codecs.

It's the intensity joint stereo in MP3 that gave joint stereo a bad rep.

Some of the old MP3 encoders didn't do intensity stereo encoding well. That was a decade ago. Modern MP3 encoders are better. Intensity encoding isn't used isn't used for higher quality mp3 encoding.

 

A major reason for doing lossy encoding is to get the best possible audio quality in a smaller file size at a lower bit rate. If you didn't care about file size or lower bit rates then might as well do lossless. So for lossy, the best option would be the smaller file assuming both have the same audio quality. Why do lossy if your goal is to make the file almost as big as a compressed lossless file?

 

There has been tons of mis-information about what joint stereo is, and tons of audio people tend to believe it because it sounds plausible to them. but, it just isn't so. J-Stereo does not degrade the sound, it actually can improve it by allowing more bits to be used for where they are needed. And, variable-bit + J-stereo is even better!

Several months ago, last December, I think, I posted a link to a great webpage that explained it all in great detail. I cannot find the link now.

 

Lots of music has a high correlation between the left and right channels. Watch a stereo correlation meter and you can see that lots of music is bouncing around with a high degree of correlation. A high degree of left/right correlation means that the mid channel in mid/side will have most of the music info. For music with a high correlation you can convert to mid/side and give the mid channel the most bits and the side channel fewer bits. That results in a more efficient way to encode the music.

For example, say you're encoding at a 200 kbps rate and the music has a high left/right stereo correlation. If you went with left/right stereo you'd have to give the left channel 100 kbps and the right channel 100 kbps. With mid/side you could give the mid channel 130 kbps and the side channel 70 kbps and end up with something that sounds better because the bits are used on the channel where they are more needed.

 

I don't believe so. There's a reason why JS is used with all the LAME presets that I'm aware of.

http://wiki.hydrogenaudio.org/index.php?title=Joint_stereo