First of all hello to the readers of this topic!

As I am a newbee, plz let me know if I by mistake offend someone by posting such question, but I actually have a doubt regarding safe joint stereo of Lame.

I recently came to a site (using google search) named mp3-tech.org,
I found an information regarding limitations of Jt. St.

Joint stereo limitations:
Mp3 can not switch joint stereo mode for specifics scalefactor bands. If joint stereo is used, it has to be used for all the bands. This is rather inoptimal, and is limiting the use of joint stereo. As an example, imagine the following situation:
The lower frequencies are featuring an instrument playing on the far left, and frequencies around 1500Hz are featuring a singer in the middle of the stage.
In such a situation, it is not possible to use joint stereo with Mp3 because of the lower frequencies part which is too different between both channels.

So, by any chance lame's using the technique mentioned?
By any chance that's what's known as Safe Jt. St.?

Reference:
www.mp3-tech.org/content/?Mp3%20Limitations]http://www.mp3-tech.org/content/?Mp3%20Limitations
(I don't know whether to post external links, hence I am not using http:. Also help me on that, plz)

LAME with presets (that is if you don't try and tweak it with options) will use the most sensible stereo mode.

I think its fine. High mathematics is a weak spot of mine, so welcome

This quote only refers to a limitation of the formats design -not being able to toggle the notation used for the scalefactors - this would introduce a little of its own inefficiency as an extra bit to toggle each band would have to be allocated, but ~maybe it would be merited.

I havent visited the whole page, but its unlikely to fault Lame's JS (saying there is a limitation in the format, is different from saying its used poorly by all encoders)

Theres a button to add links above the composition frame, marked "http://"

regards'

I imagine it could make quite some difference in some cases - eg. if there's a centre panned bass and two guitars panned left and right playing different things. Does anyone know if it's implemented in any other more modern codecs?

thanx for ur warm response,
and do not worry about math
for that, i am always at ur service
of course, about LAME, i miht be needing ur mathematics <i mean knowledge >

and i am just curious about the safe jt. st.,
It's been said that too much switching betn M/S & St. will worsen the quality.
so, in the samples in which there are major variations (sound going from left to right, remaining same in bith channel & then shiftin to left channel) how would this be Safe?

I can't see it being an issue, unless it was a naff decoder being used that couldn't switch very well. M/S and L/R are still representing the same signal, however fast they're switched between.

Its a bit like pixels on a display. Theres about 100 blocks of sound per second, each written as either L/R or M/S - in a variety of accuracy and sizes. So aprox every 100th of a second there are notation changes more significant than changing from L/R or M/S
Changes in sound are resolved through the individual blocks of sound but not so much within them - if at all (?)

Theres alot recent talk about js mode in a nearby thread (and more and more) in this forum with some input from more experienced. Potential problems have been explored here...

I would not be an issue if the encoding was perfectly transparent, but most of the time it is not.
The fact is that often the changes in artifacts are more annoying than the artifacts themselves.


No, this is a design limitation of the mp3 format, and Lame can't do anything against it.
note: this limitation is corrected in AAC.

Thanx Gabriel. U r the owner of the site I visited, isn't it? Thanx for ur response.

But, if this is limitation of mp3, if completely transp. encoding is not possible, then why the latest version of lame is using switching betn LR & MS?
Is it just experimental or is it really improving quality over the simple stereo mode?

M/S is still a (fairly large) gain if the channels are similar.

If they are not, it's a loss.

So switching is a win. The switching possibilities in mp3 are just not optimal.

But not switching ("simple stereo") would generally be a large loss, which is why JS is so heavily *encouraged*.

I always used to think that JS wasn't suitable for hi-bitrate mp3s. Where did that rumour come from, was it ever the case? After converting some JS mp3s of matrix quad vinyl to to 4 channel wav (using phase difference analysis to get complete separation), and reading this forum I am totally convinced otherwise.