Stereo, joint stereo or dual channel?

[nhope]

Can anyone explain to me what the difference is and which I should be using for VCD’s and DVD’s

[Sven Bent]

Joinst stereo = encode similarities on both rigjt and left and te diffrrence in the channesl. that way it doesn encod the same twicje for right and left


stereo = encode each channel for itself but can move bits usage to better fit the pictures. E.G right use 48kbits and left uses 96 kbits in a 128kbits file.


Dual channesl = both chennesl are enconde without using knowledge of the other channles. E.g in a 128kbits file each channesl woudle take 64kbits



joinstereso shoudl give the beeast quality/kbits becuas it doesn't encode the same data for both left and right. it only encoded it wans end therfore saving a lot bandwidth. whcih then can be used to encode details.

[Cokes]

Originally posted by Sven Bent
Joinst stereo = encode similarities on both rigjt and left and te diffrrence in the channesl. that way it doesn encod the same twicje for right and left

stereo = encode each channel for itself but can move bits usage to better fit the pictures. E.G right use 48kbits and left uses 96 kbits in a 128kbits file.

Dual channesl = both chennesl are enconde without using knowledge of the other channles. E.g in a 128kbits file each channesl woudle take 64kbits

joinstereso shoudl give the beeast quality/kbits becuas it doesn't encode the same data for both left and right. it only encoded it wans end therfore saving a lot bandwidth. whcih then can be used to encode details.

Not at all:

Joint stereo (JS) stream is over-set of three modes:
Stereo (S), Middle side Stereo (MS) and Intensity Stereo (IS) packets.

MS - when the left and the right channels are quite similar, then a middle (L+R) and a side (L-R) channels are encoded instead of left and right. This allows to reduce the final file size by using less bits for the side channel.

IS - some frequencies are then recorded as a monophonic signal followed by a few additional information in order to restore a minimum of spatialisation (Dolby Surround incompatible)

But what DVD,SVCD a VCD concerns (MPEG-1(2) Layer 2 audio): Layer 2 doesn't know anything about MS, it uses only IS, so if you encode your Dolby Surround donwmixed audio in Joint Stereo mode of any Layer 2 encoder - it destroys surround information.

Another situation is by Layer 3: For example lame (f.e. lame --r3mix) uses in joint stereo mode some advanced psychoacoustics models - it can swap bitrate, modes (S,MS,IS) on the basis of audio stream complexity (Dolby Surround compatible).

[tangent]

LAME doesn't use IS at all. That's a good thing.
--r3mix (and LAME in general) is known to have some stereo issues which is fixed by using --nspsytune with varying --nsmsfix. The --alt-presets are tuned to solve this issue (only useful at bitrates >160, below that and you waste too many bits in more crucial areas trying to fix the problem).

[mkanar]

Originally posted by Cokes

But what DVD,SVCD a VCD concerns (MPEG-1(2) Layer 2 audio): Layer 2 doesn't know anything about MS, it uses only IS, so if you encode your Dolby Surround donwmixed audio in Joint Stereo mode of any Layer 2 encoder - it destroys surround information.


Coke: Thanks for the info! I have been encoding 6 channel tracks as stereo and 2 channel tracks as joint stereo. My speculation was that 6 channel sound downmixed into 2 channels was too much for joint stereo. As it turns out, my theory was way oversimplified, but at least I recorded all of those movies correctly. Now I don't have to pause for 5 seconds every time I select between stereo and joint stereo as I understand which to use in each situation.

Are there any applications that can compare the effective output of a compressed mp3 file to the original and give some sort of information on the degration of quality? Well, didn't think so; oh well. :)

MKanar

[mkanar]

So in the case of MPEG-1-Layer-2 audio on an SVCD, assume that the source audio is 2 channel with no surround encoding, etc.

Now, should Joint Stereo be used at, say, 128-160kbps and stereo be used at 192kbps and higher? I'm just throwing numbers out.. but trying to figure out when it is best to use joint-stereo vs stereo. Obviously, I will always encode 6-channel-source audio as stereo, except maybe if I need to use a low bitrate?

Thanks!

[tangent]

If you are using MP3 and LAME, then you should not need to use full stereo at all. LAME's Joint Stereo is pretty good already and should not give you trouble, as long as you use the --alt-presets which uses further joint stereo tunings at the higher bitrates.

[mkanar]

*** MPEG-1-Layer-2 ***

This discussion is about MPEG-1, Layer-2 audio, not MP3 audio which is obviously superior. MP2 audio (often encoded with the tooLame library, not LAME for MP3) is the standard for SVCD discs.

Thank Anyhow.

Anyone else got some information?

MKanar

[nhope]

Thanks for all your feedback. To let you know a little more, I'm making original movies from Premiere normally with 3 or 4 audio tracks: parts of the original 48k stereo track from the DV tape and a couple of mixed audio tracks ripped as 44.1k WAV files from CD. I frameserve to TMPGEnc and I have tooLAME set as an external audio tool. So what we're saying in this situation is that I should just be using "stereo", right?

[mkanar]

If the source audio is 6-channel, then yes, definately use stereo (is my understanding). However, I am still not clear is reference to encoding 2-channel audio. Even the high-quality Fraunhofer MP3 codec is typically set to Joint Stereo: MS at 112kbps through 160kbps or 192kbps, MS/IS at 96kbps. MP2 is well-known to require a higher bitrate than a comparable-quality MP3 stream.

Just out of curiosity, how is it that you are encoding 3-4 audio tracks on an SVCD disk? As my increasingly failing memory recalls, SVCD discs can have upto 2 audio tracks.

Thanks all!
MKanar

[Cokes]

Originally posted by mkanar
If the source audio is 6-channel, then yes, definately use stereo (is my understanding). However, I am still not clear is reference to encoding 2-channel audio.

When your source is in example an 192 kbps 2.0 AC-3 stream with Dolby Surround, encode it in stereo mode MP2 at say for example 192 kbps to not destroy phase information.

When your sources are stereo recordings from your DV camera, with ease you can use let's say 128 kbps MP2 with joint (IS) stereo enabled. You become more space for you videostream...

That's a way what see I... :cool:

[nhope]

Sorry guys, what I mean is that I have those 3 or 4 tracks or whatever in my Premiere editing program. When I encode they are combined into a single audio track.

[nhope]

Erm, I'm back, having replaced all my gear that got nicked a few months ago. And I'm STILL not clear whether to use stereo, joint stereo or dual channel for making VCD's (i.e. MPEG-1 layer 2, 44100Hz, 224Kbit/s) encoded in TMPGEnc or CCE 2.62. Can anyone help?

[DSPguru]

use Joint stereo when your source is stereo, but if your source is a dolby surround track (meaning, it was downmixed from a 5.1 source), use stereo mode.

[frank]

Use BeSweet with tooLame.
My experience is that tooLame encodes in stereo mode regardless of your setting (s/js).
tooLame 192 kbit/s, psy model 2, stereo is a good choice even to DS2.
Measured with ce pro.