Using the --alt-preset insane, extreme, or standard settings, is 'surround channel' information preserved when I encode an MP3 using LAME? If not, what can I do to preserve it so I can listen to my MP3s through my 5.1 system without having problems with that part of my music? I heard that Joint Stereo is what causes the surround channels to either completely disappear or just partially go away but, it wasn't solid proof. Can someone cleae me up on this? I used to encode my MP3s at 320 kbps Stereo with no presets (just straight 320 kbps with everything at default) and had no problems with the surround channels but now, I notice that when everyone does these listening tests, they have to do it WITHOUT using surround channels (eg. using regular stereo headphones on equipment with no modifications). I was wondering if this was because the Joint Stereo DOES INDEED do something with the surround channels. I have some portable and home audio equipment that lets you hear this 'surround channel' as a form of 'virtual surround' modes.

Thanks...

Phase information which is used to encode surround information into a stereo signal is fully preserved by the joint stereo algorithm used in LAME. You'll hear the same surround content on a LAME-encoded MP3 as on the original WAV file.

Of course, keep in mind that MP3 encodes stereo signals, and most music from CD's is simple stereo music without surround content. You should *not* listen to such music with amp or speakers in surround mode if you want to hear the music properly. It may seem like you are uncovering added "hidden" realism by playing non-surround stereo tracks in surround mode. In reality, you're butchering your music with a rusty ax. Such phantom surround effects would definitely get in the way of a listening test meant to spot more subtle encoding artifacts.

Here is a thread on the subject listed in the MP3 FAQ for this site:

From the horse's mouth; stereo image myths

not true, sound wishy-washy, sea-shell pre-echo crappy to me


um, if you insist on using Lame, 320/Stereo q=0 will achieve the best rear channel accuracy; also I find 3.93.1 the best for CBR (not --alt-preset insane which is joint stereo) big waste of space IMO, less than satisfactory sound at a high cost


I measure a codec's integrity on how well it preserves the musics rear channel info, say... when the music is processed by a digital/pro-logic amp (from an optical source).

codec's that I find to have near perfect accuracy to the original wave are:
Nero 2.6.1.1 LC-ACC \ CBR 224 (flawless, thank you nero!)
& Musepack SV7 1.14 \ -quality 8 (braindead mpc's are great too, but too big!)

I personally enjoy streaming these codec's to my amp, good natural sound.
full bandwidth too, i think... anyway it's important to me, oh yeah & Lame doesn't compare

thank you HA for teaching me what I needed to know, adam breakaway from mp3 go mp4

ooops, AAC

Where did you come from???


Have you ever heard of an ABX test?

Read the Terms of Service:
-> 8.

Hydrogenaudio is supposed to be an objectively minded community that relies on double-blind testing and relevant methods of comparison in discussion about sound quality. The usual "audiophile" speak of non-audio related terms which are completely subjective and open to redefinition on a whim, are useless for any sort of progression in discussion.

This rule is the very core of Hydrogenaudio, so it is very important that you follow it.
You can read how to easily perform double blind listening tests here

--alt-preset insane is about 99,5% normal stereo and 0,5% joint stereo.

sorry dreamliner, just realized NOW this is a tech board, sounded like adam was looking for an opinion. I hope my comments don't offend anyone

was does CBR and joint stereo has to do with each other?
this sounds like complete nonsense

seems lile rjamorin has do do a dolby prologic mp3 testing soon :-)

Sorry, I was/am a little burnt out and dealing with the trivial BS is just getting to me.

Stick around for a while...

Thanks for whatever advice you could give me but, I must use MP3 for my portable. If I wanted something for JUST my PC, I'd use FLAC only, like I do album archiving on my PC and across DVDs. MP3 is the most common compressed audio format out for portables so, I need information regarding it only. For the record, I don't have my equipment in 'movie mode', like some on here may think. Some of my music modes have a version which have both the regular front channels and the surround information are within the front channels. Others also don't 'split' the front channel sound either (the surround information is in the rear channels and all of the regular information (like what you hear with headphones without enhancements) is still in the front channels). In fact, on some portable CD/tape/flash/HDD/etc. players, you can pull the headphone jack out halfway or so and hear just the 'surround' information (which is in monoural).

What you hear in this case is the difference between stereo channels (= substraction). Using this 'DSP' (in fact it's similar to simple karaoke/vocal cut DSPs) you'll hear compression artifacts with most lossy encoders/settings, especially if there's a loud signal in the centre of soundstage like the singer's voice, that masks these artifacts under normal conditions. Are you sure this is what is coming from the rear channels? Seems too simple to me...

very few (compared to mainstream 2 channel) 4 channel releases exist. And here's a big one, why don't you just test it out?

Actually, yes. when I remove the centre channel (a la karaoke mode, vocal removal, etc.), I would get artefacts on LAME 3.93.1 MP3-encoded audio encoded at a straight 224 kbps (and sometimes 256 kbps) bitrate without any tweaks. Now, since --alt -preset standard and extreme generally uses lower bitrates, I was wonder if this would disappear. I don't have 'golden ears' like many people on here probably do so, that's why I'm bringing the issue up here. Maybe someone who does karaoke and removed the centre channel from their music can explain their experience? Thanks! I want to use one setting that will cover all music and never have to look back.

I doubt that something like this is even possible with psychoaccoustic compression. You might want to try wavpack lossy - or just use lossless.

If you intend to use any form of sound manipulation on your audio, you shall not use psychoacoustic lossy compression. Such codecs removes sounds (add noise) in places where it is masked (hidden) by other sounds in the music. If you afterwards remove/change these masking sounds, you will here artifacts...

For this kind of use, your only choice is to use lossless compression (or non psycho-acoustic lossy compression, like wavpack lossy).

We've had three threads on this before, and found that both lame and MusePack aps/standard/q5 completely mess up the surround channel. This is obvious if you listen to the surround channel in isolation (or put your ear right next to the speakers). You don't need to ABX!!!

The final thread suggested methods for making the surround channel encode without problems using Lame and MusePack. I also suggested that, if you sit in a sensible central position, the problems are probably not audible anyway. One other person tried this and agreed, but it's hardly been extensively tested!


Here is the third thread which also refers to the two earlier threads, and provides some answers. Hope it helps...

http://www.hydrogenaudio.org/forums/index....showtopic=12004


Cheers,
David.

P.S. I've checked, and I can't see these "surround" threads in the FAQ. If they're really not there, can someone add them please?