Stereo vs. Joint Stereo

[Lite]

Well, everything went demagoguery. Well not really the same! Audio speakers is also a sound system, and the difference can be heared on it too. If you think so, would not exist different bitrate, and would be one fixed.

If you think this way, there would be no different bitrates, and would be one a low fixed bit rate. But many specially listening to lossless quality, and you are reasoning that there was no difference at such low bitrates. It is strange.

But Ok. Here are another problems.
At 48k you uses AAC+v2 or v1. Although this is a sufficient bitrate for AAC, but there could only hear whistle of high frequencies and much lost of detail because signal is restoration from low frequencies with the use of SBR technology, or even sooner Parametric Stereo (PS) on very small bitrate. AAC+ is the only way to use such low bitrate. This is a good choise.

The following is 96k which is more than enough for truly high-quality playback when using the AAC+. But there you are using MP3, which at this bitrate even with the use of Joint Stereo simply can not provide any quality at all. Quality awful in itself, and distortions are so high that their presence does quality absolutely bad. And it already turns out quality lower, than at AAC+ on 48k.
But you call quality at this bitrate average, than more ordinary users keep misconception that the quality depends on the bit rate, although in reality it is not. On such bitrate it is necessary to use at least OGG Vorbis, but nevertheless it is better AAC+.

The following is 192k on which you are using MP3. But even here you have worsened the options of refusing to use Joint Stereo.

You have not selected the correct settings on two of three bitrate. I thought that this progressive community. Why AAC+ puts severe limitations such as the use of only 48k. Why use 96k stream MP3, not AAC+? As that crookedly you have distributed formats. You have decided to follow an example my useless country, where all upside down? Oh, don't do this, please.

By the way, if you suddenly turn on Joint Stereo on a 192k stream and will make poll about sound quality on 192k channel not specifying the reasons, it will be interesting to look at results.

And here that I still wished to tell about Joint Stereo. Speaking easier... Why it's worth that to include even 320k? Because 320k is 4.5 times less than 1411k. Even what heights has now reached the best MP3 codec Lame, it still lacks the bitrate in order to portray the source material without distinguishable from hearing loss. Setting codec to Joint Stereo mode, we are helping to achieve a higher quality, because we increasing the virtual bitrate.

[XDR]

But, now what if a DJ's set that was compressed to Stereo mp3 is broadcasted on a Joint-Stereo stream, would you still hear any difference between broadcasted stereo or joint-stereo.

Or the other way around? Broadcasting a Joint-Stereo mp3 on a Stereo stream? What does that do to sound quality?

Just reminding you that there's two stages of mp3 compression at work here, or in most cases actually 3 (since most DJ's play mp3s). Quality loss is everywhere.

[Lite]

Yes it so. But files used by them have much higher bitrate. In this case losses is much smaller. Plus there is usually used the qualitative codec - Lame with quite good options of quality. And such system is not adhered to real time, therefore it is possible to expose though the highest quality, and thus coding consumes many resources. And when the announcement strongly good options of quality is carried out will not put because of a binding to real time. Also it is absolutely not known what codec you use. Therefore tracks coded on standard, instead of the broadcasting codec will be more qualitative, in the absence of restrictions on consumed processor resources. Plus even if to take an ideal case when the sound takes from lossless source the last stage will be solving. As and at us. It some kind of a bottle neck.
Still such example. If on an average on quality of a sound of audio the equipment in the end to include cheap ear-phones the sound will be awful. And if to connect expensive ear-phones the sound will be much better in any case! Therefore always it is necessary to try to keep the greatest possible quality at any stage.

[dfx]

why don't we all just go back and use uncompressed PCM everywhere?

[Lite]

Well only it is not necessary so to exaggerate. Speaking about attempts of preservation of quality I had in view of reasonable decisions and compromises. As an example besides I will result obligatory use Joint Stereo and a reasonable choice of the used codec. That is at bitrates lower than 128 it is obligatory to use AAC+ etc. The head is needed not only to carry the cap and eat...
I stake on progressive technologies their correct application and adjustment! Knowledge of features of formats and their realisations. Knowledge of quality of decoders eventually (Snibatch, MAD, MP3 Decoder of Apollo Player. Last one is most precise of existing decoders).

[Lite]

Also use of higher quality options of coding. In Lame, for example, it is enough of them.

[dfx]

Originally Posted by Lite

Well only it is not necessary so to exaggerate.

well tbh, i think it's you who's exaggerating about how much of a difference it all makes.

care to participate in a single-blind listening test?

[t4e]

Originally Posted by Lite

If you think this way, there would be no different bitrates, and would be one a low fixed bit rate. But many specially listening to lossless quality, and you are reasoning that there was no difference at such low bitrates. It is strange.

no, you need to read more carefully what people say, i said there is no audible difference between 192 and 256

Originally Posted by Lite

Setting codec to Joint Stereo mode, we are helping to achieve a higher quality, because we increasing the virtual bitrate.

are you implying that using joint stereo increases the bitrate?

i'm sorry, but most times i have a hard time following what you're trying to say

[Lite]

Well I also do not know that else it is possible to tell. If after all explanations you and have not understood importance of all of it... I Can add only what to experiment and study I all it has begun when for the first time has got acquainted with Lame. It then was on version 3.93.1, that is it was somewhere in 2002. I much experimenting. Traced each subsequent version Lame and well heard differences in sounding of these versions. Also I can describe it. It differed very strongly, believe me! And always used and will use the best option, that is "-q0". The unique version which I have passed it 3.95.

There was also a period when I could not be defined that better Stereo silt Joint Stereo. But as a result all the same has understood that Joint Stereo it is much better. Then I began to always use "-mj" setting.

[Lite]

Originally Posted by t4e

are you implying that using joint stereo increases the bitrate?

i'm sorry, but most times i have a hard time following what you're trying to say

If to paraphrase effect which we receive from Joint Stereo the virtual increase of bitrate turns out. Really so it is difficult to understand it? I repeat - virtual!

[t4e]

Originally Posted by Lite

If to paraphrase effect which we receive from Joint Stereo the virtual increase of bitrate turns out. Really so it is difficult to understand it? I repeat - virtual!

that which i highlighted above makes no sense whatsoever

its really useless continuing this discussion, no offense but try using simpler english


i'm done with this subject

[Lite]

I use translaters, sorry.

[dfx]

ok then, you didn't reply to my question, but here we go anyway :D

i took a random recent track in WAV quality and extracted a 1-minute part of it. i encoded this part in 4 (well, 5) different ways:

1) 320 kbps joint-stereo
2) 320 kbps stereo
3) 192 kbps joint-stereo
4) 192 kbps stereo
5) uncompressed - original WAV

i jumbled those around and gave the samples random numbers. the original WAV is also provided seperately a second time (labelled as such) for comparison.

all files -----> Index of /blind

everybody is invited to participate! but read the rules!

this is supposed to be a blind listening test, which means you may listen only, no looking at spectrograms or comparing files or anything else allowed. listening only!

also this is supposed to be a blind test, which means do not post your results here, instead send me a PM with your results! i will reply and collect the messages and then post the collected results after a while

ps: encoder used was LAME 3.98.2 and decoder used was libMAD 0.15.1b.

[a'ndY]

Originally Posted by dfx

why don't we all just go back and use uncompressed PCM everywhere?

i second that motion

anyway, settings for highest quality mp3s with lame is -b 320 -q0 (lame automatically uses -m j). if that makes an audible difference to -b 320 -h or -b 320 -h -m s is another story.

[Tarek]

Originally Posted by dfx

why don't we all just go back and use uncompressed PCM everywhere?

..and hand out 6 hour DJ sets on Blu-Ray discs