Can Someone Please Explain In Fairly Good Detail As To Why There Is So Much Impact To Audio Played Threw (Surround Sound Pro Logic Or Simulation Presets) once Encoded Be It MP3, WMA, Or Any Other Loosely Compression Merhod. AKA DEAD Channels, Or DSP's Can't Even Simulate A Field. Even At Rates As High As 3xx.Kbs MP3's. Is This Because Of Such High Compression? Or Is It Due To An encoding Not Found Any Other Format Than 48Khz (CD AUDIO) Never The Less Any Further Info Would Be Great. Thanks!! Please Reply To

How are you encoding your MP3s.
BTW, the capitalising of every first letter of every word is very annoying.

why are people always using so much CAPITALS

darn i had to read some of the stuff twice to figure out was he was meaning
alwmost as anoying and lame as to write3 In miX3d l33t taLk

Why respond to an email ???

The reason for having forum is that other people with the same question can seach and get an answer, and not have people to write question all the time when the question has been answered once

Sven, please try to avoid unnecessary ranting. I combined your messages to one, also edited some in the other thread.

Alright, I tried searching this board for answer but didn't find a relevant thread, so I don't think there will be any harm answering this. I'm sure however that this information is available at several places over the internet so a bit of googling wouldn't harm.

So if I understood correctly, you are basically asking why Dolby Pro Logic surround information is lost when applying lossy compression.

Dolby Pro Logic has only two discrete channels so additional control signals are required to represent up to six separate channels. Like in FM radio, where a single channel + control signals -> two channels perceived. These control signals have high frequency and low amplitude thus they are in inaudible range.

Since psychoacoustic compression (MP3, WMA, others) relies on removing inaudible information these control signals are usually lost, unless a really low masking treshold is used. My bet would be that only a tweaked MPC --insane would retain full surround control information.

This situation has been solved through Dolby Digital where there are six discrete, digitally stored, channels to begin with.

My experience proves otherwise. Yes, there can be problems: the kind of poor encoders that people here wouldn't use (e.g. Xing) really mess up the surround channels. And coding noise can sometimes be unmasked in the surround channels. But generally, it's not really a problem.

To the original poster: Are you using the Radium codec, or FhG ProdPro? If you are, you should know that it has a bug - it will destroy out of phase information.

Otherwise, I can't see how you are experiencing "DEAD Channels, Or DSP's Can't Even Simulate A Field". Are you using a Dolby Pro Logic decoder or a Dolby AC-3 (a/k/a Dolby Digital) decoder? (the latter WILL NOT WORK). Does it know to expect a Pro Logic stream?

mp3 at 320kbps played through a Dolby Pro-Logic decoder is very very good. It's not perfect, but it's not much worse than mp3 320kbps in normal stereo, and certainly much better than most people's threshold of thinking "I don't think that sounds right".


How can I put this less subtly? If it doesn't work, you're doing something wrong!

Cheers,
David.
http://www.David.Robinson.org/

P.S. - nik - I see why you thought it would be a problem, but Dolby Pro Logic is nothing like FM radio! You're thinking of an FM stereo signal with the pilot tone at 19kHz. But Dolby Stereo drives the rear channels with the difference between the two stereo channels, and the centre channel with sum of the two stereo channels. Dolby Pro logic adds a clever mechanism to mute the two front channels when there's only difference or only sum information, but that's all. It's all there in the audible frequency range of the two stereo channel. You can get at the "rear" information just by attaching a single speaker across the positive connectors of the normal stereo speaker outputs (though some amplifiers blow up if you try this - don't come crying to me if this happens - but I've never had it happen).

Dolby Pro Logic has only two discrete channels so additional control signals are required to represent up to six separate channels. Like in FM radio, where a single channel + control signals -> two channels perceived. These control signals have high frequency and low amplitude thus they are in inaudible range.

DLP doesn't work that way. It uses phase differences to derive the surround channel. If a signal is 180 degrees out of phase with the front channels then it gets steered to the surround channel. Maybe you're encoding with the IS (intensity-stereo) mode using some non-LAME encoder? The IS mode is known to alter or completely scramble phase differences between channels.

Well I'll be damned! You always learn something new . I recall picking up this piece of information years ago from a person I trusted to know what he's talking about.

I guess I (or he) had indeed mixed DPL with FM Stereo at some point of time. It seemed perfecly logical however.

Using a bit of a priori deduction, I would think that poor or too aggressive joint stereo implementations (and I don't mean just IS) especially at lower bitrates could do serious damage to a DPL sound field. Seemingly subtle stereo image distortions and quantization noise could cause pretty weird effects to the surround field. But you get what you pay for, whereas 320 kbps MP3 may indeed retain the surround field, there is no reason to expect that a 64 kbps WMA does.

I'm seem to be a bit sceptical however. Psychoacoustics compression and meant-to-be-innoticeable control information in signal just doesn't make any sense to me. In digital world we can easily use n discrete channels for surround sound instead of these hidden signal control "cheats" .

...if only MP3 (standard) supported n discrete channels... but it doesn't (or does it?) so let's go with dpl or switch to OGG / wait for MPC/SV8

I have my doubts on that. The Mid-Side stereo encoding process is exactly the same as the decoding process in Dolby Surround. They both take the difference of the two channels. MS stereo encodes this difference as the Side channel, while Dolby Surround dumps the difference into the rear channel. So none of the Dolby Surround information is lost when you do MS Joint Stereo encoding. It might even be possible that Dolby Surround information will be retained better with MS JS than with stereo.

A lot of people have reported that Lame handles perfectly pro logic info. Of course it won't work at 100kbps because of the lowpass filter. It seems that it starts working from 160kbps.


Also once again: MP3 SUPPORTS 5.1 DISCRETE CHANNELS. (there is just no implementation yet)

Quote: Also once again: MP3 SUPPORTS 5.1 DISCRETE CHANNELS. (there is just no implementation yet)

That brings up an interesting question: How would LAME handle joint stereo with all those channels. I know Dolby's AC-3 does something like this, but I'm not sure of the details. So how would LAME do it using joint stereo?

I don't really know how Lame would handle it. What is sure is that right now, Lame can't handle more than 2 channels, and I don't think this will change soon.

But even without encoder support for 5.1, there could be a way:
producing several mp3 with a regular mp3, then process the bitstream with an extra tool to merge the tracks into a 5.1 track.