MP3 Surround Format Debuts Today

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

bmoura

2K Club - QQ Super Nova
Since 2002/2003
Joined
Mar 2, 2003
Messages
9,518
Location
Redwood City, CA
From today's NewsWire, the debut of the MP3 Surround Format:

mp3 SURROUND Set for Launch
Fraunhofer IIS, Thomson, and Agere Systems Unveil
Free mp3 SURROUND Evaluation Download


Full Versions Of mp3 surround Encoder And Decoder
Available at www.mp3surround-format.com

Paris, December 23, 2004 – Thomson (Euronext Paris: 18453; NYSE: TMS) today announced that mp3, the world’s most popular audio compression format, goes multi-channel. Fraunhofer Institute for Integrated Circuits IIS, inventors of mp3, together with Thomson, co-inventors of mp3 and exclusive licensing representative for their mp3 patents and software, and Agere Systems (NYSE: AGR.A, AGR.B), have joined to launch mp3 surround into the consumer and commercial markets. This technology will enable 5.1 channels surround-quality sound for a broad spectrum of applications including web-based music distribution, broadcasting systems, PC-related audiovisual or gaming applications, consumer electronics and automotive systems.

mp3 surround supports high-quality multi-channel sound at bit rates comparable to those currently used to encode stereo mp3 material, resulting in files half the size of common compressed surround formats. At the same time, the new format offers complete backward compatibility to any existing mp3 software and hardware devices.

Attracting attention at several recent industry events, the new mp3 surround format has become eagerly anticipated by the vast mp3 user community. The web site www.mp3surround-format.com now provides users immediate download access to free mp3 surround evaluation software, demo samples and detailed technology information.

Thomson, which has partnered with Fraunhofer IIS since the codec’s launch in 1992, is adding Fraunhofer Institute’s new mp3 surround technology to its licensing program.

“The free mp3 surround evaluation software will enable everyone to test the latest feature of the digital audio standard. We encourage music artists, engineers and producers to experience this new multichannel audio format and realize the benefits of its portability and backward compatibility,” said Henri Linde, Vice President, Growth Initiatives, Audio & mp3 program of the Intellectual Property and Licensing unit for Thomson.

The evaluation encoder enables the creation of mp3 surround files out of 5 or 6-channel wav material. The Fraunhofer IIS mp3 surround player is capable of decoding and playing back the surround format’s files as well as stereo mp3 material. This software-only solution runs on any standard PC with multi-channel audio capabilities.

mp3 surround was developed by Fraunhofer IIS in collaboration with Agere Systems. Using a psychoacoustic technique called binaural cue coding, mp3 surround captures the spatial image information of multi-channel sound. This method is critical in preserving the compact file size that mp3 users expect.

“We are delighted to advance mp3 into the area of multi-channel audio, with the help of Agere Systems. As surround sound is increasingly gaining importance in the marketplace, we are confident that mp3 surround will attract the attention of music production professionals and consumers alike,” said Juergen Herre, Chief Executive Scientist at Fraunhofer IIS.

“Together, Fraunhofer IIS and Agere are developing technology that will contribute to the delivery of a compelling compressed multi-channel audio solution into the marketplace,” said Peter Kroon, chief multimedia architect for Agere’s Mobility business. “This work is part of Agere’s overall commitment to provide advanced multimedia technology that enables consumers to stay connected.”

The evaluation period of the mp3 surround encoder will expire on the 31st of December 2005, while the Fraunhofer IIS player will continue to remain fully functional. The use of the software is allowed for personal and non-commercial purposes only. Professionals can license mp3 surround from Thomson at www.mp3licensing.com.
 
Oh god, here we go yet again.
To be quite honest, this is the last thing that i think surround needs - to be taken down the same road that redbook went and crucified on the altar of MP3.
We already have compressed surround formats - Dolby ProLogic/Dolby Surround, Dolby Digital and DTS.
DD already runs at 192Kbps in stereo, and a 192 bitrate for surround is frankly a joke when you stack it up against the original 24/96 bitrate of 9.8 MegaBits/second - compressed losslessly with MLP. At 24/96 uncompressed we are talking a data stream of over 13 Megabits/second.
Yet these people keep trying to tell me that it does not matter if I throw away most of my Audio data as it is "redundant".
If it were redundant - it would not be there in the first place.

Please, can we just not mention this abomination in the same place as DVD-A/SACD/DTS and DD multichannel, and to even think about how dreadful it would be if some moron put the Alan Parsons DSOTM on MP3 surround?
God forbid.

There is absolutely no need at all - apart from greed - to even consider introducing this disgusting parody of a format.
MS have it covered with WMA Pro - which sounds far better but is still not that good. DD is better still, and DTS rules the lossy compressed roost by a long , er, crow.

This thing has been introduced by someone who is trying to make money by reinventing the wheel, but as Hengist Pod did in the film "Carry On Cleo" by making it a square rim. (Sorry about the obscure ref - now you'll have to chase it up to find out what I mean. "Hengist Pod, Square Wheel Makers")

TO BE AVOIDED AT ALL COSTS
 
bmoura said:
From today's NewsWire, the debut of the MP3 Surround Format:

Brian,

i have downloaded the mp3surround package by a while (it's available since dec. 11), made some test and then thinked about it.
There are several issues with MP3surround, and all are not on the "technical" (code) side.

1) Hardware problems
MP3s can sound decent and while it's easy to spot a mp3s file - played in stereo - on a serous hifi setup, for the typical JoeSixpack A/V centre that-plays-everything (dvd svcd vcd cd mp3 jpeg etc etc etc) it sounds ok: no difference from a dts-encoded cd or a redbook cd on such players.
However, there are some issues to consider:
a) at present time, there's no hardware that can play mp3s files - this could be easily solved if Fraunhofer do a little partnership with 4 or 5 chipset producers of dvd players (ess, mediatek, sigma, sunplus...) with 5.1 analog output for a firmware update, replacing the stereo mp3 code with a mp3s code. It may seems impossible but the dvd market is *full* of clones and the actual chipset produced in the last 6 years are +- a dozen, nothing more.
Does it makes sense? Yes *only* if the "software issues" are solved (see point 2). Otherwise, the market has already DD, DTS, WMA that all can handle lossy mch file and are widespread.
b) a positive for mp3s is a good stereo compability, something that wma handle with more difficulties.

2) Software problems
Here are the hardest issues:
a) there's no need to mention here the slooooooow pace of musical - not video - surround releases in *any* format (dd, dts, dvda, sacd, dualdisc); if Fraunhofer and their friends are not able to persuade the Majors to release a *lot* of their catalogue in surround and make it available on online stores, what would you listen? 4 demo songs?
b) MP3 had a good shot for two reasons that now are not true for the surround format: first, rip a cd was a easy task and there was already software that could do it well when mp3 came around; second, mp3 was pratical for sharing and downloading a lot of files even with a 56k modem, and all for "free" (they used to call it piracy). Nowadays current surround releases are hard to "convert" in other format and, more than ever, are very scarce. Furthermore, dsl/cable connections are more widespread and thus there's less need to have a smaller filesize for download - another format such as dts is already playable on a ton of machines without a upgrade and still not to hard to dl with a fast connection. All this make us going back to previous point: what would you listen?
c) converting to mp3s the old back-catalogue (namely quad) would make us happy (or at least some of us) but what appeal can have for the mp3-people? Next no nil. Need to have a Eminem in surround - for the artistic point of view it could be a zero value but for a marketing plan that's what millions want.

Conclusion: do we need mp3s?
That's more a software issue than a hardware issue: if there are files to download - legally or not - that are good and not available in other surround format, mp3s can have a chance; otherwise, it's just a tech showcase that is out of time and out of sense.
But... will make sense for a Major to release, for example, Eminem and U2 in mp3s and *not* in dvd-a or sacd? Probably not. Thus... what chance has mp3s to live? IMHO, very few.
 
JonUrban said:
The future look bleek for real music and sound

The real market for technologies like MP3 Surround is 5.1 Radio Broadcasting.

There are several systems like this one out there - competing to be the technologies used for Surround delivered by Sirius, XM, HD Radio, etc.

As to which one wins (if any), we'll have to see.
 
bmoura said:
The real market for technologies like MP3 Surround is 5.1 Radio Broadcasting.

There are several systems like this one out there - competing to be the technologies used for Surround delivered by Sirius, XM, HD Radio, etc.

As to which one wins (if any), we'll have to see.
Hey, that's interesting. I never thought of it along those lines.
I'm sure there are a lot of things that I haven't thought about! :eek: ;)
 
I'd like this technology only in the context of "5.1 Radio Broadcasting". Otherwise, I'm kinda 'ho-hum' about it and don't see a need for it.

Back to radio... I somehow doubt we'd ever be fortunate enough to have real surround music take off in a radio broacast---terresstial or satellite. If it did catch on, that'd be great---and a boon to those of us who like surround.
 
Eugenius said:
I'd like this technology only in the context of "5.1 Radio Broadcasting". Otherwise, I'm kinda 'ho-hum' about it and don't see a need for it.

What purpose on 5.1 radio broadcasting if 5.1 music is lacking?
A 5.1 of a live show may be as well broadcasted in stereo with a DPL2 encoding, to recreate a hall ambience is enough. No need for a new format.
 
MP3 is the devil's work. Especially at the stupidly low bit-rates regularly used. Hell, it's (almost) as bad as Musicam, the obsolete piece of junk the UK government is attempting to foist on us all as DAB radio, and then switch off analogue so they can flog off the bandwidth to communications companies for marginally less than what you'd find down the back of the sofa.
This is bad. Very bad. Perhaps inevitable, the way things are heading, but that's hardly sugaring the pill.
Perhaps there will be some light though. The masses don't care about sound-quality, correct? But they do like their gadgets. Well, that's fine. Because what goes around comes around. In a few years time, someone, somewhere with some marketing ability (unlike, say, Sony or the DVD-A people) is going to make decent sound-quality fashionable again. Whilst I have little time for fashion (an excuse for mugs to pay extortionate prices for clothes et al made by slave-labour at regular intervals), sometimes, it has side-effects... here's hoping.
 
winopener said:
What purpose on 5.1 radio broadcasting if 5.1 music is lacking?
A 5.1 of a live show may be as well broadcasted in stereo with a DPL2 encoding, to recreate a hall ambience is enough. No need for a new format.

It would be just like the early days of MTV when there were only, say, 25 videos available. ;)
 
I am still naive enough to believe the only reason the masses seem to prefer MP3 and overcompressed garbage like it is out of ignorance, and not choice.

I still love my old DTS transfers of the older Quad titles - but the 24/8.2 versions on DVD-A sound infinitely better.
Most of those who listen to MP3 are comparing them to current CD's which are far too loud due to far too much heavy handed use of boxes like the dreadful finalizer, and the equally abhorrent L1/L2 maximizers. I mean, how can you take any tool seriously where it says in the manual that a good place to start is with -6dB attenuation! My god, that is a full 50% of the original dynamics just written off.

CD can be very good when done properly. Trouble is that most of it isn't due to the insane volume wars these days.
Digital will allow this type of butchery to be committed on innocent mixes - but just because we can do something does not mean we should.

As for MP3 - how anyone can believe that it is okay to simply throw away 11/12 of the original audio and still expect it to sound the same is quite beyond me.
192kbps? get a grip please - this is nowhere near enough for decent surround.
We all know how much better DTS sounds over Dolby Digital, and how much more so DVD-A than DTS.
There should be no room for MP3 surround anywhere in the world IMHO, although given that MS have already got WMA Pro in Surround, it was sort of inevitable.
 
neil wilkes said:
Oh god, here we go yet again.
To be quite honest, this is the last thing that i think surround needs - to be taken down the same road that redbook went and crucified on the altar of MP3.
We already have compressed surround formats - Dolby ProLogic/Dolby Surround, Dolby Digital and DTS.
DD already runs at 192Kbps in stereo, and a 192 bitrate for surround is frankly a joke when you stack it up against the original 24/96 bitrate of 9.8 MegaBits/second - compressed losslessly with MLP. At 24/96 uncompressed we are talking a data stream of over 13 Megabits/second.
Yet these people keep trying to tell me that it does not matter if I throw away most of my Audio data as it is "redundant".
If it were redundant - it would not be there in the first place.


?? 'Redundancy' is determined by audibility testing, and driven by such real and well-established psychoacoustic phenomena as acoustic masking. It has been demonstrated numerous times that data can be removed without ANY audible loss. Audible artifacting is dependent on bitrate , the algorithms/codecs employed, and the nature of the signal being encoded.

The question is always whether a given combination of these *does* result in audible difference. It is not a given that it *will* (or will not).


MS have it covered with WMA Pro - which sounds far better but is still not that good. DD is better still, and DTS rules the lossy compressed roost by a long , er, crow.

This has by no means been settled; both Dolby and DTS have been presenting evidence in their favor for years, and the arguments pro and con are easily found on the Internet.


MP3 surround at 192 kbps will undoubtedly find its niche, as MP3 @128 has.
As have small desktop radios, computer sound systems, car audio, and other 'less than hi fi' appliances. It's good to have choices.
 
Sorry, I do not buy into this at all.
There is no way on gods green earth you or anyone else will ever convince me that it is okay to throw away vast chunks of data and still have it exactly the same as it originally was.
I can assure you played over the right system - not headphones, not PC speakers and certainly not the garbage that passes for MIDI systems so common today, it is perfectly easy to pick the differences between even a DTS version or an MLP lossless version of exactly the same mix.

I have both DTS and DVD-A copies of Sinatra at the Sands. The DTS version has nothing like the same openness and detail.
Moving swiftly on then.
Queen, A Night at the Opera.
Even at the 24/96 DTS version used, you can hear the difference between the DTS and the MLP.
Another example - Lennon's Imagine Quad version...
I have a DVD-A transfer at 24/88.2, plus a DTS version I encoded from the same PCM audio files. You can hear the difference.

So, if this is true with DTS, how much more so with MP bloody 3?

A throwaway format for a throwaway generation.
Next thing we know, it will be maximizers & multichannel finalizers, so that the studios can butcher the multichannel mixes too, kill all the dynamics, and pretend that an MLP version sounds no better than the MP3.
Which crucified like that, it wouldn't.

But I can assure you that any piece with dynamics and a good mix that has not been brickwalled will be far better in PCM than MP3.

I repeat - you cannot throw away most of your data and still have it sound the same. Similar, but not the same.
You'll be trying to tell me next that JPEG is as good as bitmaps.
Or MPEG-2 is as good as broadcast video.
 
This is potentially bad news for surround. Especially if there is an attempt to replace the hi-rez disc formats with the surround MP3 format. The way things are going I think the general public would accept this, as a great deal of people have taken to stereo MP3's as a replacement for stereo PCM delivered on a CD.

On the DTS Topic, if you A/B DTS and hi-rez LPCM/PPCM then there is no doubt a difference, favouring hi-rez LPCM/PPCM. I remember when I first got DVD-A capability. There were a lot of DVD-A's in my collection where I was playing the Dolby/DTS streams. When I switched to the hi-rez stream I noticed a huge difference as I was at that point use to the Dolby/DTS streams on those titles having listened to them numerous times. However, if you listen to something like The Sessions for Robert J. in DTS it sounds fantastic to me. If the title comes out in DVD-A there will be an opportunity to do an A/B comparison.
 
Sorry, I do not buy into this at all.
There is no way on gods green earth you or anyone else will ever convince me that it is okay to throw away vast chunks of data and still have it exactly the same as it originally was.
I can assure you played over the right system - not headphones, not PC speakers and certainly not the garbage that passes for MIDI systems so common today, it is perfectly easy to pick the differences between even a DTS version or an MLP lossless version of exactly the same mix


Actually, good headphones would be a *more* sensitive means for hearing differneces between a lossy compressed and native track, especially if they manifest themselved during very quiet sections. And as I said, it depends on 1) how much data is 'thrown away' (bitrate), 2) *how* it is thrown away (codec), and 3) source material. If you choose not to believe it, fine, but unless you've done blind testing involving these variables, as has been done often on www.hydrogenaudio.org, the claim has no objective force.

Note that I'm not saying there are no audible differences; I'm saying there can be. You seem to be saying there *always* are. In the case of your DTS to DVD-A comparison, it's possible you heard a real difference. Or not, since you don't note whether levels were matched and the test was done blind. It's also possible that the difference is manifest on the mateial you heard, but might not be on other material (see criterion #3 , above) or using a codec other than DTS (#2 and #1).

But I can assure you that any piece with dynamics and a good mix that has not been brickwalled will be far better in PCM than MP3.

I can assure you that if you set the bitrate high enough, you are very unlikely to detect a difference on most musical sources. But don't take my word for it. Encode some music using hydrogenaudio.org's recommended LAME compile, using the --insane or --extreme presets, and do a blind ABX of them . If you can't hear a difference , move on teo the -standard preset and see if you can do it then.

I repeat - you cannot throw away most of your data and still have it sound the same. Similar, but not the same.
You'll be trying to tell me next that JPEG is as good as bitmaps.
Or MPEG-2 is as good as broadcast video.

You cannot make credible sweeping statements about lossy codec sound based on mp3s where 'most of the data' has been thrown away, and sighted comparisons.
 
Last edited:
On the DTS Topic, if you A/B DTS and hi-rez LPCM/PPCM then there is no doubt a difference, favouring hi-rez LPCM/PPCM. I remember when I first got DVD-A capability. There were a lot of DVD-A's in my collection where I was playing the Dolby/DTS streams. When I switched to the hi-rez stream I noticed a huge difference as I was at that point use to the Dolby/DTS streams on those titles having listened to them numerous times. However, if you listen to something like The Sessions for Robert J. in DTS it sounds fantastic to me. If the title comes out in DVD-A there will be an opportunity to do an A/B comparison.


Are you sure the DVD-A and DTS versions on the same disc are identical *except* for their encoding?
Are you sure your system outputs DTS and DVD-A tracks at the same levels, crossover, and EQ?
Was your comparison done blinded?

If the answer is no to *any* of those questions, then it's unreasonable to assert conclusively that DVD-A and DTS formats sound different based on your tests.
 
'fraid I'm still with Neil on this one. The old Linn LP12 catchphrase of garbage in, garbage out leaps to mind.
I'm a cynic about many audio related matters. Fancy cables; the odd notion that solid state power-amps can in some way be microphonic; mega-priced caps... I could continue.
So, the high-end is daft -we all know that, just as £1000 of CD spinner will give 95% of the quality, if not more, of one costing 10 times that. But there is truth there. Grab a CD and stick it on something like a Meridian 507 feeding, say, a decent entry level amp and speakers and it will sound excellent. Reverse this; get £10,000 worth of amp and speakers fed by an mp3 file, and it'll sound insipid. 16 bit done well is very good. Done badly, it's poor, like anything, but it's not improved by chucking out half the data.
I mentioned Musicam earlier. Not mp3, but it's also a compression system, and one supported by the UK government for DAB radio. I've done what amounted to an double-blind test between £600 of Arcam DAB tuner and £99 worth of Denon TU260L FM tuner using BBC Radio 3, which uses the lowest compression possible in DAB.
Night and day -what I identified very quickly (and correctly I hasten to add!) as FM mugged it in every area. Draw your own conclusions methinks. But chucking away large quantities of data might well be part of the problem!
 
'fraid I'm still with Neil on this one. The old Linn LP12 catchphrase of garbage in, garbage out leaps to mind.

Indeed it does. Garbage assumptions in, garbage conclusions out. That catchphrase long precedes Linn, btw. (And the head of Linn makes some pretty garbagy claims in the most recent issue of TAS, btw.).

Reverse this; get £10,000 worth of amp and speakers fed by an mp3 file, and it'll sound insipid.

*AGAIN*, that depends on the mp3. Try it with a well-encoded and decoded high-bitrate mp3 of music in a blinded level matched test. The assumptions behind your claim are many, and not all well-supported. I'm rather sure I could make an mp3 for you that you would find impossible to audibly distinguish from the redbook source, using your proposed system in a properly controlled comparison.



16 bit done well is very good. Done badly, it's poor, like anything, but it's not improved by chucking out half the data.

And what does this supposedly prove about mp3 in general? That the more data you discard, the greater the *possibility* of audible artifacting (all other variables being unchanged)? True. No one has said otherwise. its' one of the three main criteria I mentioned as affecting the sound of a compressed audio file. Btw a crummy recording might be even *less* sensitive to audible mp3 artifacting at lower bitrates, simply because there's so much else that's inhernetly crummy-sounding about the *source*.

Your Musicam comparison as described doesn't make sense to me. Were you comparing tuners or were you comparing lossy compressed FM to native FM? How large is 'large quantities of data' in the case of Musicam? DAB radio? Bitrates might indeed be part of the problem. That's always true of a lossy codecs. But it's not true that *any* lossy encoder will *always* produce audible artifacts, which is what seems to be the underlying assumption.

AGAIN: bitrate, codec, source material. Any can affect the level of audible artifacting in lossy compression.
 
You'll have to forgive Neil his overzealous biases. (I hope he'll forgive yours'!)
After all, a main thrust of his business is DVD-A. As such, it's in his best interests to be it's proponent.
Now, mind you, I'm not saying he's wrong. Actually he's pretty much correct about the audio quality issues.
He's just not as objective as he could be.
I myself have been in these same kind of positions and was just as hardheaded, if not worse.
Nothing worse than having an endorsement deal only to find something, (that might not even be better), about to drive it away. :eek:

Back On Topic
While not endorsing MP3s, I would like to say that downloadable files are 5.1's only hope.
After all, most of the music buying public is moving away from discs. (And movies aren't far behind!)
If they can find a system that sounds OK, folds down to stereo properly and is easily transportable,
(yet copy-protectable), the labels might find it more economically palettable if not profitable to pursue.
Really, the practical needs have to come first, then the issues of higher quality.
Hey, maybe someone should be investigating MLP based 'ipod' type devices!!!
 
My main overzealous bias is towards the rule that conclusions go as far, but no farther, than evidence warrants. That also means knowing the pros and cons of different kinds of evidence. Guess I'm kooky that way.


IMO, the best reasonable hope is that technology advances to the point where lossy compression isn't really commerically necessary -- we're already there for physical media, and we're probably going to get there for downloadable audio soon , as data transmission rates speed up, storage become cheaper, DSL-type connections replace old modems in homes, and lossless compression algorithms improve their compression ratios. Video will take considerably longer, since the amount of data tthere is so much greater than even 5.1 audio.
 
Back
Top