MPEG Surround standard has been finally released.

[Oki]

Finally the technologies from Coding Technologies, Philips, Fraunhofer IIS and Agere were merged by MPEG for formalizing the new perceptual coding of 5.1 multichannel sound into the, so called, Reference Model 0 (RM0). In July 2005 MPEG and ISO approved this standard. You can find a lot in info about this technology in the official Tutorial on MPEG Surround Audio Coding (http://www.chiariglione.org/mpeg/technologies/mpd-mps/index.htm).

MPEG Surround coding can be viewed as an enhancement of known techniques, such as a more flexible method of joint stereo coding, a generalization of MPEG Parametric Stereo already used in 3GPP rel6 (aacPlus v2) but for 5.1ch or 7.1ch sources or an extension of Binaural Cue Coding (BCC).

MPEG Surround exploits inter-channel differences in level, phase and coherence to capture the spatial image of a multi-channel audio signal relative to a transmitted downmix signal. It encodes each of these cues separately taking account of their distinctive characteristics such that the cues and the transmitted signal can be decoded to synthesize a high quality multi-channel representation allowing higher compression than separate channel coding. The downmix signal is a stereo track for high quality 5.1ch or 7.1ch coding or a mono track for 2ch or medium quality 5.1ch coding.

The MPEG Surround encoding process is agnostic to the compression algorithm used for the transmitted channels; it could be any of a number of high-performance compression algorithms such as MPEG-1 Layer III, MPEG-4 AAC or MPEG-4 High Efficiency AAC, or it could even be PCM.

You can test something very close the MPEG Surround using MP3 as the basic carrier. This technology from Fraunhofer is called MP3 Surround (http://www.iis.fraunhofer.de/amm/download/mp3surround/index.html) and it provides transparent 5.1ch compression at 192Kbps. An encoder and a MP3 Surround player is provided for free for non-comercial use.

This technology is backward and forward compatible and you can play the file or stream with a legacy player without the multichannel part.

Regards,
Oki

[bond]

You can test something very close the MPEG Surround using MP3 as the basic carrier. This technology from Fraunhofer is called MP3 Surround (http://www.iis.fraunhofer.de/amm/download/mp3surround/index.html) and it provides transparent 5.1ch compression at 192Kbps. An encoder and a MP3 Surround player is provided for free for non-comercial use.as i read it this fhg encoder is not compliant to the mpeg standard!!!

afaik divxnetworks licensed the fhg surround encoder (i take it more as a proof of concept encoder) for divx6 and will now push this not compliant stuff onto hardware players

haha, let the mess begin :D


edit: transparent 5.1 at 192kbps, i hope noone believes this...

[Oki]

as i read it this fhg encoder is not compliant to the mpeg standard!!!

afaik divxnetworks licensed the fhg surround encoder (i take it more as a proof of concept encoder) for divx6 and will now push this not compliant stuff onto hardware players

haha, let the mess begin :DDivxnetworks played too quickly. The final standard was not the Fraunhofer IIS/Agere Systems porposal but a mix of their porposal and the Coding Technologies/Philips one. If Divxnetworks does not follow the standard then they will be alone, leaving them in a very difficult position since the approved standar is even better than their licensed technology.

IMHO and based on the listening test graphs described in the official "Report on MPEG Spatial Audio Coding RM0 Listening Tests" (http://www.chiariglione.org/mpeg/working_documents/mpeg-d/sac/RM0-listening-tests.zip) document, the best proposal was the Coding Technologies/Philips one. The conclusions of this report seems more focused on trying to reach a consensus between the contenders than getting the best solution for the MPEG4 Call for Proposal on Spatial Audio Coding.

[bond]

IMHO and based on the listening test graphs described in the official "Report on MPEG Spatial Audio Coding RM0 Listening Tests" (http://www.chiariglione.org/mpeg/working_documents/mpeg-d/sac/RM0-listening-tests.zip) document, the best proposal was the Coding Technologies/Philips one. The conclusions of this report seems more focused on trying to reach a consensus between the contenders than getting the best solution for the MPEG4 Call for Proposal on Spatial Audio Coding.hm do you know if this mpeg surround will be only specified once (mpeg-d) or will every mpeg audio standard (mp2, mp3, aac aso) get their own amendment specifying the exact use of mpeg surround for the specific format?

[Oki]

hm do you know if this mpeg surround will be only specified once (mpeg-d) or will every mpeg audio standard (mp2, mp3, aac aso) get their own amendment specifying the exact use of mpeg surround for the specific format?Time will tell, but it is sure that MP3 Surround is not MPEG Surround compliant because it was a closed standard in Dec'04 and the MPEG Surround reference model 0 was approved in Apr'05 as an improvement of both main proposals.

Nevertheless MP3 Surround could be an adhoc standard if people start using it, but the official MPEG MP3 Surround will not be compatible with FhG MP3 Surround.

IMO HE AAC Surround will be the first audio object derived from MPEG Surround. This combination looks very attractive to DAB and DVB since it can be played in HE AAC players and even in AAC players and can be used with multichannel and stereo receivers.

Regards :)
Oki

[KpeX]

So it would be possible to use this standard to produce a fully standards-compliant backwards-compatible Audio-CD with matrixed 5.1 audio?

[Oki]

So it would be possible to use this standard to produce a fully standards-compliant backwards-compatible Audio-CD with matrixed 5.1 audio?Negative. The Red book standard doesn't allow any kind of multiplexed infomation in the CDDA tracks like tags, video or text. DVD-Audio and Super CD Audio standards already have multichannel and many more features like more bandwith and bit depth per audio sample.

What MPEG meant with that statement is that you can create a MP4 file or stream with MPEG Surround using any other allowed audio format including basic PCM for the downmix signal.

Backward compatibility means that you can decode the mono or stereo downmix signal without the 5.1ch or 7.1ch information with a legacy MPEG decoder ignoring the Spatial Audio Coding multiplexed information.

Forget about matrixed audio, this is completely different, the multichannel information is not sored into the downmixed signal but in a different data stream, giving another level of accuracy compared to Dolby Pro Logic phase encoding and leaving a pure stereo signal without artifacts for legacy playback.

Regards ;)

[mic]

@ Oki
Interesting - :thanks:

Don't see it replacing discreet 6 channels for movies (yet), but that's not the claim either as I read it. OTOH, I think I could see it replacing (or at least competting strongly with) the old niche for Dolby surround (not 5.1 & above).

Again, Thanks

[Oki]

@ mic
Dolby surround was invented in an efford to store a surround channel into a stereo signal but the stereo signal was degraded during the process and the surround effect was almost a joke. So please do not compare MPEG Surround with the very limited matrixed codings. Have you tried the FhG MP3 Surround encoder and player with your favorite multichannel sample?

Parametric Stereo is very close to real stereo and in the same way MPEG Surround is very close to encoding each channel independently providing, at the same time, the highest quality stereo; such that the main market for this technology will be:
- DAB: high quality 2ch or 5.1ch using a single stream.
- DVB: AV Broadcasting quality 2ch and 5.1ch using a medium or low bitrate as low as 48Kbps for the audio track.
- portable AV devices with support for multichannel audio but limited in memory.
- 1CD movie backup using 5.1ch @ 64Kbps HE AAC Surround (less size for the audio means more video quality).

:)

[mic]

So please do not compare MPEG Surround with the very limited matrixed codings.

I'm sorry -- musing out loud I didn't explain properly I guess... I was referring more to commercial markets or market share. You're right as far as quality comparisons go of course. :)

Right now there is a lot of broadcast, cable, & possibly sat TV that still uses old Dolby surround setup, in fact where I'm at, Southeast Michigan, US, DD 5.1 over digital cable is still pretty rare (maybe 1 - 2% of movies).

They want to move to some version of mpeg4, but as of yet I haven't heard of much progress on anything surround audio. My thought was that this might replace the use of DD stereo (98% digital cable) & old Dolby stereo surround (~ 100% analog cable).

As folks move to HDDVD in the future, wouldn't be surprised if this didn't show up there, assuming tools & hardware become available that are cheaper then Dolby licensing. And I do realize that's a stretch, closer to gazing in a crystal ball then something I could back up technically.

That said, think the difficulty in replacing discreet 5.1 would be total channel separation for rear only sound effects, though I'm pretty sure it could get very close.

Thanks

[Kurtnoise]

such that the main market for this technology will be:
- 1CD movie backup using 5.1ch @ 64Kbps HE AAC Surround (less size for the audio means more video quality).
I'm ok with the first sentences but not this considering the dvd boom market. 1CD to store a movie will be completely useless in future.

PS technology is great for streaming too...Think about Dream.

[Oki]

1CD to store a movie will be completely useless in future.I agree with you, in the future it will be impossible to fit a HD movie into 1CD; but now, my friend, most of the backups are still done in 1CD. Backups of 1 or 2GB are not yet usual. And if you can get more video quality maintaining the audio quality and the total size, why not doing it?

:)

[jerrykey]

IMHO and based on the listening test graphs described in the official "Report on MPEG Spatial Audio Coding RM0 Listening Tests" (http://www.chiariglione.org/mpeg/working_documents/mpeg-d/sac/RM0-listening-tests.zip) document, the best proposal was the Coding Technologies/Philips one. The conclusions of this report seems more focused on trying to reach a consensus between the contenders than getting the best solution for the MPEG4 Call for Proposal on Spatial Audio Coding.

That's an interesting document. The Coding technologies/philips proposal seems way better than the frauenhofer one. So coding technologies and philips seem to be the leading companies here. I recently read that coding tech and philips also developed parametric stereo (in aacplus). Also here no word from fraunhofer... Are other parties taking over the lead in audio coding?

[Oki]

That's an interesting document. The Coding technologies/philips proposal seems way better than the frauenhofer one. So coding technologies and philips seem to be the leading companies here. I recently read that coding tech and philips also developed parametric stereo (in aacplus). Also here no word from fraunhofer... Are other parties taking over the lead in audio coding?FhG is not having good luck since CT stuff spun off from FhG. You are right, MPEG PS was developed by CT and Philips, but FhG was also out of other recent MPEG standards like HE-AAC: HQ-SBR was developed by CT, Low Power SBR by Matsushita Electric (Panasonic), NEC and CT, even AAC was developed by Dolby Labs and MLP by Meridian.

It seems that CT is taking over the place where FhG used to be. Dolby will be there as usual and Philips always was a leader in R&D but focused in audio and video, not only in audio. Others like Meridian, Matsushita, Agere, etc... are having good ideas from time to time. FhG is now in this last group.

Regards,
Oki