Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: New MP3 multi-channel method by FhG (Read 9915 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

New MP3 multi-channel method by FhG

from http://www.mi2n.com/press.php3?press_nb=63979
Quote
Fraunhofer Institute for Integrated Circuits IIS presents the brand new MP3 Surround technology as a CeBIT premiere. A next generation of MP3, the MP3 Surround technology, provides a multi-channel listening experience with minimum effort and maximum compatibility.

New MP3 multi-channel method by FhG

Reply #1 –
Seriously, i'm really into aac now for surround.
I thought Fraunhofer were too, being one of the original developers.
Is this gonna just end up as another barrier to aac uptake?

Interestingly reading the descrption seems to say that the surround channels are extrapolated from the main stereo channels using some auxillery data and that this data takes a very low bitrate.

If this is a new technique then couldn't this be applied to AAC (or any other codec) so we could end up with 5.1 HE-AAC files at 64 kbps (or lower) rather than the 128kbps HE-AAC files i'm making right now.

[Edit: Multichannel sound for AAC requires a roughly proportional data increase for every stereo pair added i thought.]

This reminds me though that you can do something very similar with Dolby ProLogic II downmixing and extrapolate apperently decent 5.1 from a regular bitrate stereo file which would mean this technology is not revolutionary at all. However this relies on matrix mixing stuff and doesn't require the extra bits this new method seems to describe.

"a small amount of side information is transmitted characterizing the spatial distribution and attributes of the multi-channel sound image."

Perhaps this new one is better. Can anyone elaborate?

New MP3 multi-channel method by FhG

Reply #2 –
Hard to say if it's better than DPL2 without hearing it. The disadvantage here is that the codec decoder needs support for this side-information in order for the multichannel to work. Where as in DPL2's case any encoder that encodes the phase information correctly can support it, no support in the codec decoder is needed (of course at the end of the line there must be DPL capable audio hardware or software processing which takes advantage of it).
Quote
If this is a new technique then couldn't this be applied to AAC (or any other codec) so we could end up with 5.1 HE-AAC files at 64 kbps (or lower) rather than the 128kbps HE-AAC files i'm making right now.
I'd be very surprised if this wasn't in the making or at least in the planning stage already.. 
There's no public information about this from MPEG, so even if I happened to have some information about this from some source, I couldn't unfortunately say it....
Juha Laaksonheimo

New MP3 multi-channel method by FhG

Reply #3 –
Quote
I'd be very surprised if this wasn't in the making or at least in the planning stage already.. 
There's no public information about this from MPEG, so even if I happened to have some information about this from some source, I couldn't unfortunately say it....

Hmm.. actually I found something public:
http://www.aes.org/technical/cas/
"
113th AES Convention: Workshop on recent MPEG-4 Audio Extensions, Workshop on "Coding of Spatial Audio", 2nd Release of CD-ROM
114th AES Convention: Workshop on MPEG-4 Audio Extensions, Workshop on "Coding of Spatial Audio"
"

Infact there's more that can be found from google about this new planned extension for MPEG-4 audio..
Juha Laaksonheimo

New MP3 multi-channel method by FhG

Reply #4 –
found this

New MP3 multi-channel method by FhG

Reply #5 –
Quote
found this

Yes, MP3 specs indeed allow true multichannel coding. But FhG chose to use its proprietary non-standard 2-channels+extra spatial info. The advantage is that FhG's approach needs clearly less bits, but of course true multichannel would be higher quality.

Also I don't know if FhG's method will be added to the MP3 specs, but MPEG-4 Spatial Audio will be of course standardized when it will (most likely) become another official extension.
Juha Laaksonheimo

New MP3 multi-channel method by FhG

Reply #6 –
hehe, 7.1 audio at 48kbps here i come!

seriously, how many channels can you hear on your MPEG-4 enabled mobile phone  works for me though

joking aside there were 4 representatives from Fraunhofer at the 113th meeting, but there would be anyway right?

[Edit: this reminds me surely bitrate shouldn't be proportional to the number of stereo pairs used as there is clearly similarity and therefore redundancy between all channels used in a sound field. To say you need for instance 128k for stereo and to say you therefore need 5.1/2 * 128 for 5 channel sound is only true if you had to code 2 completely separate stero tracks plus a hypothetical 1.1 track. You could maintain true multichannel sound (as true as any stereo coupling techniques that we use are anyway) at lower bitrates than would be expected if extended proportionally

This further remnds me that i heard someone talking about floating channels or something, whose spatial position can shift and the other channels are derived from, that can improve efficiency or something]

New MP3 multi-channel method by FhG

Reply #7 –
interesting fact about mp3 surround format, from doom9.org :
Quote
We might get more from the MP3 surround format than we want: It has not been confirmed yet, but Fraunhofer and Thomson are also working on DRM technology for the MP3 format. And it might just come as a free "upgrade" hidden inside the multichannel feature.

New MP3 multi-channel method by FhG

Reply #8 –
Hmm, Just when we thought that MP3 had reached stagnation in terms of development, this comes along!!! Looks like the LAME development project is here to stay for a long time!!!! 

New MP3 multi-channel method by FhG

Reply #9 –
Quote
Is this gonna just end up as another barrier to aac uptake?

might sound off topic but i believe so. Just now that AAC is starting to sprout in Sipacks computers we get this. MP3 will just lure away people from other technologies maybe only because its "cool".

now on topic, AAC has far more potential than MP3 in this respect, its just obvious AAC with the same kind of technology would outperform mp3 with a lower bitrate. but ohh well, FhG is all about $$

New MP3 multi-channel method by FhG

Reply #10 –
Quote
Quote
Is this gonna just end up as another barrier to aac uptake?

might sound off topic but i believe so.

I don't think so. It doesn't depend only on FhG, it also depends on players supporting this technology. If they make specs available, they might get some widespread support. If not, they will only be supported in maybe Winamp and a few other players that use FhG's decoding libraries.

New MP3 multi-channel method by FhG

Reply #11 –
Quote
they will only be supported in maybe Winamp and a few other players that use FhG's decoding libraries.

like every Microsoft Operating System after a small update?

New MP3 multi-channel method by FhG

Reply #12 –
Quote
Quote
they will only be supported in maybe Winamp and a few other players that use FhG's decoding libraries.

like every Microsoft Operating System after a small update? 

haha i very much doubt that m$ will do that

mp3 5.1 is a direct competitor to their wma9 pro 5.1 codec
I know, that I know nothing (Socrates)