MPEG has recognizes a “low-power” decoder variant for High-Efficiency AAC. This low-power decoding method requires 40% less processing power and decodes HE AAC bitstreams with only a slightly reduced audio quality.

Who can tell me something about the "low-power" decoder?

Well, it's important for hardware implementations of AAC+, what else would you like to know? I don't think that there is already an application publicly available yet, but it's interesting that they have developed this in parallel to the software codec so it's "in store" if a hardware manufacturer wants to build an AAC+ compatible consumer player. My guess would be Philips again as the first company with such a thingy, maybe RCA, because both have mp3PRO hardware already that has upgradable firmware. And the minimum system requirements for AAC+ (CPU speed, memory) are almost the same for both formats.

By the way, there are of course AAC+ hardware implementations for professional usage like a receiver for Digital Radio Mondiale, Mayah's all-in-one realtime encoder for several formats or a very interesting portable device from Telos for reporters that need to plug into a telephone network from anywhere in the world to send their live reports to the home station.

But it will take some time until a consumer player will be available, don't know how long...

I want to know more about SBR technical detail. What documents shall I have?
(I have two iso document N5203-Text of ISO/IEC 14496-3:2001/FPDAM 1, Bandwidth Extension and N4611.)

sorry... MPC CPU and resources usage is lower than MP3...

And How does MP3 pro cpu and resources usage compare to MPC
And AAC + compare to MP3??

Is this low power decoder only avalible on Hardware?? Or could a software does that same thing?? ( I .e have a low power software decoder on a PC ? )

Well, if you've got the draft version of the specification already, what else do you need? I already mentioned in another thread that the ISO/IEC International Standard status for AAC+ is planned for May 2003, and it's unlikely that they will hand out any final documents or even reference software before this date. You could also have a look at the different Ad-hoc Groups (AHG) that were installed after the last MPEG meeting. They are the ones that do most of the work between meetings, i.e. formulating documents etc., so this would give you an idea about what's going on right now:

http://mpeg.telecomitalialab.com/meetings.htm

Or are you looking for something shorter and easier to understand than the specs? Then the new White Paper about MPEG-4 High-Efficiency AAC from M4IF would be a good source, and the Wiki of Audiocoding.com of course...

Thank you hans-jürgen !

So what?



I don't have the exact values at hand right now, but both SBR codecs use of course more resources than the ones without. For example I can't play a mp3PRO file on my 5x86/133 MHz PC at all, because it stutters all the way through. But it plays MP3 and normal AAC/MP4 files in Winamp 2.81 with no problems (if I use John33's compiles). So this might give you an idea perhaps...



At the moment nothing is available at all, and I guess we can only wait what Menno will come up with in this quarter. Of course this low-power decoder could be implemented in a software application, but this is probably not the main purpose for it.

Good news is that future ISO reference software will have a low power SBR implementation (you can turn it on by #define switch during compile process)

There is no point of using low power HE AAC decoder on a PC platform anyway, but it will significantly accelerate adoption of the HE AAC on a hardware platforms.

LP AAC basically uses real-valued QMF filterbank, instead of complex QMF filterbank used in "high power" HE AAC decoder, and some additional filters to cope with aliasing.