Hello Developers,

I'm working on DRM and DRM+ for quite a while now and a few years ago I've written a software modulator called Spark (www.drm-sender.de). At the moment, the software uses the FAAC encoder library which delivers reasonable mono quality at bitrates around 24kBit/s. A license for the optional Dolby AAC+ encoder is way to expensive for the "normal" user, who just want to play around. Therefore, and because of people asking me about AAC+ support or rather the lack thereof, I thought about interfacing either the 3GPP source code or the Nero encoder. The problem with both is the transform length of 1024 samples per frame instead of 960. However, this problem could be solved by resampling the audio signal before encoding. Now, there are two problems left:
a) When using the 3GPP source code, people have to compile their own libraries and it does only support a few bitrates.
b) When using the Nero executable I need to have a disk-swap wav file and must extract the AAC frames out of the MP4 container, and I have no clue how the format is made up.

What do you think is the best solution?
Have you got experience in getting the AAC frames out of the MP4 container format?
Would you like to support me? :-)

Thanks,
Michael

I don't think you can just extract frames from a 1024 samples per frame encoder and make that into 960 samples per frame packets?

Can't you prepare your software to use the aac dll from used in winamp?

Edcast (formerly oddcast) did the same, so he doesn't distribute, but makes it possible to use it (by copying or referring to the winamp version already installed on your system).

Regards,

Arkadas

Not unless it supports Digital Radio Mondiale.

Do you use aac he vers 1 or vers 2?
If you want minimum band(for example 24kbps) maybe is better use vers 2 and stereo 44100..
Am I wrong?

No, I would resample the audio signal before the encoding.


It's HE-AACv2 at 24 kHz or 12 kHz for bitrates around 20 kbps.


Is there an SDK for the library out there? Does it support the 960 transform length? If you have further information, please get in touch.

What about the SBR support for FAAC? Is there actually anything going on in the open source domain?

That doesn't help, you will still get 1024 samples per frame LC AAC files.

What you need is AAC files encoded with the Error Resilient tools and 960 sample framesizes. It might be that some DRM decoders even decode files without the ER tools, but they are officially mandatory for DRM.

why do you use 24-12khz sampling and mono?
bacause an user of doom9 has explained me that it's better user 44100 stereo because aac he vers 2 is parametric stereo...
I don't find original post of doom9, but this:
http://www.hydrogenaudio.org/forums/index....showtopic=59284

That's the only samplerates that DRM allows (and 48-24kHz with HE AAC of course).

ok..ma ideally if you can use other sampling it's better use 44100 right?
I tried shoutcast for winamp and at low bitrate I can use with aac-he vers 2 44100 stereo (default)

What you are saying makes no sense.

The post you linked to gives no reason to use 44.1kHz with DRM. As menno said, it's not even allowed to do so.

that post is only a resume of a detailed post of doom9(that I don't found)..
in that post is explained also with many graphs these values

Try to get more info from Ed at http://www.oddsock.org/ for using the aac encoder from winamp!

regards,

Arkadas

Please read TOS 8.

That's still not going to create DRM streams

tos8?

comrice,try shoutcast for winamp