wonder how this "japanese mp3 killer" siren14 codec compares to AAC-HE and ogg vorbis at lower bitrates... apparently one can get the source code and it is royality free, might be
useful for IP telephony and low latency streaming too...

http://www.cmpnetasia.com/ViewArt.cfm?Arti...tid=6&subcat=61
http://www.polycom.com/company_info/1,1412...2-10713,00.html
http://www.polycom.com/company_info/1,,ce-...2-10713,FF.html

download:
http://www.polycom.com/pw_files/Siren14PCExecutable.zip

I remember another japanese MP3 killer some years ago

Did it start with a "V" and end in "F"?

That too, but click the devil...

Reading the discription on Reallyrarewares,i think Keiji Mafuka implimented
the first reminence of DRM. It actually sounds like a DRM scheme.

forgive me using the "killer" argument. might be of some interest
for the hydrogene community if it's a fake or not.

as far as it is described you can get it up to a resolution of 48khz (ultrawideband?)...
did someone find the source code and compile it for linux?

there is another link in the asterisk list..
http://lists.digium.com/pipermail/asterisk...ust/014464.html

there is little probality that would the ITU standardize a fake codec
http://www.itu.int/itudoc/itu-t/aap/sg16aap/history/g722.1/

Oh, I never meant to say Siren is fake. It's just that I get tired of the name "MP3 killer" being thrown around as a buzzword. None of the codecs so far introduced as "MP3 killers" came even close to affecting its popularity.

...and i think Siren14 is not meant to be an mp3 killer (quality-wise). But it's ment to be something of a similar quality level but faster to encode and/or decode. If it really is I think Siren14 has its place (ie embedded devices, phones, ...)

It's probably something like AAC+LTP in a light version with a quick and simple psychoacoustic model.

Sebi

I encoded a few tracks and I think it does pretty well.
It is obviously tuned to retain speech and make it intelligble.
I'd say AMR-WB+ is clearly better at 24kbps, especially for music. BUT this is low latency, low dsp power and has a free to use licence

then google talk could evaluate it..



3GPP AMR-WB+ is coming from the other side of the mobile phone industry, with aacplus, DRM and closed source licenses.

so siren14 is mainly positioned against ogg speex? considering the possibility to listen
to a webcast via a normal mobile phone AND a voip phone, such codecs might allow
quite a good interoperability if combined with the right protocols (sip...)

in the moment speech codecs allow a quality of mp3 128kbps or mpeg2 192kbps,
with an easy implementation in existing dsp environments through firmware updates and
easy licensing scheme, it is gonna get interesting, and google could make it a quasi
standard.

imagine listening to radio stations or podcasts via google talk with your mobile phone.