OK, time for another nooby question, but i didn't find this in the FAQ. What's the difference between MP3 Layer I, II, III? LAME seems to pick whichever one it likes, i didn't find a parameter to tell it which one to pick.
Thanks,
glauber
From what I know LAME only encodes Layer 3, which is what MP3 is based off. Layer 2 was a subband codec and Layer 1 was ???? (enlighten me ) whereas Layer 3 was a transform codec. There is a website that describes all of this but I can't remember the name... you can always try Gabriel's MP3 Site (http://www.mp3-tech.org/).
Regards
AgentMil
Darnit! Sorry, i meant to ask about MPEG-1, -2 and -2.5. Sorry, fingers faster than brain. I'll check the URL you gave.
g
Layer 1 was ???? (enlighten me )
Subband too.
MP1 is also known as PASQ: Precision Adaptive Subband Coding
whereas Layer 3 was a transform codec.
Actually, it was more or less of a "hybrid" between transform and subband (the infamous hybrid filterbank...). Examples of purely transform codecs are AAC and Vorbis.
Sorry, i meant to ask about MPEG-1, -2 and -2.5.
The frequencies supported are different for each MPEG version:
MPEG1: 48, 44.1 and 32kHz
MPEG2: 24, 22.05 and 16kHz
MPEG2.5: 12, 11 and 8kHz.
Layer 1 was ???? (enlighten me )
Subband too.
MP1 is also known as PASQ: Precision Adaptive Subband Coding
whereas Layer 3 was a transform codec.
Actually, it was more or less of a "hybrid" between transform and subband (the infamous hybrid filterbank...). Examples of purely transform codecs are AAC and Vorbis.
Sorry, i meant to ask about MPEG-1, -2 and -2.5.
The frequencies supported are different for each MPEG version:
MPEG1: 48, 44.1 and 32kHz
MPEG2: 24, 22.05 and 16kHz
MPEG2.5: 12, 11 and 8kHz.
Heh... thanks for fixing that one up... I thought that as well but wasn't too sure, but I do recall it now from the discussion about SCF "errors" with MPC and why MP3 don't have them.
Regards
AgentMil
One more to add for the difference
MPEG 1 - 11172 part 3 (Pretty Old 1992)
MPEG 2- 13818 part 3 (1996)
MPEG 2.5 non ISO Standard (Fraunhoffer's)
One question. MP3 is so called polyphase hybrid filter bank which has subband filter (output still in time domain) followed by frequency mapping my a MDCT bank.
Why in the new edition (AAC ofcourse) the sub band was taken off and only MDCT retained? (Pure transform coder? ) Is it due to the lessen the complexity of the codec or is there any other reason.
ISO standards use two FFT running in lengths of 1024 and 256 for threshold calculation as well as for window switching. I have heard some encoders doesn't use FFT and instead use MDCT itself for these calculation. Is this a reason for the migration from Hybrid to pure transform coder?
And why is the filter bank called "poly phase" ?