Help - Search - Members - Calendar
Full Version: WHERE is the implementation for synthesis filter bank of ISO-11172 com
Hydrogenaudio Forums > Hydrogenaudio Forum > Scientific/R&D Discussion
ieangel
the standard specfified a medium complexity implementation for the synthesis filter bank. The specification is shown below
IPB Image

now, I want to seperate the 32 subbands out, and I will check the frequency component of every subband.
but i have no idea how the proposed flow chart come from.


So, can anybody show me any tips?


Best Regards
SebastianG
QUOTE(ieangel @ Apr 21 2008, 14:24) *

the standard specfified a medium complexity implementation for the synthesis filter bank. The specification is shown below [...]
now, I want to seperate the 32 subbands out, and I will check the frequency component of every subband.

Ok, then, what's the problem?

QUOTE(ieangel @ Apr 21 2008, 14:24) *

but i have no idea how the proposed flow chart come from.

Does it matter where it came from?

QUOTE(ieangel @ Apr 21 2008, 14:24) *

So, can anybody show me any tips?

Try rephrasing your question.

Cheers,
SG
ieangel
I am sorry, my question is:
in the alogrithms of synthesis filter bank, the synthesised result should be the sum of 32 subbands: result=sb0+sb1+....+sb31;
IPB Image
but in the proposed implementation, i can't find the 32 subbands.

In fact, I want to calculate the spectrum of each subband, e.g. the spectrum of sb0, sb1....sb31. But I can not find the signal of each subband in the proposed algorithms of ISO-11172.

is it clear?
Gabriel
They come from the exact same PQMF filter as Layer II.
menno
Don't you want to look at the analysis filterbank if you want to create the subbands?
SebastianG
QUOTE(ieangel @ Apr 23 2008, 07:14) *

[...] But I can not find the signal of each subband in the proposed algorithms of ISO-11172.

I still fail to see what you're asking for. Didn't you already check the effect of zeroing some subbands prior synthesis?

You might want to tell us what your goal is. I'm guessing here that you're trying to exploit the filterbank to do cheap lowpass filtering -- either at the encoder side (like LAME does) or at the decoder side (for whatever reason).

Cheers,
SG
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2008 Invision Power Services, Inc.