[M4IF Technotes] What's low complexity about A

[Panos Kudumakis]

Ivan 
Thanks for the clarification
Panos
____________________Reply Separator____________________
Subject:    Re: Re[2]: [M4IF Technotes] What's low complexity about AAC 
Author: "Ivan Dimkovic" <admin   aac-tech.com>
Date:       09/11/01 11:04
Dear Panos,
> Prediction for sure impact on quality but talking about MPEG quality is
nosense
> Thats because MPEG defines the bitstream and how to decode it in order to
be
> compliant  The encoders (which make the difference in the quality) in the
> reference software are non-optimised Even the decoders can further
optimised at
> least in terms of complexity The companies in MPEG contributed to code
they do
> not give away their competitive advantage that is their optimised codecs
and
> that open up competition in the market
>

When I talked about small quality difference between AAC Main and AAC LC I
was talking about encoders optimized for quality, used in MPEG listening
tests, and not reference software encoder (ISO refsoft).
> For more info check out "MPEG-4, why use it ?" publicly available in MPEG
site
>

For futher information please check document N2006 "Report on MPEG-2 AAC
Stereo Verification Tests", where AAC Main 128 scored slightly less
than -0.2 (difference from original) while AAC LC 128 scored slightly more
than -0.2 - both of them are in the range of 'excellent' quality. Encoders
used in these tests were optimized for maximum possible quality. In the most
AES documents the overall conclusion is that AAC Main does not offer much
better quality for a performance loss.
Also,
"Considering the results of the MPEG-2 AAC Stereo Verification Test, it is
clear that there is only a marginal difference in performance between AAC
Main and AAC LC Profile at the tested bitrates. On the other hand there is a
significant increase of complexity (both required RAM and instructions) for
the Main profile compared to the LC profile"
(quoted from "Implementing ISO/MPEG-2 Advanced Audio Coding in Realtime on a
Fixed Point DSP", Johannes Hilpert,Michael Braun,Manfred Lutzky,Stefan
Geyersberger,and Rainer Buchta Fraunhofer Institute for Integrated Circuits,
AES article 4822)
Regards,
*************************************************
Ivan Dimkovic, Technical Manager
PsyTEL Research
Multimedia Coding Solutions
Belgrade Yugoslavia
phone:  +381 64 11 40 600
fax:    +381 11 32 25 275
email:  dim   psytel-research.co.yu
www:  http://www.psytel-research.co.yu
*************************************************
____________________________
BN3 Hosted Customer Service Solution, basic service FREE.
CRM enable your web site in 5 minutes! http://www.bn3.com
_______________________________________________
Technotes mailing list
Technotes   lists.m4if.org
http://lists.m4if.org/mailman/listinfo/technotes
**********************************************************************
This communication contains information which is confidential
and may also be privileged.  It is for the exclusive use of the
intended recipient(s).  Please note that any distribution, copying
or use of this communication, or the information in it, is strictly
prohibited.  If you have received this communication in error,
please notify us by email or by telephone (+44 20 8848 9779)
and then delete the email and any copies of it.
This communication is from Central Research Laboratories Ltd.,
whose principal office is at Dawley Road, Hayes, Middlesex,
England.
This footnote also confirms that this email message has been
checked for the presence of computer viruses.
**********************************************************************