Newsgroups: comp.speech
Path: pavo.csi.cam.ac.uk!doc.ic.ac.uk!pipex!howland.reston.ans.net!math.ohio-state.edu!cs.utexas.edu!asuvax!ncar!uchinews!cs.umd.edu!afterlife!jpcampb
From: jpcampb@afterlife.ncsc.mil (Joe Campbell)
Subject: Re: How to compare different coding systems?
Message-ID: <1993Sep25.021345.9042@afterlife.ncsc.mil>
Organization: The Great Beyond
References: <sliawati.748902125@bluering.cowan.edu.au>
Date: Sat, 25 Sep 1993 02:13:45 GMT
Lines: 25

In article <sliawati.748902125@bluering.cowan.edu.au>,
S.LIAWATIMENA <sliawati@scorpion.cowan.edu.au> wrote:
>  I would appreciate if somebody could tell me what are the characteristics
>that have to take account when comparing different coding systems? (e.g. 
>LPC, Vector Quantization, CELP, Trellis, etc.).
>  Thank you very much in advance.
>
>S Liawatimena                    
>S.Liawatimena@bluering.cowan.edu.au 

There are may "characteristics," including: speech quality, speech
intelligibility, complexity (RAM, ROM, processing power, floating pt,
power dissapation, ...), coding delay, robustness to channel errors, 
transcoding/tandeming performance with other coders, etc.
Meaningful measures of speech quality and intelligibility, especially for
mid- to low-rate coders, are usually determined subjectively by responses
from a group of listeners.  DRT & DAM tests (see papers by Voiers) and
MOS/DMOS (someone please chime in with *the* reference) are popular tests.

Joe
-- 
.............................................................................
; Dr. Campbell  N3JBC  jpcampb@afterlife.ncsc.mil  74040.305@compuserve.com ;
; My opinions are mine!   Happiness = Reality - Expectations, Click & Clack ;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
