can the SNR, THD, THD+N, IMD use as measuremet to the audio codec
quality? if could, how to define the quality ?
if not, got other solution to measure audio codec quality in
non-network environment ?
i oso cant find the MOS tool to measure the G729 codec quality... tat
is the problem .
hope a reply from u soon..

wrote in message
oups.com

can the SNR, THD, THD+N, IMD use as measurment to the
audio codec quality?

Not really, because these are simple static measurements, and codecs
generally do well enough with simple static measurements. There's no harm in
doing them, and there's no harm in getting worried about a codec that does
really badly by them.

if could, how to define the quality ?

The gold standard for evaluating codec SQ is blind listening tests based on
snippets of music that are known to make codecs misbehave and sound bad.

thx for ur reply.
however, i cant do the blind listening test becoz it not show any
output of data.
so i need another solution to measure the sound quality. i hope u can
suggest me a software to measure the codec sound quality which without
network.

wrote:
thx for ur reply.
however, i cant do the blind listening test becoz it not show any
output of data.
so i need another solution to measure the sound quality. i hope u can
suggest me a software to measure the codec sound quality which without
network.

There really is no good objective measurement, other than to use a listening
panel. The problem is that the codec is designed around the human perception
and that has to be taken into account.

A codec that works well in one language may work very poorly with another.
I am surprised people can understand tonal languages through the cellphone
at all.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

On 8 Mar 2006 10:20:35 -0500, (Scott Dorsey) wrote:

wrote:
thx for ur reply.
however, i cant do the blind listening test becoz it not show any
output of data.
so i need another solution to measure the sound quality. i hope u can
suggest me a software to measure the codec sound quality which without
network.

There really is no good objective measurement, other than to use a listening
panel. The problem is that the codec is designed around the human perception
and that has to be taken into account.

A codec that works well in one language may work very poorly with another.
I am surprised people can understand tonal languages through the cellphone
at all.
--scott

I have heard from research chaps at Nokia that the German female voice
is the hardest to accommodate in a GSM codec.

d

Pearce Consulting
http://www.pearce.uk.com

wrote in message
oups.com
thx for ur reply.

however, i cant do the blind listening test becoz it not
show any output of data.

The data that blind tests generate takes the form of statistical analysis of
listener judgements.

so i need another solution to measure the sound quality.

http://www.pcavtech.com/techtalk/compare_mp3/index.htm

http://ff123.net/signatures.html