VHFcontesting
[Top] [All Lists]

[VHFcontesting] Is There a Codec In Your Data Stream? Should You Care?

To: vhfcontesting@contesting.com
Subject: [VHFcontesting] Is There a Codec In Your Data Stream? Should You Care?
From: Roger Rehr W3SZ <w3sz73@gmail.com>
Date: Fri, 24 Feb 2017 17:13:48 -0500
List-post: <vhfcontesting@contesting.com">mailto:vhfcontesting@contesting.com>
Hi All,

With the broader use of SDRs and "Network Radios" and remote operation,
it has become increasingly common for a codec to be in the path of our
received and transmitted signals.

In some cases those using commercial radios may have a codec included in
their TR/Rx path of which they are totally unaware.

I use my own software for all terrestrial and some EME operation, and
when using this software my signal path always includes the use of a
codec with both received and transmitted signals.

It does not need saying that the operation of a codec potentially
changes a digital signal in ways that may be unapparent to the casual
user, but which may markedly degrade the performance of our beloved
digital modes.

The severity of this degradation may be a function not only of the codec
and its parameters such as bandwidth, bitrate, and type (VOIP or AUDIO,
for example) but may also depend upon the particular digital mode being
used and the quality of both the RF and the digital signal paths for the
signal in question.

Earlier this year I did some preliminary studies and found that under
some circumstances the codec that I use in my software, OPUS, could
result in 50% or more of possible signal decodes being missed when the
default codec parameters with automatic dynamic parameter optimization
were used.  https://opus-codec.org/

By altering the codec parameters I was able to markedly improve the
digital signal decoding performance of the path.

I had started to collect more data on this issue and I had planned to
present all of this at either one of the spring VHF/UHF/Microwave
conferences here in the US, or at one of the fall conferences.

However, I have been tasked with other talks both for this spring and
for the fall, and so it now looks as though I will NOT be able to resume
formal work on this until the last quarter of this year.

The purpose of this email is to warn folks using the digital modes to
CONSIDER the possibility that  there might be a codec in their signal
path and that this codec might be degrading the digital signal decoding
performance of their system.  This might be of particular importance for
you if your digital signal decoding performance with that "newfangled
radio" that you just got just doesn't seem to be up to what your old
radio could do, or if you are operating remotely and find that the
digital signal decoding performance of your station when you operate
remotely is inferior to your station performance when not operating
remotely.

I WILL complete my work on this subject but it will likely NOT be before
the last quarter of this year that I can resume it.  So it is likely
that I will not have anything ready for presentation before early next year.

If you have a codec in your data stream and if you are prepared to
discuss issues that you think you may be having with it with some degree
of technical facility, by all means drop me a note.

My software is a test bed for my use/amusement and is not directly
suitable for use by others.  It is described at the URL below.  There is
only a little discussion of the OPUS codec there, but I do indicate
there what codec parameters I am currently using for terrestrial digital
work, based on my preliminary observations.  So if you are writing your
own software as well, those comments and the pertinent portions of the
source code that can be downloaded from there may help you get started:
http://www.nitehawk.com/w3sz/CSharpsdrclientANDserverVersion2pt0.html

73,

Roger Rehr
W3SZ
_______________________________________________
VHFcontesting mailing list
VHFcontesting@contesting.com
http://lists.contesting.com/mailman/listinfo/vhfcontesting

<Prev in Thread] Current Thread [Next in Thread>
  • [VHFcontesting] Is There a Codec In Your Data Stream? Should You Care?, Roger Rehr W3SZ <=