On Sat, Apr 17, 2010 at 6:59 AM, Pino Zollo <[email protected]> wrote:

> Hi,
>
> I made two QSO's in Contestia....
>
> It works fine but the water-fall was not showing anything.
>
> Is this because Contestia is so sensible to decode extremely weak signals
> or
> is a problem ?
> The water fall was showing other PSK31 signal in the band.
>

Contestia and Olivia are virtually identical CODECs. Their performance is
going to be within one or two dB of each other so what you expect from
Olivia you can expect from Contestia.

Contestia uses a small character set so it needs fewer bits transferred for
the same message. This also means that you don't get lower case letters and
many punctuation symbols you can send with Olivia. Also Contestia has a
smaller block size and shorter interleave. This means less delay but less
resistance to long fading (QSB). If you increase the interleave to provide
QSB resistance you can expect longer delays (latency) before copy starts.

In this same family of protocols you will find MFSK16, DominoEX, and Thor.
Olivia, Contestia, and MFSK16 all share the same base modulation -- m-ary
frequency shift keying. The differences are in the alphabet/character-set
they use, the symbol (baud) rate, and the robustness of the forward error
correction (FEC). If you set the #tones (always a power of 2, e.g. 2, 4, 8,
16, 32), symbol rate (determines tone spacing), and level of FEC the same on
these protocols, they will provide the same speed and weak-signal
performance.

(Soapbox -- We would be better served by having just one of these modes and
then allow selection of the tones, baud rate, convolution polynomial, block
size, interleave factor, and Viterbi decoder settings and then transmit that
in something equivalent to an RSID header.)

DominoEX and Thor use a novel and, frankly, better modulation scheme than
the pure MFSK used in MFSK16/Olivia/Contestia. All other things being equal,
if you were to put DominoEX with FEC or Thor up against on the three other
above but ensure that you are using like parameters for baud rate,
convolutional coding, and FEC, DominoEX and Thor would be easier to tune and
more resilient in the presence of doppler. In other words, they would
flat-out work better.

I apologize for what is a really long answer to your question. I hope it
helps to understand the differences and to set reasonable expectations for
performance.

Hmm, it would be great if there were a table that listed the basic parameter
differences between modes so we could pick the one that best met conditions
at the time so as to provide maximum throughput at 100% copy. Frankly, most
of the time Olivia is overkill but it also provides the most robust FEC and
interleave so it will get through under the worst conditions when the other
modes are failing.

-- 
73 de Brian, WB6RQN/J79BPL
_______________________________________________
fldigi-alpha mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/fldigi-alpha

Reply via email to