MT63 2.2 KHz/20/ch/sec = 110 KHz-sec
Should be
MT63 2.2 KHz/20/ch/sec = 110 Hz-sec
2200/20 = 110
Isn't it more meaningful to say that for each 110 Hz of bandwidth, you get one
character? But since MT63 uses interleaving to obtain error free copy and not
ARQ, you might better realize that MT63 uses 2200 Hz for 80 characters or 27.5
Hz to get one character.
To obtain a better BER you have two choices...more FEC or more ARQ (tries) or a
combination of either.
As someone said, comparing ARQ modes with FEC modes is comparing apples and
oranges. Only if you use both ARQ and FEC can you make a valid comparison.
Walt/K5YFW
-----Original Message-----
From: digitalradio@yahoogroups.com
[mailto:[EMAIL PROTECTED] Behalf Of KV9U
Sent: Thursday, January 11, 2007 5:31 PM
To: digitalradio@yahoogroups.com
Subject: Re: [digitalradio] Re: Pactor versus Olivia
I have seen the claim for a good operator able to copy down to about -15
db S/N for CW operation. Some of the digital soundcard modes are
supposed to be able to still have throughput down around -15 depending
upon other factors such as doppler, ISI, etc.
Using Multipsk, and I don't know how accurate this is, I can at least
copy down to at least -10 with CW according to the readout display on
S/N ratio, but even though I am over 60, my hearing is still quite good.
I would not be willing to make too close a comparison between an ARQ and
non-ARQ mode, but the footprint numbers were posited by Rick, KN6KB in
his RFfootprints Powerpoint presentation. In comparing different modes,
he claims that P3 is the superior mode, even to RDFT which he used to
develop the SCAMP mode. He uses a spacing of 200 Hz between signals and
that may or may not be entirely fair since the really narrow modes can
work a lot closer than that from my experience. He comes up with a
calculation of KHz-seconds by dividing the bandwidth by the characters
per second, thus:
MT63 2.2 KHz/20/ch/sec = 110 KHz-sec
PSK31 .25 KHz/4 ch/sec = 62
HF Packet 1.7 KHz/37 ch/sec = 46
Pactor 1 .55 KHz/20ch/sec = 28
RDFT 2.2 KHz/97 ch/sec = 23
Pactor 2 .7 KHz/50/ch/sec = 14
Pactor 3 2.4KHz/225 ch/sec = 10
Of course this assumes everything is flowing perfectly and under most
conditions some of the modes will seriously degrade, particularly HF
Packet. MT-63 and PSK31 would stay at the same rate until they fail to
get through. At that point, the Pactor modes would have slowed down to a
fraction of their high speed and their numbers would not look as good.
Perhaps they would move into the 30 to 60 range?
At only 5 cps for a wide mode like P3 (but narrower at the slower speed
levels at maybe 2000 Hz including a guard band), that would be 2000/5 =
400 KHz-sec and be extremely poor. Even at 20 cps, P3 would not have a
favorable footprint anymore.
By the way, I have often wondered why the B2F binary compression system
used with the Winlink 2000 system has never been used for nearly a 2:1
compression for improved throughput. This could be applied to any
system, including keyboarding.
73,
Rick, KV9U
DuBose Walt Civ AETC CONS/LGCA wrote:
>Well I can't hear a CW signal at a -5 dB SNR. Can you? But I don't think
>minus teens...but that is what some of the newer PSK modes claim.
>
>The only thing that will really kill or stop Pactor is excessive amounts of
>doppler or until the detector can no longer decode the signal.
>
>You are actually making a case for RF footprinting. You want to see what the
>throughput is for bandwidth unit. We could measure it in Hz or KHz. 20 cps
>at 1 KHz is .04 cps/Hz (MT63-1K RAW). 5 cps at 100 Hz is also 4.5 cps is
>.05625 cps/Hz (PSK31). But when decoded, MT63 is almost 100% error free and
>PSK31 can have up to 10% errors.
>
>So it not all just about throughput, its about the overall robustness of the
>mode to include your requirement for how much error free copy you want. You
>are going to get many few errors with Pactor III than PSK31. If you used ARQ
>and something else to to the the BER of PACTOR III, you might find that the
>throughput was less than 5 cps.
>
>Walt/K5YFW
>
>
Announce your digital presence via our DX Cluster telnet://cluster.dynalias.org
Our other groups:
http://groups.yahoo.com/group/dxlist/
http://groups.yahoo.com/group/themixwgroup
http://groups.yahoo.com/group/contesting
http://groups.yahoo.com/group/wnyar
http://groups.yahoo.com/group/Omnibus97
Yahoo! Groups Links
