Helmut, David, Thanks for the feedback, it aligns pretty much with my view, and the "filtering power" in a good cpu is probably the dominating factor.
The implication of this is that a fairly simple receiver would be sufficient, and provided it delivers quadrature outputs, all information is available. Gain need not be large, in fact keeping gain down ( but still within A/D resolution) and having good phase balance, would perhaps be a good recipy. Studying a waterfall of a weak signal, it can often be seen that only a single carrier is erased when subject to QSB. WHICH carrier i affected varies with time, probably due to frequency dependence of signal cancellation. (path length varying in athmosphere) This can be seen as "diagonal" lines of carrier erasure. This leads to thinking that IF this could be addressed, with some more elaborate fec scheme it would add to weak signal robustness. Anyway, I have had good results decoding freedv at indicated SNR 4 and up, so this already works very well. Just a note: When using external audio input, my transceiver does NOT disconnect the microphone, so I got this pointed out by a fellow ham, I have to disconnect the mike, or else I get dual dv + SSB. I have heard other stations suffering from this problem as well, so unplug the mike or add a mute switch. Worth bringing to attention... Regards and happy QSOing, Gullik / SM6FBD On 01/21/2015 08:52 PM, David Rowe wrote: > The demodulator has it's own optimal filtering so the SSB rx filtering > won't have any impact. > > Flat tx filtering is a good idea, so that all carriers have equal power. > > - David > > On 21/01/15 23:01, Gullik Webjörn wrote: >> Dear list, >> >> Currently the typical freedv / codec2 gear is a SSB transceiver and a >> codec2/fdmdv modem. >> Observing the signal on the waterfall, I can see that a fdmdv signal is >> about half the bandwidth >> of a SSB signal. Thus, I was thinking about changing the filter in the >> transceiver, to gain an extra >> dB or so by eliminating the difference in noise bandwidth. >> >> However, maybe this is NOT the way to go, maybe the signal processing in >> the fdmdv modem is so >> much superior due to cpu power, that THAT would be the area to focus on. >> >> We have discussed I/Q modulation/demodulation, and that even though this >> is easily implemented, it >> has not been done (but maybe tested??) yet. >> >> Is it your belief that a "dedicated" transceiver would significantly >> improve codec2/freedv operation? >> >> Where is the biggest mismatch? Crystal filters ( passband shape / group >> delay) , AGC, general amplitude/phase >> in the signal chain (that is optimised for CW/SSB/RTTY)?? >> >> Gain / AGC ( passband mismatch)?? >> >> In short, how would a "near optimum" transceiver for this mode look >> like, what can be eliminated, >> what should be enhanced? >> >> Would a signal chain with independent carrier processing of the 16/17 >> carriers gain anyting, I am here >> thinking of individual carrier erasure due to multipath, or are we close >> to optimal here? >> >> Your thoughts... >> >> Gullik / SM6FBD >> >> PS >> Saw the video from linux-dev NZ, nice... >> >> >> >> ------------------------------------------------------------------------------ >> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. >> GigeNET is offering a free month of service with a new server in Ashburn. >> Choose from 2 high performing configs, both with 100TB of bandwidth. >> Higher redundancy.Lower latency.Increased capacity.Completely compliant. >> http://p.sf.net/sfu/gigenet >> _______________________________________________ >> Freetel-codec2 mailing list >> [email protected] >> https://lists.sourceforge.net/lists/listinfo/freetel-codec2 >> > ------------------------------------------------------------------------------ > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > GigeNET is offering a free month of service with a new server in Ashburn. > Choose from 2 high performing configs, both with 100TB of bandwidth. > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > http://p.sf.net/sfu/gigenet > _______________________________________________ > Freetel-codec2 mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/freetel-codec2 ------------------------------------------------------------------------------ New Year. New Location. New Benefits. New Data Center in Ashburn, VA. GigeNET is offering a free month of service with a new server in Ashburn. Choose from 2 high performing configs, both with 100TB of bandwidth. Higher redundancy.Lower latency.Increased capacity.Completely compliant. http://p.sf.net/sfu/gigenet _______________________________________________ Freetel-codec2 mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/freetel-codec2
