Hello Gullik, Yes the fft is used as part of the frequency offset estimation algorithm. Frequency offset estimation is necessary, as without knowing the frequency offset the demod cannot recover the data.
Yes the issue of 48kHz is to overcome flaws in sound cards. Otherwise, 8 kHz sampling is just fine. I tend to use 8kHz sampling for all of my software simulation of the modem. Using use the I channel you would get two modem signals but if you tine the demod to the upper one it will work just fine. Cheers, David On Tue, 2012-06-12 at 12:36 +0200, Gullik Webjörn wrote: > Am I correct in that the only use of fft in fdmdv is the > identification of the pilot (to autotune), except > for the get waterfall, that is nice, but not necessary? > > Is it correct that the whole issue of sampling at 48 khz is > just to overcome flaws in sound cards, or is there any other > reason I would not want to use a "reasonably good" 8 ksamples > A/D, such as filtering, aliases? > > I am attempting to build a "fdmdv" test generator, that > will initially produce DSB, using only the "I" channel. > Transmit just generates and sums the carriers, n'est pas? > > Regards, > > Gullik /SM6FBD > ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Freetel-codec2 mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/freetel-codec2
