> There are about 10.76 Msymbols/sec in an ATSC signal, so you are going > to have Nyquist problems if you try to represent them in 8M complex > samples/sec. You'll need at least 10.76 complex MS/s. (Indeed you > may notice that we were skating inside the edge by using 20 > non-complex MS/s on the old hardware.) If we had a faster USB bus, > the USRP could sample at a nice high rate, like 32 Msamples/sec, then
So... this probably stems from my shoddy understanding of signals and/or information theory, but here is the question: I have 8M complex S/s. This means that I *should* be able to do some sort of quadrature demodulation or something along those lines to get a full 8MHz of spectrum, no? Is this logic wrong? The FFT displays the signal from -4 to +4 MHz... I was under the impression that the ATSC signal was only 6MHz wide (if not a little less), so why do you need 21.52MS/s (real) to sample a 6MHz signal? Is this due to the quantization of the measurements, and I would normally be right with infinitely precise samples, but in the world of reality we live in, we just have to oversample? > The pcHDTV card uses a chip that's specialized and engineered to do > this job very well in a variety of environments. By contrast, you are > only about the fourth person to ever run the ATSC input code. It's a > "dancing bear", which took more than a year of solid engineering just > to get it dancing at all. It hasn't been optimized, either for CPU > usage or for non-perfect signal reception. Point well taken. There's a lot of multipath going on where I'm receiving this.... might have to try for something different. I do have a different antenna source; I'll try playing with it. > I'll attach my favorite point script below. I thought it was in the > source trees, but now I can't find it except on my hard drive. > Yeah, the wiki referred to it, but it wasn't in the gnuradio-0.9 tarball. Thanks for the suggestions... I'll give them a whirl and see if I can get anywhere. --- Ilia Mirkin [EMAIL PROTECTED] _______________________________________________ Discuss-gnuradio mailing list [email protected] http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
