So after several sessions of comparing new code to old code, I found the
problem. The root raised cosine filter was not being inserted into the
graph
before the fpll. The patch is pasted below. I successfully watched some
video received from a local station (KFXL). I didn’t get any audio, but
I
only captured for a couple seconds so I’m not sure if that is related or
not.
Now that decoding is working again, I would like to work on
improving/optimizing it where I can. Anyone have any pointers for places
they know are current bottlenecks?
-Bryce
— a/gr-atsc/src/python/fpll.py
+++ b/gr-atsc/src/python/fpll.py
@@ -70,7 +70,7 @@ def main():
out = gr.file_sink(gr.sizeof_float,"/tmp/atsc_pipe_3")
# out =
gr.file_sink(gr.sizeof_float,"/mnt/sata/atsc_data_float")
-
tb.connect(u, fpll, lp_filter)
-
tb.connect(u, rrc, fpll, lp_filter) tb.connect(lp_filter, iir) tb.connect(lp_filter, (remove_dc,0)) tb.connect(iir, (remove_dc,1))