hello, I found a simple way to interface sound applications with USRP
and gnuradio; maybe it is already widely used, but I guess it is worth
mentioning it - this method work provided that the application is able
to get samples from file, my experience has been done with drm, that
called with the option --fileio is able to get 16 bit integers from a
file.
I created a tmpfs on my linux by adding this mount point to /etc/fstab:
ramfs /mnt/ramfs tmpfs defaults 0 0
and then i issued the command “mount ramfs”
then I created a pipe with the command “mkfifo /mnt/ramfs/drm.raw”
The game is almost done - start drm with
“drm --fileio /mnt/ramfs/drm.raw”
and start your python code providing a file sink with the same name as
the pipe you created, in my case:
…omissis…
out = gr.file_sink(gr.sizeof_short,"/mnt/ramfs/drm.raw")
…omissis…
Maybe it’s not going to work nicely for high sample rates, but it works
lovely for this case as well as for gmfsk started in “demo” mode with
the command:
cat <your_pipe_in_tmpfs> | gmfsk --testmode
MC