Question regarding multiple flow graphs in tx_voice.py

GNU Radio
1

Hi all,

I am working on tx_voice.py and I got something I don’t quite
understand:
As I know that audio_rx samples from audio source and applies GSM
encoding
with 33bit packages outputs, but there’s no connection between txpath
and
audio_rx. How could 33bits packages be sent to the transmit path? What’s
the
purpose of using multiple flow graphs here? Below is the code FYI.

------------------------------Part of
tx_voice.py-----------------------------------------------
class audio_rx(gr.hier_block2): ### get voice
samples
from audio source input
def init(self, audio_input_dev):
gr.hier_block2.init(self, “audio_rx”,
gr.io_signature(0, 0, 0), # Input signature
gr.io_signature(0, 0, 0)) # Output signature
sample_rate = 8000 ###Change to 44100 next time
src = audio.source(sample_rate, audio_input_dev) ### samples
from
audio source
…
self.connect(src, src_scale, f2s, voice_coder, packet_sink)

def get_encoded_voice_packet(self):
    return self.packets_from_encoder.delete_head()

class my_top_block(gr.top_block):

def __init__(self, modulator_class, options):
    gr.top_block.__init__(self)
    self.txpath =

usrp_transmit_path.usrp_transmit_path(modulator_class,
options)
self.audio_rx = audio_rx(options.audio_input)
self.connect(self.txpath) ###########No connection
###########
self.connect(self.audio_rx) ############# Here!
###########

…

Thanks for all the help,

Sincerely,

Yuan