17 Oct
2017
17 Oct
'17
7:51 p.m.
I have an older system working fine on the master branch and I'm currently
trying to set up a new VM running the max branch. The goal is to send audio
from two talkgroups to darkice. The problem is that I'm not getting any
audio out of rx.py.
I know the darkice config works, because I can do this and hear pink noise
on the stream:
speaker-test -r 8000 -D hw:Loopback,0,0
But starting rx.py with the same audio device like this, I never hear
anything:
./rx.py --args rtl=0 -N LNA:42 -f 774081250 -S 1320000 -v 5 -2 -T
ss911_tacoma.tsv -O hw:Loopback,0,0
The receiver seems to be working. I can see the name of the talkgroup I'm
interested in at the bottom of the screen while it's active, I just don't
hear anything on the stream:
"""
NAC 0x1e1 WACN 0xbee00 SYSID 0x1e7 774.431250/804.431250 tsbks 2390
voice frequency 773.431250 tgid(s) 40306 40107 0.3s ago count 77
voice frequency 774.681250 tgid(s) 59197 40106 0.3s ago count 166
Frequency 774.431250 Talkgroup ID 40106
PCSO East
"""
Is there something obvious I'm doing wrong here?
P.S. Is there a better way to pipe audio to darkice than snd-aloop?
Tom
17 Oct
17 Oct
11:34 p.m.
More command line options are needed:
-O Loopback:0,0 -V -2 -U
Make sure you see an entry in the log file that says "audio device:
Loopback:0,0" or whatever alsa device you are using.
In reality you're probably better off making sure op25 plays audio through the default
device before you worry about getting it going through the loopback.
As a secondary recommendation, the alsa buffering is considerably better in the version of
sockaudio.py found in my github repo. I suggest you upgrade to that version as it has
other fixes too.
Graham
11:47 p.m.
On Tue, Oct 17, 2017 at 2:34 PM, gnorbury(a)bondcar.com [op25-dev]
<op25-dev(a)yahoogroups.com> wrote:
More command line options are needed:
-O Loopback:0,0 -V -2 -U
Thanks, I must have missed that detail with the conversion to rx.py. I
already have -2, but not -V and -U.
This fixed it. I'm getting audio through the loopback now!
Make sure you see an entry in the log file that says
"audio device: Loopback:0,0" or whatever alsa device you are using.
I don't see this line in the output. I also haven't specified a log
file. Is there a default log path? Is this something I should expect
to see in stdout or stderr?
In reality you're probably better off making sure
op25 plays audio through the default device before you worry about getting it going
through the loopback.
I tried that, but virt-manager isn't even passing the speaker test
audio, so I think that's a lost cause. There is no physical sound
output on this system.
As a secondary recommendation, the alsa buffering is
considerably better in the version of sockaudio.py found in my github repo. I suggest you
upgrade to that version as it has other fixes too.
It's unclear to me what sockaudio.py does and how I would use it with
darkice. Is there an example of this somewhere?
I tried your repo, but got an error, so I went back to max:
tom@stream:~/op25/op25/gr-op25_repeater/apps$ git status
HEAD detached at boatbod/master
Untracked files:
(use "git add <file>..." to include in what will be committed)
../../../build/
ss911_tacoma.tsv
ss911_tgs.tsv
nothing added to commit but untracked files present (use "git add" to track)
tom@stream:~/op25/op25/gr-op25_repeater/apps$ ./rx.py --args rtl=0 -N
LNA:42 -f 774081250 -S 1320000 -v 5 -2 -T ss911_tacoma.tsv
linux; GNU C++ version 4.8.2; Boost_105400; UHD_003.005.005-0-unknown
gr-osmosdr 0.1.1 (0.1.1) gnuradio 3.7.2.1
built-in source types: file osmosdr fcd rtl rtl_tcp uhd hackrf bladerf rfspace
Using device #0 Realtek RTL2838UHIDIR SN: 00000001
Found Rafael Micro R820T tuner
gain: name: LNA range: start 0 stop 0 step 0
setting gain LNA to 42
supported sample rates 250000-2560000 step 24000
Exact sample rate is: 1320000.047207 Hz
set_center_freq: 774081250
Using Volk machine: sse4_2_64_orc
Traceback (most recent call last):
File "./rx.py", line 671, in <module>
tb = p25_rx_block()
File "./rx.py", line 220, in __init__
self.open_usrp()
File "./rx.py", line 606, in open_usrp
self.__set_rx_from_osmosdr()
File "./rx.py", line 542, in __set_rx_from_osmosdr
self.__build_graph(self.src, capture_rate)
File "./rx.py", line 283, in __build_graph
symbol_rate = self.symbol_rate)
File "/home/tom/op25/op25/gr-op25_repeater/apps/p25_demodulator.py",
line 173, in __init__
self.lpf = filter.fft_filter_ccf(decimation, lpf_coeffs)
AttributeError: 'module' object has no attribute 'fft_filter_ccf'
Tom
18 Oct
18 Oct
4:40 a.m.
If you have audio I guess it's working so there's no need to worry about the audio
device line in the log file :)
The log is generally produced by redirecting stderr to something. Typically
"stderr.2" if you follow the example in the README. You can "tail -f
stderr.2" to see a real-time output of log messages if there are any.
I'm not sure what went wrong with my repo, but you can access it pretty easily as
follows:
cd ~
git clone https://github.com/boatbod/op25 https://github.com/boatbod/op25 op25.boatbod
cd op25.boatbod
mkdir build
cd build
cmake ../
make
sudo make install
I suspect the problem might have been that you tried to clone my repo into the existing
max branch, which comes from a different source and branch. Best to delete and start
over, or clone into a different subdir as shown above.
Graham
4:45 a.m.
Addendum: the filter error is something I saw once before. For some unknown reason your
version of gnuradio does not have the fft_filter_ccf() routine, so you are forced to use
fir_filter_ccf() instead. Max's code uses the FIR version by default, but I'd
changed mine to use FFT because it was much more cpu-efficient for the lightweight
hardware such as Rasp PI 3.
You can edit the p25_demodulator.py file, line 173 and change 'fft' to
'fir' and it should work on your system.
6:49 p.m.
On Tue, Oct 17, 2017 at 7:45 PM, gnorbury(a)bondcar.com [op25-dev] <
op25-dev(a)yahoogroups.com> wrote:
Addendum: the filter error is something I saw once before. For some
unknown reason your version of gnuradio does not have the fft_filter_ccf()
routine, so you are forced to use fir_filter_ccf() instead. Max's code
uses the FIR version by default, but I'd changed mine to use FFT because it
was much more cpu-efficient for the lightweight hardware such as Rasp PI 3.
You can edit the p25_demodulator.py file, line 173 and change 'fft' to
'fir' and it should work on your system.
Lower CPU usage sounds nice. I'd like to take advantage of it. My
installation has the Ubuntu 14.04 gnuradio package (having followed these
instructions:
http://op25.osmocom.org/trac/wiki.=/wiki/InstallInstructionsPage#InstallIns…).
Is there a different distribution or gnuradio package you recommend to get
fft_filter_ccf()?
Tom
19 Oct
19 Oct
12:21 a.m.
I just pushed a fix for the fft_filter_ccf() / fir_filter_ccf() error that a couple of
people have experienced. Fundamentally I don't know why some gnuradio installations
have both while others only have one, but the p25_demodulator code now falls back to fir
if fft is not present.
Anyone who wants the update will need to "git pull" to download the change.
2388
days inactive
2389
days old
6 comments
2 participants
participants (2)
-
gnorbury＠bondcar.com [op25-dev] -
Tom Hayward tom＠tomh.us [op25-dev]