OP25 Rx.py vs. multi_rx.py efficiency

Status
Not open for further replies.

maus92

Member
Premium Subscriber
Joined
Jun 23, 2004
Messages
8,270
Location
The OP
I finally got around to configuring op25 to run multi_rx to on a single RTL device, basically mirroring the same setup I use for rx.py. Does it make sense that multi_rx *seems* to be more efficient than rx? Both setups run on the exact same hardware, and use the same whitelist and tg-tags files.
 

NebraskaCoder

Member
Joined
Oct 26, 2016
Messages
325
Location
Omaha, NE
I believe that he is adding improvements to multirx and leaving rx.py alone. I know it's that way for new features, not sure about improvements.
 

maus92

Member
Premium Subscriber
Joined
Jun 23, 2004
Messages
8,270
Location
The OP
I believe that he is adding improvements to multirx and leaving rx.py alone. I know it's that way for new features, not sure about improvements.
It does seem to be able to be more accurately fine-tuned - the same sdr is all over the place with rx.py
 

nick0909

Antenna flicker
Feed Provider
Joined
Jan 4, 2003
Messages
140
I may have to try multi again, I've been using rx.py and last time about a year ago when I tried to switch multi_rx would drop the signal and stop decoding after a day or two.
 

Outerdog

T¹ ÆS Ø
Premium Subscriber
Joined
Jul 1, 2016
Messages
667
I may have to try multi again, I've been using rx.py and last time about a year ago when I tried to switch multi_rx would drop the signal and stop decoding after a day or two.

Try it again. Either version. I run Osmocom (Max) and with 45 million TSBKs decoded, it has been running smoothly since Nov 21. I've had it as high as 89 million before I had to restart the machine for an unrelated issue.
 

Attachments

  • Screenshot 2021-12-04 182836.png
    Screenshot 2021-12-04 182836.png
    35.1 KB · Views: 38

boatbod

Member
Joined
Mar 3, 2007
Messages
3,417
Location
Talbot Co, MD
Try it again. Either version. I run Osmocom (Max) and with 45 million TSBKs decoded, it has been running smoothly since Nov 21. I've had it as high as 89 million before I had to restart the machine for an unrelated issue.
My streaming server usually runs unmolested for weeks/months on end, that is right up until my inquisitive cat steps on the SDR and borks the USB connection. I think I've moved it out of his reach, but apparently he's smarter than me!
 

Outerdog

T¹ ÆS Ø
Premium Subscriber
Joined
Jul 1, 2016
Messages
667
I think I allowed a Windows update to take place which caused the terminal window containing OP25 to obviously go away. Running it this way might not be best, but OP25 is headless here, so... meh. TSBK counts are about as useless as uptime counts, but some of us still do it...
 

boatbod

Member
Joined
Mar 3, 2007
Messages
3,417
Location
Talbot Co, MD
I finally got around to configuring op25 to run multi_rx to on a single RTL device, basically mirroring the same setup I use for rx.py. Does it make sense that multi_rx *seems* to be more efficient than rx? Both setups run on the exact same hardware, and use the same whitelist and tg-tags files.
What you see is definitely possible. In my version of multi_rx, the frame assembler passes a block of symbols to the p25 handler with nid check performed just once, then comes and waits for the remaining symbols to complete the frame before proceeding with decode. In contrast the rx.py frame assembler passes each individual symbol to the p25 handler and that in turn incurs a lot more bit shuffling and sync flag/nid checks. I've had a couple people observe that it seems to be a bit more efficient, but as yet no hard data to quantify by how much.
 

boatbod

Member
Joined
Mar 3, 2007
Messages
3,417
Location
Talbot Co, MD
I think I allowed a Windows update to take place which caused the terminal window containing OP25 to obviously go away. Running it this way might not be best, but OP25 is headless here, so... meh. TSBK counts are about as useless as uptime counts, but some of us still do it...
In my view the sole purpose of the tsbk counter is to see that the system is still alive and receiving data ;)
 
Status
Not open for further replies.
Top