Does lowering Display FPS or Waterfall Rate Reduce Network Bandwidth?

  • 1
  • Question
  • Updated 3 years ago

Remote SmartSDR operations is coming soon to a radio near you :-) , therefore, understanding how to best utilize available bandwidth so that one can maintain optimized audio seems to be a reasonable concern.  

I understand that in most cases LAN based remote operations the 6000 series won't come anywhere near exceeding available bandwidth.  This will not necessarily be true with WAN Remote Ops when v2.x is released next year.  With this in mind I have the following questions:

  1. Will reducing the Frames Per Second (FPS) Rate and/or the Waterfall Rate on SmartSDR cause the 6000 to throttle the number display packets created and sent across the network or does this only change how the client reacts to a constant number of display packets sent by the 6000?  

  2. Will changing the SmartSDR setting for the display "AVG" or "Weighted Average" affect the number of packets transmitted by the 6000?

Having this knowledge will help to develop tactics on how to best accommodate low bandwidth conditions while maintaining optimized receive and transmit audio. 

Flex Community...please reply ONLY if you have either measured empirical evidence or have direct knowledge of how SmartSDR/Flex 6000 is programmed to work.  I would really like this thread to be based on fact and not feelings or assumptions.


Thanks in advanced to all.

Photo of Jay -- N0FB

Jay -- N0FB, Elmer

  • 534 Posts
  • 210 Reply Likes

Posted 3 years ago

  • 1
Photo of Richard Clafton W4/G7EIX

Richard Clafton W4/G7EIX, Elmer

  • 455 Posts
  • 117 Reply Likes

If you want results you can trust as opposed to dealing with assumptions...  Install WireShark.   Run some tests.   Get results.  Job done.

Merry Christmas!  (Merry Boxing Day!)

Photo of Jay -- N0FB

Jay -- N0FB, Elmer

  • 534 Posts
  • 210 Reply Likes
Thanks Richard I understand.  There is a method to my madness.  I'm hoping for a definitive statement from the good folks at Flex Radio. 

Merry Christmas to you!
Photo of Jim Gilliam

Jim Gilliam

  • 459 Posts
  • 71 Reply Likes
This answer might not fall into your specific question, but it seems there should not be a problem based on that I remote the 6500 over the Internet using Teamviewer with a local remote client. I see no interruption of the panoramic display nor the waterfall. I am using Charter cable at about 10 Mbs. So perhaps, the constraining factor(s) will be that  of the bandwidth availability from your ISP rather than the software/firmware of the Flex.
Photo of Jay -- N0FB

Jay -- N0FB, Elmer

  • 534 Posts
  • 210 Reply Likes
Hi Jim, you are right, the evidence you are providing is more of a testament to TeamViewer's ability to re-transmit screen activity of your home computer to a remote computer running Teamviewer.   What the Flex 6000 does when communicating directly between itself and to a remote computer running SmartSDR will be different.   Thank you for your observations!
Photo of Dave KD5FX

Dave KD5FX

  • 83 Posts
  • 4 Reply Likes
You can easily test this yourself if you are using Windows. Turn on the Task Manager and look at the Networking tab.
Here are the results I found on my PC:
FPS    Net%
30       .20
20       .14
10       .08
 1        .02

No change with weighted avg on or off or by changing the AVG slider.
Adding a 2nd Panadapter nearly doubled the network traffic.

73, Dave
Photo of Jay -- N0FB

Jay -- N0FB, Elmer

  • 534 Posts
  • 210 Reply Likes
Good stuff Dave!  Thanks!
Photo of Jim Gilliam

Jim Gilliam

  • 459 Posts
  • 71 Reply Likes
It will be very interesting to compare the differences when V2 comes out.
Photo of George Molnar, KF2T

George Molnar, KF2T, Elmer

  • 1529 Posts
  • 541 Reply Likes
Can't imagine that changing the video activity wouldn't affect bandwidth. In my old life in TV broadcasting, it was quite an exercise to trade off picture quality for bandwidth in a digital multicast. I will do some testing when I get home next week, but suspect ther results with SSDR 1.4 will be somewhat different than 1.38 due to the extensive coding changes being made. Either way, it's always fun peeking under the hood!
Photo of Jay -- N0FB

Jay -- N0FB, Elmer

  • 534 Posts
  • 210 Reply Likes
Here are some findings from my testing:  Three processes associated with SmartSDR consume bandwidth while SmartSDR is actively connected to my Flex 6300.  My tests show differences when changing the Panadaptor Frames Per-Second (FPS) and Waterfall Rate.  My testing showed no measurable differences when Averaging or Weighted Average are enabled/disabled.  These individual measurements are not included below.



Environment
Wireless AC Router (5GHz Band)
Computer: Dell with Intel P5 running @ 2.4 GHz (2 Hyper-threaded Cores)
Computer Network Connection:  Wireless AC Adapter connected at a 886 Mbps link speed
Radio:  Flex 6300 tuned to 28.480. 3.3 KHz pass band filter, 1 Slice Receiver Active
Radio to Router Connection:  1 GB Cat-6
Panadaptor width:  100 KHz  (28.420 - 28.520 MHz)

Results Interpretation

It appears that the Frames Per Second & Waterfall Rate have a significant impact to the overall bandwidth utilization with SmartSDR 1.3.8.   Enabling a DAX channel also has an equal impact to bandwidth utilization as setting FPS and Waterfall Rate to maximum settings.  This means that in the future, when WAN Remote is made available, should the operator find themselves on a less than optimal connection to the internet, they can reduce the bandwidth overhead by reducing the Frames Per Second and/or Waterfall rate percentage.  The reduced visual panadaptor information is a good trade off (in my estimation) for providing adequate bandwidth for receive and transmit audio.

As I am not a FRS Beta tester (would love to be), I don't have access to the pre-released version of 1.4.  For this reason I am unable to include what the compressed voice data bandwidth utilization with this analysis.  I would bet that it will likely be significantly less than the requirements for a single DAX enabled channel, but this is only conjecture.
(Edited)
Photo of Les W6JKA

Les W6JKA

  • 11 Posts
  • 1 Reply Like
I have been using a Flex6700 with a Dell Optiplex XE2 Server for quite awhile now and have come up with the same conclusion. Reducing the waterfall rate definitely reduces bandwidth. I travel in my motor coach quite a bit and a decent connection is critical. Hope the new 1.4/1.5 deals with the audio bandwidth. Presently steering to Skype for audio. Using Ultra VNC for the server with Mirror Driver.
Photo of Jim Gilliam

Jim Gilliam

  • 459 Posts
  • 71 Reply Likes

You will find using RemAud has better quality audio with much less latency than Skype. Most important to hams, it's free.  


http://www.latimhttp://www.df3cb.com/remaud/


Photo of Richard Clafton W4/G7EIX

Richard Clafton W4/G7EIX, Elmer

  • 455 Posts
  • 117 Reply Likes
Photo of KM4CQG

KM4CQG

  • 222 Posts
  • 34 Reply Likes
Waterfall will give you the best results from what I have tested. Turn it off and you can run all your receivers with no worries even on a mid level processor.
Ian
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9046 Posts
  • 3435 Reply Likes
Jay - your experiment and results are correct.  Reducing the frame rate does reduce the number of bits on the wire.  Turning Averaging on or off will not make an appreciable difference.