Wi-Fi bandwidth

Post by pinnerite
I have been experiencing jerkyness streaming live HD TV from MythTV over wi-fi to my work computer. SD is fine and both are fine on the MythTV server.
So, I concluded that it must be limits across the wi-fi.
is there /are there any programs that can demonstrate the wi-fi bandwith and perhaps how it is bewing sliced up?
I couldn't find anything from a brief search via DuckDuckGo or google.

The bandwidth is probably more determined by that last 20 feet than
anything else, and that is NOT determinable by any generic statements.
Ie, the wireless part between your computer the modem. It will depend on
environmental noise (is your microwave running) by intevening walls (eg
steel studs), the positioning of the antennae, etc.

Try using some of those speed tests you can find on the web.

Post by pinnerite
TIA
Alan

Bud Frede

2021-04-04 10:06:21 UTC

Post by William Unruh

The key thing is usually airtime. It _is_ possible to measure what's
actually going on, but AFAIK it requires more sophisticated test
equipment than most people have access to.

Paul

2021-04-02 23:45:37 UTC

These are the instructions for Windows. I believe "ttcp"
started on another platform, check your package manager
and see if it is there. I include this, just so I won't
have to re-write it.

https://web.archive.org/web/20130606180737if_/http://www.pcausa.com/Utilities/pcattcp/PCATTCP-0114.zip

On the second (perfectly working) machine, set up the receiver first.

ipconfig # This gives the address of this machine (192.168.1.3)

pcattcp -r -4 # The receiver sits and waits for the transmitter

On the broken machine, do

pcattcp -t -4 192.168.1.3 # Substitute IP address from ipconfig result...

*******

It's almost as basic as Ping, but not quite.

And here is the Linux version.

https://linux.die.net/man/1/ttcp

It's a very basic test.

You could also set up an ftp server on one machine,
and an ftp client on another machine, and do some
testing that way. Some platforms allow setting up
an FTP server with one click - that was MacOSX :-)
Usually, it's more work to set up ftp or http servers
for LAN usage in testing.

There's more than one program out there to test with,
but there is also the tendency to go overboard with
this stuff, and some of the others might be more trouble
than they're worth.

Paul

Bit Twister

2021-04-03 03:03:51 UTC

Do not know about wi-fi bandwith but you might look into iperf3 for throughput testing.

You start it as a server on one node, then run it on the client node.

For example I have three nodes wb, tb, and mtv. wb is the server and the command
I would run from the client nodes. Following is from my brain book.

$ uh network speed
_network_debug_speed_testing_server_wb_ iperf3 -s
_network_debug_speed_testing_client_mtv_ iperf3 -c wb
_network_debug_speed_testing_client_tb_ iperf3 -c wb

------key words for uh search -------- ---- command to execute--------

Henry Crun

2021-04-03 04:46:31 UTC

For basic info try installing wavemon (runs on Ubuntu)

--
Mike R.
Home: http://alpha.mike-r.com/
QOTD: http://alpha.mike-r.com/qotd.php
No Micro$oft products were used in the URLs above, or in preparing this message.
Recommended reading: http://www.catb.org/~esr/faqs/smart-questions.html#before
and: http://alpha.mike-r.com/jargon/T/top-post.html
Missile address: N31.7624/E34.9691

pinnerite

2021-04-03 11:01:50 UTC

On Sat, 3 Apr 2021 07:46:31 +0300

Post by Henry Crun

For basic info try installing wavemon (runs on Ubuntu)

I installed wavemon on the mythtv server machine via ssh.
I am not sure whether what it reports is about the internal wi-fi stuff or the quality of its connection via the router to the outside world.

For now, I will stick with viewing live TV in SD.

Thank you for your advice.

Alan

--
Mint 20.04, kernel 5.4.0-42-generic, Cinnamon 4.6.7
running on an AMD Phenom II X4 Black edition processor with 8GB of DRAM.

Bit Twister

2021-04-03 11:51:43 UTC

Post by pinnerite
On Sat, 3 Apr 2021 07:46:31 +0300

Post by Henry Crun

For basic info try installing wavemon (runs on Ubuntu)

I installed wavemon on the mythtv server machine via ssh.
I am not sure whether what it reports is about the internal wi-fi stuff or the quality of its connection via the router to the outside world.

Personally, I would suggest you would want to compare results on both
server and client machines.

Do run ifconfig wi-fi_here and check the Tx/Rx lines for problems on
the client machines.

Chris Elvidge

2021-04-03 12:09:59 UTC

Are you using 2.4GHz or 5GHz?
5GHz will give a better data rate.
HD needs more data then SD.

--
Chris Elvidge
England

Jonathan N. Little

2021-04-03 13:11:13 UTC

Post by pinnerite
I have been experiencing jerkyness streaming live HD TV from MythTV
over wi-fi to my work computer. SD is fine and both are fine on the
MythTV server.
So, I concluded that it must be limits across the wi-fi.
is there /are there any programs that can demonstrate the wi-fi
bandwith and perhaps how it is bewing sliced up?
I couldn't find anything from a brief search via DuckDuckGo or google.
TIA
Alan

Are you using 2.4GHz or 5GHz?
5GHz will give a better data rate.
HD needs more data then SD.

Additionally, what it distance between server and tv? What is between
server and tv? What channel are you using and proximity to other
networks on the same channel? 5GHz will have higher through put than
2.4GHz but at the expense of range and sensitivity to interference. All
the advantages of the wire over wireless.

--
Take care,

Jonathan
-------------------
LITTLE WORKS STUDIO
http://www.LittleWorksStudio.com

2021-04-03 14:34:28 UTC

Post by Jonathan N. Little

Post by pinnerite
I have been experiencing jerkyness streaming live HD TV from MythTV
over wi-fi to my work computer. SD is fine and both are fine on the
MythTV server.
So, I concluded that it must be limits across the wi-fi.
is there /are there any programs that can demonstrate the wi-fi
bandwith and perhaps how it is bewing sliced up?
I couldn't find anything from a brief search via DuckDuckGo or google.
TIA
Alan

Are you using 2.4GHz or 5GHz?
5GHz will give a better data rate.
HD needs more data then SD.

Agreed. When I got my first dual-band router, I thought I'd use the 5GHz
band on my laptop, leaving the lower frequency to my brother so we
didn't have to share the bandwidth on the lower frequency. But, I soon
learned that when I put a couple of walls and a floor between my router
and laptop, 5Ghz was actually the slower of the two.

TJ

Jonathan N. Little

2021-04-03 15:26:27 UTC

<snip>

Post by TJ

Post by Jonathan N. Little
Additionally, what it distance between server and tv? What is between
server and tv? What channel are you using and proximity to other
networks on the same channel? 5GHz will have higher through put than
2.4GHz but at the expense of range and sensitivity to interference. All
the advantages of the wire over wireless.

Absolutely, you may have a higher initial base bandwidth value but with
a weak and obstructed signal lose all advantage as dropped packets
require resending.

--
Take care,

Jonathan
-------------------
LITTLE WORKS STUDIO
http://www.LittleWorksStudio.com

Melzzzzz

2021-04-03 14:47:36 UTC

["Followup-To:" header set to alt.os.linux.mint.]

Are you using 2.4GHz or 5GHz?
5GHz will give a better data rate.

Yes, but weaker signal.

Post by Chris Elvidge
HD needs more data then SD.

--
current job title: senior software engineer
skills: x86 aasembler,c++,c,rust,go,nim,haskell...

press any key to continue or any other to quit...

Bud Frede

2021-04-04 10:10:17 UTC

Are you using 2.4GHz or 5GHz?
5GHz will give a better data rate.
HD needs more data then SD.

HD should still be quite possible with 2.4GHz channels. It also
penetrates better, so in some cases will perform better over a distance.

All of this also depends upon where you live and whether or not there
are competing WiFi networks within range (and how busy those networks
are).

Abandoned_Trolley

2021-04-03 15:01:28 UTC

Why not stop worrying about "bandwidth" (whatever that means to you) and
check the actual connection speed, just like any normal phone or laptop
user would ?

Theres no mention of it anywhere in this thread

AT

Bud Frede

2021-04-04 10:19:33 UTC

Why not stop worrying about "bandwidth" (whatever that means to you)
and check the actual connection speed, just like any normal phone or
laptop user would ?
Theres no mention of it anywhere in this thread

It can vary from moment to moment, and may not mean much anyway. It's
like the "bars" that you see on a typical cell phone that are supposed
to indicate signal strength in a meaningful way. :-)

I've been told about some consumer network equipment that recognizes
when one of the popular speed test sites is being used and prioritizes
that traffic so that speed test results look better. (Doing a "VW," in
other words.)

I know of some other consumer network equipment that learns traffic
patterns and changes its behavior based on them. A streaming 4K video
session will be optimized because it takes place over a long period of
time. A speed test is too short in duration to be optimized for (unless
you're cheating like the other company), and it's highly abnormal usage
as well, so the results shown may not reflect real-world usage of the
network.

Abandoned_Trolley

2021-04-04 10:51:50 UTC

Post by Bud Frede

Why not stop worrying about "bandwidth" (whatever that means to you)
and check the actual connection speed, just like any normal phone or
laptop user would ?
Theres no mention of it anywhere in this thread

It can vary from moment to moment, and may not mean much anyway. It's
like the "bars" that you see on a typical cell phone that are supposed
to indicate signal strength in a meaningful way. :-)
I've been told about some consumer network equipment that recognizes
when one of the popular speed test sites is being used and prioritizes
that traffic so that speed test results look better. (Doing a "VW," in
other words.)
I know of some other consumer network equipment that learns traffic
patterns and changes its behavior based on them. A streaming 4K video
session will be optimized because it takes place over a long period of
time. A speed test is too short in duration to be optimized for (unless
you're cheating like the other company), and it's highly abnormal usage
as well, so the results shown may not reflect real-world usage of the
network.

I am well aware of all of that. However, there is still no mention at
all of the connection speed anywhere in this thread.

More importantly, theres no mention of any attempt to ascertain the
actual (wired or fibre) download speed from the ISP to the router (which
DOES NOT vary from moment to moment) which is more likely to be the
cause of the problem.

I would check the admin page of the router and see what the downstream
line rate is and work from there. If it looks like its OK then (if
possible) it might be an idea to make a temporary WIRED connection to
the router and see if that cures the problem. If it doesnt, then clearly
its not a wi-fi issue.

Nothing I have seen in this thread so far conclusively proves that the
wi-fi is at fault.

AT

Paul

2021-04-04 12:02:44 UTC

Post by Bud Frede

Post by pinnerite
I have been experiencing jerkyness streaming live HD TV from MythTV
over wi-fi to my work computer. SD is fine and both are fine on the
MythTV server.
So, I concluded that it must be limits across the wi-fi.
is there /are there any programs that can demonstrate the wi-fi
bandwith and perhaps how it is bewing sliced up?
I couldn't find anything from a brief search via DuckDuckGo or google.
TIA
Alan

Why not stop worrying about "bandwidth" (whatever that means to you)
and check the actual connection speed, just like any normal phone or
laptop user would ?
Theres no mention of it anywhere in this thread

It can vary from moment to moment, and may not mean much anyway. It's
like the "bars" that you see on a typical cell phone that are supposed
to indicate signal strength in a meaningful way. :-)
I've been told about some consumer network equipment that recognizes
when one of the popular speed test sites is being used and prioritizes
that traffic so that speed test results look better. (Doing a "VW," in
other words.)
I know of some other consumer network equipment that learns traffic
patterns and changes its behavior based on them. A streaming 4K video
session will be optimized because it takes place over a long period of
time. A speed test is too short in duration to be optimized for (unless
you're cheating like the other company), and it's highly abnormal usage
as well, so the results shown may not reflect real-world usage of the
network.

I am well aware of all of that. However, there is still no mention at
all of the connection speed anywhere in this thread.
More importantly, theres no mention of any attempt to ascertain the
actual (wired or fibre) download speed from the ISP to the router (which
DOES NOT vary from moment to moment) which is more likely to be the
cause of the problem.
I would check the admin page of the router and see what the downstream
line rate is and work from there. If it looks like its OK then (if
possible) it might be an idea to make a temporary WIRED connection to
the router and see if that cures the problem. If it doesnt, then clearly
its not a wi-fi issue.
Nothing I have seen in this thread so far conclusively proves that the
wi-fi is at fault.
AT

Pinnerite setup, described in OP

Some-Wifi-Router ---+---------------------+
| |
Client MythTV -- OTA
(Wifi) (Wifi)
PC PC

This is a question of LAN performance, nothing
about WAN or pulling Netflix or anything like that.

MythTV is capable of complicated setups, but we're not
being told about that. There can be front ends and
back ends in the picture. Tuners doing capture,
would burn up as much bandwidth as client sessions
(2 megabytes/sec per tuner?). A capture tuner could be
competing for bandwidth, with a client session trying
to watch a recording.

Loading Image...

You could do an air survey or whatever. See whether
there's a better Wifi channel or band choice maybe.

https://www.pcwdld.com/wifi-tools

Or any sort of simple LAN transfer would indicate
whether something is broken. If you can only get
11 megabits/sec during a file transfer, that's going
to doom TV streaming.

Doing speedtest.net won't help, because it's not
a WAN issue.

If a router is not supporting QOS (quality of service)
protocol options, then the router might be using nothing
more complicated than round robin. That's what the LAN
to WAN router on my network does. The higher speed portion
is done with a GbE switch. If one machine opens too many
WAN-directed connections to the router, the other machines
might not get as many slices of the WAN pie as a result.
I can just about starve a second machine (can't web surf),
if the first machine opens a shitload of connections. But
this is an issue of getting fair share on the WAN end,
which is not particularly the problem here.

Networking questions can benefit from drawings,
but what are ya gonna do.

Paul

pinnerite

2021-04-06 16:59:48 UTC

On Sat, 3 Apr 2021 16:01:28 +0100

Post by Abandoned_Trolley
Why not stop worrying about "bandwidth" (whatever that means to you) and
check the actual connection speed, just like any normal phone or laptop
user would ?
Theres no mention of it anywhere in this thread
AT

If I disconnect the phone connection at the router, I still have wi-fi access between my devices.

Recording TV does not depend on the wi-fi. the signal comes from the TV aerial directly to both the TV and the HTPC.

Both my work computer and the TV only handle 2.4Ghz. The remaining items are enabled for that and 5Ghz including a TP-Link wi-fi extender.

Now there is no problem on the combined frontend / backend server presumably because wi-fi is not involved. Similarly, playback to the TV is fine because it is connected by an HDMI cable. Sadly to the remote frontend on my work-machine it is.

The extraordinary thing is that playback of HD via wi-fi to my work machine is fine,
Live TV is not. Logically I would expect them behave the same.

As Ian Lang often says "I blame the EU".

--
Mint 20.04, kernel 5.4.0-42-generic, Cinnamon 4.6.7
running on an AMD Phenom II X4 Black edition processor with 8GB of DRAM.

Paul

2021-04-07 03:13:04 UTC

Post by pinnerite
On Sat, 3 Apr 2021 16:01:28 +0100

If I disconnect the phone connection at the router, I still have wi-fi access between my devices.
Recording TV does not depend on the wi-fi. the signal comes from the TV aerial directly to both the TV and the HTPC.
Both my work computer and the TV only handle 2.4Ghz. The remaining items are enabled for that and 5Ghz including a TP-Link wi-fi extender.
Now there is no problem on the combined frontend / backend server presumably because wi-fi is not involved. Similarly, playback to the TV is fine because it is connected by an HDMI cable. Sadly to the remote frontend on my work-machine it is.
The extraordinary thing is that playback of HD via wi-fi to my work machine is fine,
Live TV is not. Logically I would expect them behave the same.
As Ian Lang often says "I blame the EU".

But you know your MythTV better than anyone here.

The software (not MythTV) for my tuner card, does it this way.

"Live TV"

1) Comes into tuner from OTA antenna.
2) Tuner packets are written to disk. Disk could be on a backend machine.
3) Streaming software reads from disk (delayed maybe a second or two).
4) MPEG2 video now needs to be decoded and sent to a screen, from (3)
This uses a bit of CPU (in the old days), or it uses
the Video SIP on your video card.

The "feature" that comes with this, is while you are watching LiveTV,
you can "commit" what has just been shown, and the file stored on
the hard drive is then kept for later.

The activity in that sequence, might involve two passes on
Wifi, thrashing at the hard drive (reads and writes could be
smoothed a bit via caching), and CPU activity for decode
(which could be moved to a block on the video card).

With your knowledge of how MythTV works, maybe you'll find
a bottleneck similar to the above.

Paul

pinnerite

2021-04-09 09:15:08 UTC

On Tue, 06 Apr 2021 23:13:04 -0400

Post by Paul

Post by pinnerite
On Sat, 3 Apr 2021 16:01:28 +0100

If I disconnect the phone connection at the router, I still have wi-fi access between my devices.
Recording TV does not depend on the wi-fi. the signal comes from the TV aerial directly to both the TV and the HTPC.
Both my work computer and the TV only handle 2.4Ghz. The remaining items are enabled for that and 5Ghz including a TP-Link wi-fi extender.
Now there is no problem on the combined frontend / backend server presumably because wi-fi is not involved. Similarly, playback to the TV is fine because it is connected by an HDMI cable. Sadly to the remote frontend on my work-machine it is.
The extraordinary thing is that playback of HD via wi-fi to my work machine is fine,
Live TV is not. Logically I would expect them behave the same.
As Ian Lang often says "I blame the EU".

But you know your MythTV better than anyone here.
The software (not MythTV) for my tuner card, does it this way.
"Live TV"
1) Comes into tuner from OTA antenna.
2) Tuner packets are written to disk. Disk could be on a backend machine.
3) Streaming software reads from disk (delayed maybe a second or two).
4) MPEG2 video now needs to be decoded and sent to a screen, from (3)
This uses a bit of CPU (in the old days), or it uses
the Video SIP on your video card.
The "feature" that comes with this, is while you are watching LiveTV,
you can "commit" what has just been shown, and the file stored on
the hard drive is then kept for later.
The activity in that sequence, might involve two passes on
Wifi, thrashing at the hard drive (reads and writes could be
smoothed a bit via caching), and CPU activity for decode
(which could be moved to a block on the video card).
With your knowledge of how MythTV works, maybe you'll find
a bottleneck similar to the above.
Paul

That is a very useful explanation but most MythTV users will be connected much the same way, yet I have not come across simmilar complaints.

However!

This morning I had to reboot the router because our cellphones registered "No Wifi" ot no "Internet connection". This has been happening throughout the year. Not frequently but often enough. I am wondering whether the router (supplied by TalkTalk) is actually faulty.

-
Mint 20.04, kernel 5.4.0-42-generic, Cinnamon 4.6.7
running on an AMD Phenom II X4 Black edition processor with 8GB of DRAM.

Paul

2021-04-09 17:55:35 UTC

Post by pinnerite
This morning I had to reboot the router because our cellphones
registered "No Wifi" ot no "Internet connection". This has been
happening throughout the year. Not frequently but often enough.
I am wondering whether the router (supplied by TalkTalk) is actually faulty.

It's possible. My very first router was flake-city and
the most likely culprit was how the CPU was clocked.
Rather than the firmware being buggy. Embedded CPUs can
use a quartz crystal and an internal oscillator, or can
be driven from an external four pin tin can oscillator.
The tin can only has one job to do, and is thus more
trustworthy.

Sometimes, an excuse could be made for failures, based on
what the box-at-the-corner is doing.

All the other network boxes with a CPU inside since then,
have had the normal level of reliability. I don't consider
bad CPU design to be endemic. As the industry matures, things
should improve.

Another possibility is bad thermal design. As a recent example,
TPLink makes a new pair of 10GbE switches. One with five ports,
one with eight ports. The eight port one has a fan, the
five port one, does not have a fan. Hmmm. And the units
are expensive enough at retail, for a fan to be included
if it is needed. There was a time when GbE ports first
came out, where there were premature failures related
to thermals. One guy dremeled a hole in his and fitted
a regular heatsink to the main chip.

You can spot lots of really stupid stuff, when it
comes to engineering. I especially like disk enclosures
equipped with a fan, where there are no vent holes for
an air intake. And it works just as well as you would
expect :-)

Paul

Bit Twister

2021-04-10 00:21:16 UTC