[mythtv-users] Wireless

Stephen Worthington stephen_agent at jsw.gen.nz
Sun Nov 29 10:39:33 UTC 2015


On Sun, 29 Nov 2015 11:07:49 +0800, you wrote:

>Hi
>I installed mythfrontend on a (new) macbook.
>Jerkeyness on steroids 1/3 second pause every 1-2 seconds, Totally unwatchable.
>I tried youtube on the same machine and it is smooth.
>I tried 8player on my wife’s ipad and without transcoding it plays (uPnP) smoothly.
>I tried various (bubbleUpnp, mythplayer, others) on my galaxy tab s and the results are good..bad.
>
>[The macbook has a single USBC port only, so trying wired is a big deal]
>
>I’ll setup and try some linux wireless frontends.
>The modem is a TP-LINK which claims (11) a b g n
>The macbook claims b g n.
>My speed tests show 500k bytes/sec at 5m and 160k at 20m.
>
>[for the ignorant like me 11n is 300M bits/sec]
>Something does not compute, and I am ignorant, can anyone caste opinions please.
>Thanks
>James

In order to have smooth playback, the speed of the WiFi connection
needs to exceed the highest bit rate in a recording or video file.
Exceeding the average bit rate is not enough as there are often huge
peaks in the bit rate (eg when there is a lot of movement in the
picture).  So it depends on exactly what you are playing whether it
works smoothly or not.  Some of my 1080i HD DVB-T recordings are
around 10,000 kbits/s average (10 Mbit/s), so you would think that
they would play on a reasonable 802.11n WiFi connection (say 50
Mbit/s), but the peak bit rate it more likely to be up to 10 times the
average, so in fact they will not play properly.  My lower bit rate
recordings from DVB-S are likely to play OK though.

The mediainfo program on Linux or Windows can tell you the bit rate of
a file, but normally what it tells you is only the average bit rate,
not the maximum.

In any case, your WiFi speeds are not very good.  At 500 kbytes/s,
that is only 4 Mbits/s, which is way below what should be expected
from a good 802.11n connection.  At 160 kbytes/s, that is only 1.2
Mbits/s, which is pathetic for 802.11n, unless there are lots of
obstructions in the signal path.

The 300 Mbits/s that is often quoted for 802.11n is a theoretical
maximum, and is only attained in real life when the two aerials are
centimeters away from each other.  Real 802.11n speeds are typically
less than 100 Mbits/s, and across a real size house, through walls and
so on, it will be much less.  If both of your devices are 2.4 GHz and
5 GHz capable, you should try both frequencies.  Typically, 2.4 GHz
goes longer distances, but 5 GHz provides faster speeds at closer
distances.

If you want to be reasonably sure that you can play most video on a
WiFi connection, then it is best to have 802.11ac, which is much
faster, although it can slow to 802.11n speeds at longer distances.

You also need to be comparing apples with apples - a typical YouTube
video is likely to be a way lower bit rate than a typical TV recording
from an HD source.  So you need to be playing the same file on each
device for it to be a fair test.

And the YouTube software does rather more buffering than MythTV, so as
long as the periods of peak bit rate are not too long, it will be able
to continue playing out of the buffer and will only stall when the
buffer is empty.  I would suggest trying to play a MythTV recording
that stutters in mythfrontend on your MacBook using a player program
like VLC.  That would involve setting up a way for VLC to browse the
recording directory on your mythbackend box (such as Samba).  If VLC
can play it, but mythfrontend can not, then the difference is going to
be the amount of buffering being done.


More information about the mythtv-users mailing list