Microsoft Teams Quality Issue

[QuantumPhys]

Hello all,

I've got an odd issue which I'm hoping someone can offer a steer on.

I have the EE 1.6 Gb Full Fibre Busiest Home package with a Smart Hub Plus.

All seemed to be working fine until my original Smart Hub Plus failed in Jan 2025. This was attributed to a bad firmware update, and a new hub was sent out. I connected this up and again everything appeared to be ok, until colleagues at work started complaining that my video during MS Teams calls was regularly freezing. On my end, I didn't see any issues - I could see and hear everyone else fine so it came as a surprise. Setting up a test meeting that just included me, I recorded myself and when playing back immediately found that the video and audio both freeze and stutter. This is readily reproducible, although sometimes the performance is extremely bad with the video completely frozen. Strangely, this only affects the MS Teams desktop app, with connections via the web browser still being good quality.

So far, everything else seems ok - stable and fast connection (getting around 1.6 Gbps routinely).

Here's what I've tried so far:

• Updated Teams and all drivers on laptop.
• Direct cable connection to router with WiFi turned off and all other devices disconnected. No difference.
• Laptop on WiFi on corporate network - absolutely fine, no surprise.
• Laptop on WiFi in various coffee shops and on friend's home broadband - all much slower connections than mine and Teams is absolutely fine. Often extremely good.
• Used phone as 4G hotspot - absolutely fine - extremely good video and audio.
• Replaced the Smart Hub Plus with an old Asus router I used with a previous ISP - connected to ONT. Connects via PPPoE fine with reasonable speed (around 750 Mbps although using 2.5G port as WAN port, so a little surprised it's not faster).  Connection appears very stable and no obvious issues shown by ThinkBroadband BQM. However, Teams behaves exactly the same as with the EE router - terrible video and audio.
• Configured Hotspot VPN on Asus router - set location to South England. Connection speed is around 100 Mbps. However, Teams now works absolutely fine. Very good audio and video. Turned off VPN - fast connection restored and Teams video and audio revert to being terrible.

This suggests to me it might be some network infrastructure issue, rather than a problem with the Smart Hub Plus or with the laptop / Teams. Could the route the traffic is following make a difference? If so, how would I go about pinpointing this? The Microsoft 365 connectivity tests don't flag up anything obvious.

After much testing to convince myself there was an issue with the connection I called EE and an engineer visited. He indicated there had been some problems with the Smart Hub Plus on fast connections, tried to run a series of tests but these didn't complete. Reason unclear. However, he was confident that the new Smart Hub Pro would resolve the issues although he wasn't allowed to issue one himself.

Engineer left, job appeared to be closed, then I got a message from OpenReach asking me to arrange a call so that equipment could be delivered. I did, got put through to EE, with the guide saying that I really should have had the Smart Hub Pro already and that he'd get one sent out, although the process for doing this was very difficult. That was almost a couple of weeks ago. Being concerned this wasn't being progressed I got back in touch - over multiple calls now, once I get past the agent telling me I've got a very fast connection and they can't see an issue or find open jobs, when they review previous notes there's a record somewhere to an order being placed, but no-one can find out what. It appears I'm now down as a new customer and something needs to be progressed by the "New IT" team, even though I've been with EE for over a year.

So, perhaps a Smart Hub Pro will be sent out, which will be good because it sounds like I'm entitled to one on my package.

However, I'm not convinced the hub itself is the issue, given that the Asus router performs similarly, except when a VPN is active.

Does anyone have any thoughts about what might be happening?

Thanks!

 

[HomeNetworkGuy]

The only options I see in Intel Connectivity Performance Suite are:
    1. to turn on/off prioritisation. 

    2. monitor usage details on/off.

I've turned off prioritisation and retested. 

I must admit I was very sceptical about this, as the alternative path (Vodafone 4G) worked so well and that the problem appeared to be related to the QoS marking, so it only made sense that the issue was occuring downstream. 

It looks like you found the problem!    I wish I understood how, I'm starting to think the person that I tested with over 4G was wrong and the problem did exist there too!

[HomeNetworkGuy]

[jonathan_c]

Thanks for letting us know HNG. I've actually been working on it with TC1234567 so credit should go to him. It appears the software is trying to be smart with UDP packets and this has a negative impact n the EE router/firmware/upstream exchange.

Contrary to some earlier comments, both teams peer to peer and teams meetings use UDP and failback to TCP, so it appears that meetings handle this a bit better (fallback) in this scenario.

TCP still gives a poor experience though, and friends don't let friends use Teams Web. Not when it can be avoided.

[Tryfan_Tamer]

This was my analysis earlier on in the thread  (So, my belief is the EE Smart Hub SH32B is the issue.  What I think might be happening is that the Smart Hub is trying to honor QOS markings from Teams/Zoom/Webex and somehow completely messing it up.  Older routers don't do this & that seems to be a good thing in this instance. )

I removed the problem by using an older router that effectively doesn't understand QOS markings and stripped them off.  What you are doing here is stripping off those markings at the exit point from your laptop/desktop which does the same thing.  That's an issue for me because when I go into the office or other networks my packets no longer have priority which might make my office experience poor. 

I think we can all agree that something in EE's network is trying to be clever with packet prioritisation and it's failing! 

Teams/Webex/Zoom will always tag packets with prioritisation and it was affecting all three for me which proves the issue isn't the software or the company (Microsoft/Cisco/Zoom).  The problem is EE network handling these UDP packets. 

All we need now is someone at EE to own this & sort it!! 

[HomeNetworkGuy]

You're right TT, I just took a wireshark capture of the 'good' webex session I'm in at the moment, and all my audio & video packets are showing as DSCP 0. That's why it's working correctly. I guess it was the Intel Connectivity Performance Suite allowing the WebEx application packets be honored at DSCP 46 and 38 before (which were being dropped downstream).

So while I'm happy I have a fix workaround, the network is still performing badly.  That explains why it worked OK with the Vodafone 4G Wifi tethering path.

[HomeNetworkGuy]

It seems that the Intel Connectivity Performance Suite workaround is temporary. I've not changed anything but the DSCP settings have come back this morning and my WebEx experience is bad again. 

[Tryfan_Tamer]

Good morning HNG

Is this a personal device or work one?  Quite often work polices enforce prioritisation and you can switch off but they will force back on again which is why I focused on the router side at home. 

[HomeNetworkGuy]

Morning TT 🙂  It's a work laptop, fully managed by the IT dept. The real fix is EE fixing the QoS  PHB. They never get back to me when I call support, they say they've escalated, but no one ever calls back. I'm getting closer and closer to moving ISPs.

[jonathan_c]

We found similar in testing - we were stripping the DSCP from the registry but GPO/Intune came back later and re-applied.

It might be worth toggling it on the off again in Intel suite, checking for any QoS settings in the registry at HKLM\Software\Policies\Microsoft\Windows\QoS 

You can check if if its GPO related

gpresult /H gp.html

Intune is a bit tricker, but you can find the last bunch of policies applied at HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Provisioning\NodeCache\CSP\Device\MS DM Server\Nodes

[deezle]

On my work laptop, Advanced connection manager is already disabled, but when I try and disable prioritisation, I can actually see the setting being force toggled back on & off. So no joy for me.