Speedtest result TBB vs OOKLA

I have observed a "slowness" to my general internet experience lately, so ran a speed test. The results are strange to me:

TBB reports 06-01-2020 09:19:18 41.2 Mbps 18.3 Mbpss 20ms latency

OOKLA reports 06.01.2020 9:16 67.54 Mbps 18.72 Mbps 6ms ping

What would explain the differences here, and should I be looking for 'an issue'?

PlusNet might be peering with Oookla or may even have Ookla based test servers within their network and so wouldn't be testing the peering links.

Which tester fits in with your feeling of how the service is performing?

The ookla results have a tendency to ignore some of the dips and troughs in speed, so if your lines are wiggly on the tbb test that probably explains why Ookla test is giving higher figures.

TBB was showing ups and downs during the test, but never reached >50Mbps.

My router is showing:
Connection Speed 79998 kbps 19999 kbps
Line Attenuation 11.2 db 10.5 db
Noise Margin 3.8 db 9.5 db

So I remain puzzled!

The server was GTT.net London, but I generally see fairly consistent results to several servers.

In the past, TBB also reported much higher speeds, e.g:
Sat 23/06/2018 20:46 75.56 Mbps 48.38 Mbps 19.12 Mbps

Albeit from 2018.

In reply to a post by aks:

The server was GTT.net London, but I generally see fairly consistent results to several servers.

In that case there�s clearly an issue with the TBB tester. I get similar results as well: nearly every Ookla tester gives me 300 Mbs (single thread) on all 3 of my 330 Mbps lines on BTW tails yet I never get more than 150 Mbps on the TBB tester. I gave up wondering why this is the case a long time ago (probably routing issues to the TBB tester)

You may be comparing apples and pears.

The OP specifically said they did these tests because real-use internet seemed to be running slow. That suggests to me that in this case the tbb result is the more likely to be reflecting the problem.

We know that Ookla kicks out low readings. My results from it nearly always show the highest figure seen during the test, confirmed when using my laptop by my �always-on� non-tbb meter.

Basically on mobile broadband, Ookla shows peak, tbb shows what looks real from visual analysis of that other monitor.

Ignoring the OPs issues, there's clearly an issue with the TBB tester for those on high speed (>200 Mbps) FTTP services in North Scotland (eg Inverness/Aberdeen) whether on BT Wholesale Tails or Vodafone/Cityfibre.

As I said, on three 330 Mbps FTTP lines - all on different ISPs - i'm only seeing speeds of 150 Mbps on TBB yet Ookla gives around 300 Mbps. Real life file download speeds on all three are 40 MB/s which ties in with Ookla results. One of the ISPs (FluidOne) even spent 2 hrs one evening doing NUT TCP tests on my line and yet they also got a constant 40 MB/s. If the TBB tester results are to be believed, then this means my FTTP lines are performing only 50% of their capability 24/7/365 - even taking into account the speed testing methodology differences between Ookla & TBB, I find this extremely hard to believe.

The point I'm trying to make is that results from the TBB tester shouldn't always be treated as gospel, especially since you have loads of different Ookla servers to compare against and perhaps more importantly, real life file download speeds as well. No way on earth would I be getting a constant 40 MB/s file downloads if the line speed was only 150 Mb/s.

Edited by deleted (Mon 06-Jan-20 12:19:44)

You have said all that before. More than once.

I have no problem with your concerns, but you cannot "ignore the OP's issue" when your post is completely irrelevant to it.

You are simply hijacking the thread and providing no help to the OP.

Is your router reporting an error count of any type ?

That sync rate ought to get you better throughput speeds than that I'd have thought.

Mine gets ...

Data rate:
20.00 Mbps / 80.00 Mbps
Maximum data rate:
28663 / 84063
Noise margin:
15.4 dB / 6.4 dB
Line attenuation:
9.3 dB

And just got the following via wifi

https://www.thinkbroadband.com/_assets/speedtest/but...